AN INVESTIGATION 
OF 
ADVANCED OXIDATION 
PROCESSES 
IN 
WATER TREATMENT 
GAVIN W VATI SCHWIKKARD 
MScE ng (Natal); BScEng (Natal) 
Submitted in fulfilment of the academic 
requirements for the degree of 
Doctor or Philosophy 
in the 
School of Chemical Engineering, 
University of Natal , Durban. 
February 2001 
ii 
ABSTRACT 
The deteriorating water quality in South Africa and changing legislation requiring the industrial 
implementation of waste minimisation and poUution prevention technologies has highlighted the need for 
the investigation of new effiuent treatment technologies such as advanced oxidation processes. 
TIlls investigation details the evaluation of ultrasound, an emerging advanced oxidation process, to degrade 
organic compounds during water treatment. The objectives of the investigation included the design of a 
suitable ultrasonic laboratory reactor to investigate ultrasound chemistry and the sub-processes occurring 
during sonication. Atrazine was used as a model compound to compare the perfonnance of ultrasound with 
that of ozone and hydrogen peroxide, already established advanced oxidation processes. Recommendations 
have also been made for the scale·up of ultrasonic processes . 
. A 500 mL ultrasonic cell containing an ultrasonic horn as an energy source was designed and constructed. 
The measurement of hydrogen peroxide concentration was used as a tool to indicate the process conditions 
under which the formation of free radical reactions during sonication are enhanced.. These include the 
application of oxygen and air sparging or the addjtion of a commercial source of hydrogen peroxide. It was 
found that oxygen sparging and a high acoustic p:lwer input should be used in ultrasonic processes with a 
short retention time, and conversely, that air sparging and a lower acoustic energy source should be used in 
processes with a long retention time. A flow loop system should be considered to maximise oxidation both 
within and beyond the sonicated zone, gas sparging should only occur within the sonication zone else the 
degradation of hydrogen peroxide is encouraged. Ultrasound is most effectively applied in water treatment 
as a pretreatment stage in combination ",·ith other technologies and not as a stand-alone process. 
Atrazine was used. as a model compound to compare the performance of ultrasound with ozone because of 
its persistence in the environment and resistance to degradation. Atrazine was degraded during sonication 
and ozonation. degradation increased wim the addition of hydrogen peroxide. OLone decomposition (and 
hence free radical reactions) was enhanced when ozone was combined with ultrasound or hydrogen 
JXroxide. Enhanced ozone decomposition during ozonation combined with sonication is due to the 
conditions (high temperatures and pressures) as well as the free radical reactions occurring within the 
collapsing cavitation bubbles and at the gas-liquid interface. The enhancing effect of combining ultrasound 
with ozone was greatest at the low ozone concentrations typically applied during water treatment. 
Atrazine degradation during sonication and ozonation is predominantly due to the reaction with hydro:\-yl 
radicals. Atrazine degradation products identified using gas chromatography and mass spectrometI)' were 
deethylatrazine. hydroxyatrazine and deethyldeisopropylatnlZine (tentatively identified). 
Hi 
PREFACE 
I, Gavin Schwikkard, declare that unless indicated, this thesis is my own work and that it has not been 
submined, in whole or in part, for a degree at another University or Institution. 
GW SCHWIKKARD 
February 200t 
iv 
ACKNOWLEDGMENTS 
I would like to express my thanks and appreciation to the following people and organisations who have 
contributed. to the investigation: 
The Water Research Commission and Pollution Research Group for funding the investigation. 
Diarec Diamond Sales cc for assistance in acquiring the ultrasonic horn. 
Umgeni Water for the loan of an ozone generator. 
The Foundation for Research Development for personal funding during the investigation. 
Sasol for the postpOnement of my 00rsary obligation. 
My supervisor. Prof. C Buckley. for his dedication. advice and commitment to excellence. 
Mr C Brouckaert for advice on process modelling. 
Prof. 0 Mulholland of the School of Chemistry and Applied Chemistry for the use of HPLC equipment. 
Mr K Robertson and Mr L Henwood from the workshop of the School of Chemical Engineering for the 
construction of the u1trasonic cell and for assistance with equipment maintenance. 
The laboratory technicians, Ms D Harrison,. Ms S WadJey and Mr B Pare!, for their technical support. 
Ms S Freese from Umgeni Water for advice on setting up the Olone system. 
Ms A Hansa from the ML Sultan Technikon for assistance with the identification of degradation products. 
Mr R Coetzer from Sasol Technology for assistance and advice with statistical modelling. 
Mr I Vorster from the Rand Afrikaans University for mass spectroscopy analysis. 
My feUow graduate students for their friendship and experiences we shared learning from onc another. 
My family, David, Dorothy, Mandy and Joanne, for their continual encouragement and support throughout 
the investigation. 
My wife. Sianne. for performing the HPLC atrazine analysis, assisting with the interpretation of mass 
spectra and for the sacrifices she made during the writing of this thesis. 
And to God. the Father. Son and Holy Spiril, who add meaning to life. 
Figures 
Tables 
Nomenclature 
Abbreviations 
Glossary 
1. INTRODUCTION 
I. I Water resources in South Africa 
1.2 Water legislation in South Africa 
1.3 Advanced oxidation technologies 
1.4 Project background 
1.5 Project objectives 
1.6 Thesis outline 
2_ ULTRASOUND 
2.1 History of ultrasound 
2.2 Cavitation 
2.2.1 Fonnation of ultrasonic cavitation 
2.2.2 Classification of cavitation 
2.2.3 Equations of cavitation bubble dynamics 
2.2.4 Cavitation phenomena 
2.2.4. 1 Sonoluminescence 
2.2.4.2 Ca\;t.ation noise 
2.3 Sonochemistry 
2,3.1 Chemical effects 
2.3. 1. 1 Site of sonochemical reactions 
2.3. 1.2 Effects of dissolved gas 
2.3.2 Physical effects 
v 
CONTENTS 
xiii 
xxi 
XXXIII 
~ 
X,.xxvii 
1-1 
1-3 
1-6 
1-8 
1-9 
J-ll 
2-1 
2-2 
2-2 
2-3 
2-5 
2-11 
2-11 
2-16 
2-17 
2- 17 
2- 18 
2-24 
2-26 
Contents conI. 
2.4 Applications of u1trasound 2-28 
2.4.1 PhysicaJ applications 2-28 
2.4.2 ChemicaJ applications 2-29 
2.4.2.1 Organic synthesis 2-31 
2.4.2,2 Sonocata1ysis 2-34 
2.4.2.3 Polymer chemistry 2-36 
2.4.2.4 Sonoelectrochemistry 2-38 
2.4.2.5 Sonocrystallization 2-39 
2.4.3 MedicaJ applications 2-40 
2.4.4 Industrial applications 2-41 
2.4.4.1 Water and eft1uent treatment 2-42 
2.4.4.2 Textile industry 2-48 
2.4.4.3 Food industry 2-49 
2.4.4 .4 Petroleum industry 2-50 
2.4.4.5 Membrane processes 2-51 
2.5 SonochemicaJ equipment 2-51 
2.5.1 Transducers 2-52 
2.5 .1.1 Piezoelectric transducers 2-52 
2.5.1.2 Magnetostrictive transducers 2-53 
2.5.2 Ultrasonic baths 2-53 
2.5.3 Ultrasonic horns 2-56 
2.5.4 Ultrasonic reactors 2-59 
2.5.5 Equipment design and reactor modelling 2.{56 
2.6 Concluding remarks 2-70 
3. OZON}: AND HYDROGEN PEROXIDE CHEMISTRY 
3. 1 O.lOne fundamentals 3-1 
3. 1.1 Chemical and physicaJ properties 3-2 
3.1.2 Classification of ozone reactions 3-3 
3,1.3 Kinetics of ozone decomposition 3-6 
3.2 Ozone in water treatment 3-13 
3.2 .1 History of ozone use 3- 14 
3.2 .2 Ozone generation and transfer 3-15 
Contents cont. 
3.2.3 Application 3-1 9 
3.2.3.1 Potable water treatment 3-1 9 
3.2.3.2 lndustria] effiuent treatment 3-23 
3.3 Hydrogen peroxide 3-24 
3.3.1 Chemical and physical properties 3-26 
3.3.2 Application in water treatment 3-27 
3.4 Concluding remarks 3-29 
4. ATRAZINE CHEMISTRY 
4.1 Characteristics of atrazine 4-1 
4.1.1 Discovery 4-2 
4. 1.2 Chemical and physical properties 4-2 
4. 1.3 Toxicity 4-3 
4.2 Atrazine in the environment 4-4 
4.2.1 Application and usage 4-5 
4.2.2 Atrazine in soil 4-6 
4.2.2.1 Transformation 4-6 
4.2.2.2 Retention 4-1 1 
4.2.3 Atrazine in aquatic systems 4-\3 
4.2.3. 1 Groundwater 4- 13 
4.2.3.2 Surface water 4-14 
4.2.4 Mode of action in plants 4-1 6 
4.3 Auazine in water treatment 4-1 9 
4 .3. 1 Ozonation 4·23 
4.3.2 Ultraviolet radiation 4-33 
4.3.3 Ultrasonic degradation 4-39 
4 .3 .4 Biological treatment 4-41 
4.4 Concluding remarks 4-45 
5. EXPERIMENTAL DESIGN 
5.1 Ultrasonic equipment 5-1 
5. 1.1 Ultrasonic horn 5-1 
Contents cont. 
5. 1.2 Ultrasonic cell 
5. 1.2.1 Frequency 
5.1.2.2 Acoustic power 
5.1.2.3 Horn shape 
5,1.2.4 Volume 
5.1.2.5 Vessel type 
5.1.2.6 Pressure 
5.1.2.7 Mixing 
5.1.2.8 Temperature 
5.1.2.9 Gas saturation 
5. 1.2.10 Sampling 
5.2 <Rone equipment 
5.3 Experimental procedures 
5.4 Analytical procedures 
5.4 , I Hydrogen peroxide measurement 
5.4.2 O.wne measurement 
5.4.1.1 lodometric method 
5.4.1.2 Indigo colorimetric method 
5.4 .3 Dissolved o"'Ygen measurement 
5.4.4 Atrazine measurement 
5.5 Statistical modelling 
6_ ULTRASOUND PROCESS INVESTIGATION 
6.1 Dissolved o"'Ygen concentration 
6.2 Hydrogen peroxide fonnation 
6.2.1 Effect of dissolved gas 
6.2.2 Acoustic power 
6.2.3 Regression analysis 
6.3 Hydrogen peroxide degradation 
6.4 Interval experiments 
6.4.1 Sonication period intervals 
6.4.2 Gas intervals 
viii 
5-4 
5-<5 
5-<5 
5-7 
5-8 
5-9 
5-11 
5-11 
5-12 
5-13 
5-14 
5-14 
5-17 
5-18 
5-1 8 
5-19 
5-1 9 
5-20 
5-20 
5-21 
5-21 
6-2 
6-5 
6-5 
6-8 
6-12 
6-15 
6-17 
6-18 
6-20 
Contents conL 
6.5 Commercial hydrogen peroxide experiments 
6.6 Process chemistry 
6.7 Concluding remarks 
7. OWNE P ROCESS INVESTIGATION 
7.1 Dissolved ozone concentration 
7.2 Ozone decomposition 
7.3 Hydrogen peroxide formation 
7.4 Mass balances 
7.5 Process chemistry 
7.6 Concluding remarks 
8. A TRAZINE EXPERIMENTS 
8.1 Atrazine chemistry 
8.1.1 Ultrasound 
8.1.2 Ozone 
8.1.3 Hydrogen peroxide 
8.1.4 Mass balances 
8.2 Atrazine degradation 
8.2.1 Ultrasound 
8.2.2 Ozone 
8.2.3 Hydrogen peroxide 
8.3 Product identification 
8.4 Concluding remarks 
9. CONCLUDING REMARKS 
REFERENCES 
6-22 
6-27 
6-29 
7-2 
7-8 
7-10 
7-13 
7-16 
7-17 
8-2 
8-3 
8-5 
8-10 
8-12 
8-16 
8-16 
8-19 
8-24 
8-26 
8-34 
Conlmts conL 
APPENDICES 
A. EQUIPMENT CHARACTERISATION 
B. 
A I Ultrasound equipment 
A 1.1 Evaluation of machining the surface of a horn tip 
A 1.2 Acoustic power measurement 
A 1.3 Volume experiments with the low intensity ultrasonic horn 
A 1.4 Acoustic power reduction due to the ultrasonic cell lid design 
A1.5 Mixing 
A 1.6 Rotameter calibration 
A2 Ozone equipment 
ANALYTICAL PROCEDURES 
B. I Hydrogen peroxide measurement 
B.I.I Calibration curve A 
B.1.2 Calibration curve B 
B.1.3 Calibration curve C 
B.2 Ozone measurement 
B.2.1 lodometric method 
B.2.2 Indigo colorimetric method 
8.3 Dissol\'ed ox-ygen measurement 
B.4 Atrazine measurement 
C. ULTRASOUND EXPERIMENTAL DATA 
C.l Dissol\'ed oxygen concentration 
C.2 Hydrogen peroxide formation 
C.3 Hydrogen peroxide degradation 
C.4 Interval experiments 
C.5 Commercial hydrogen peroxide experiments 
A·I 
A· I 
A·3 
A·8 
A·13 
A·IS 
A·16 
A·18 
B·I 
B·I 
B·3 
B·S 
B·7 
B·7 
B·8 
B·8 
B·9 
C· I 
C· 2 
C-4 
C"; 
C-8 
Con1enls conL 
D. OZONE EXPERIMENTAL DATA 
E. 
D.I Dissolved ozone concentration 
0 .2 Ozone decomposition 
0 .3 Hydrogen peroxide fonnation 
0 .4 Mass balances 
ATRAZINE EXPERIMENTAL DATA 
E. I Atrazine chemistry 
E. I . I Ozone 
E.1.2 Hydrogen peroxide 
E.1.3 Mass balanoes 
E.2 Atrazine degradation 
E.2.1 Ultrasound 
E.2.2 Ozone 
E.2.3 Hydrogen peroxide 
E.3 Product identification 
F. STATISTICAL ANALYSIS 
F. I Equipment characterisation 
F. I . I Evaluation of machining the surface of a horn tip 
F.1.2 Acoustic power measurement 
F. l .3 Volume experiments with the low intensity ultrasonic horn 
F. 1.4 Acoustic power reduction due to the ultrasonic cclllid design 
F. I .5 Rotameter calibration 
F. I .6 Ozone generator calibration 
F.2 Analytical procedures 
F.3 Ultrasound experiments 
F.3. 1 Dissolved oxygen concentration 
F.3.2 Hydrogen peroxide formation 
F.3.3 Hydrogen peroxide degradation 
F.3.4 Interval experiments 
F.3.S Commercial hydrogen peroxide e:o..-periments 
[).I 
D-2 
D-2 
D-3 
E-I 
E-I 
E-2 
E-3 
E-4 
E-4 
E-5 
E-8 
E-12 
F-2 
F-3 
F-3 
F-6 
F-9 
F- 1O 
F-Il 
F-II 
F-12 
F-12 
F-14 
F-16 
F-I7 
F-18 
Contents conL 
F.4 
F.5 
Ozone experiments 
F.4.1 Dissolved ozone concentration 
F.4.2 Ozone decomposition 
F.4.3 Hydrogen peroxide formation 
F.4.4 Mass balances 
Atrazine experiments 
F.5.! 
F.5.2 
Atrazine chemistry 
F.5.1.1 Ozone 
F.5.1.2 Hydrogen peroxide 
F.5. 1.3 Mass balances 
Atrazine degradation 
F.5.2.1 Ultrasound 
F.5 .2.2 Ozone 
F.5.2.3 Hydrogen peroxide 
G_ FREE RADICAL REACTIONS 
G.l Free radical chemistry 
G.2 Radical reactions 
H_ SUMMARY OF RADICAL REACTIONS 
H. l Radical reactions: ultrasound 
H.2 Radical reactions: ozone 
Di 
F-20 
F-20 
F-21 
F-22 
F-22 
F-23 
F-23 
F-24 
F-25 
F-25 
F-26 
F-26 
F-27 
F-29 
G-! 
G-3 
H-! 
H-2 
xiii 
FIGURES 
1.I 
1.2 
2.1 
2.2 
2.3 
2.4 
2.5 
2.6 
2.7 
2.8 
Reactor design flow diagram 
F10w diagram of thesis outline 
Radius versus time curves for a gas-filled bubble in water in an ultrasonic field with a 
frequency of 5 MHz (dashed line) or 15 MHz (solid line); PA = 100 kPa, Po = 400 kPa 
and Ro = 0,8 }.lDl 
Pressure distribution in the liquid surrounding a collapsing, gas-filled bubble where 
Z is the volume compression ratio (Rrz/R)3 
Maximum liquid pressured calculated for different ultrasonic frequencies during 
bubble collapse under adiabatic conditions; PA '" 100 kPa and Ro = 3,2 }.lDl 
Sonoluminescence spectrum of argon-saturated water sonicated at 333 kHz 
Cavitation noise spectrum for an ultrasonic resonance frequency (fo) of 15 kHz 
Schematic diagram of the three regions in a cavitating liquid in which chemical 
reactions take place 
Schematic diagram of the origin of cavitational physical effects 
Schematic diagram of the production of end capped JXllystyrene and 
JXlly(styrene-b-methyl methacrylate) block CO-JXllymers 
1-10 
1-11 
2-7 
2-9 
2- 10 
2-14 
2- 17 
2-1 8 
2-26 
2-37 
2.9 Schematic diagram of an ultrasonic bath 2-54 
2. iO Energy conversIOn steps in industrial sonochemisuy 2-55 
2. ) I Schematic diagram of an ultrasonic horn system 2-56 
2.12 Ultrasonic horn shapes 2-57 
2.13 Schematic diagram of a flow cell fined to an ultrasonic horn 2-58 
2. J 4 Ultrasonic reactor for use in batch, semi-batch or continuous-flow operation 2-6 1 
2. 15 Ultrasonic module of the Horn'ell sonochemical reactor 2-62 
2.16 Cross-sections of ultrasonic tube reactors; a) pentagonal, b) hexagonal and c) circular 2-62 
2.17 : Schematic diagram of the cylindrical reactor (with core cooling) designed at the 
University of Milan 
2-<>3 
Figures conL 
2.18 Schematic diagram of a Sodeva Sonirube 
2.19 Schematic diagram of a NearJie/d acoustic processor 
2.20 Schematic diagram of an ultrasonic liquid whistle 
3.1 Resonance structures of the ozone molecule 
3.2 Schematic diagram of ozone concentration in the gas bubble, film layer and. bulk 
solution as described by the two-film theory 
3.3 
3.4 
3.5 
Reaction magram for the ozone decomposition process 
Reaction diagram for the ozone decomposition process in the presence of a solute B 
Reaction diagram for the ozone decomposition process in the presence of hydrogen 
peroxide and a solute B 
riv 
2-64 
2-64 
2-66 
3-1 
3-3 
3-7 
3-9 
3-1 2 
3.6 Double-tube ozone generator 3-15 
3.7 Counter-current bubble column for ozone transfer 3-1 7 
3.8 A static mixer 3-18 
3.9 Chemical structure of Orange 11 dye 3-24 
3.10 Molecular structure of hydrogen peroxide 3-24 
3.11 Commercial production ofhydragen peroxide via lIle 3utCH)xidation of an anthraquinone 3-25 
4.1 
4.2 
4.3 
H 
4.5 
4.6 
4.7 
4.8 
4.9 
compound 
Chemical SU'Ucture of atrazine 
Reaction pathway of the industrial preparation of atrazine 
Primary and secondary metabolites of atrazine 
Microbial degradation ofalrazine by a Nocardia strain 
Mineralisation pathway of atrazine 
Formation of deethylatrazine and deisopropylatrazine from simazine. cyanazine 
and propazine 
Photophosphorylation 
Reaction pathway of auazine hydroxylauon in maize 
Glutathione conjugate of atrazine in maize leaves 
4-1 
4-2 
4-10 
4-11 
4-15 
4-17 
4-18 
4-1 9 
figures conL 
4.10 Acid and alkaline hydrolysis of atrazine 4·20 
4. 11 Reaction mechanism for the formation of dealkylated degradation products from the 4·21 
reaction of atrazine with hydroxyl radicals 
4 .12 Degradation pathway of the ozonation of atrazine 4·23 
4.13 Reaction mechanism fo r the formation of dealkylated degradation products from the 4·24 
direct reaction of atrazine with ozone at acidic pH 
4 .14 Formation pathway of hydrolysed products during atrazine ozonation 
5.1 Ultrasonic process system 
5.2 Ultrasonic horn 
5.3 Ultrasonic cell 
5.4 Schematic diagram of ultrasonic cell 
5.5 Mixing of potassium permanganate crystals in 500 mL of water during sonication 
and in the control without sonication ; (a) after 5 s of sonication and (b) the control 
after 10 min 
5.6 
5.7 
5.8 
5.9 
6.1 
6.2 
6.3 
6.4 
Schematic diagram of the cooling system of the ultrasonic cell 
Schematic diagram of the gas lines 
Schematic diagram of the ozonation experimental setup 
QLone concentration in a 2,4 mL S·l OJl.)'gen gas stream generated by a Sortios ozone 
generator at voltages between 100 and 200 V 
Reduction in dissolved oxygen concentration in oxygen·saturated 'water during 
sonication, without sonication (control) and during nitrogen sparging without 
sonication 
Dissolved oxygen concentration in nitrogen., air· and oxygen·saturated water during 
sonication in the ultrasonic cell 
Hydrogen peroxide formation in nitrogen·, air· and oxygen·saturated water during 
sonication in the ultrasonic cell at an acoustic power of 57 W 
Hydrogen peroxide formation in the control during sonication in the ultrasonic cel1 
at acoustic powers of 24 and 57 W 
4-25 
5-1 
5-2 
5-4 
5-5 
5-12 
5-13 
5-15 
5·16 
6-3 
6-4 
6"; 
6-9 
F&gUTe:J conL 
6.5 
6.6 
6.7 
6.8 
6 .9 
Hydrogen peroxide formation in nitrogen-saturated water during sonication in the 
ultrasonic cell at acoustic powers of 24 and 57 W 
Surface response diagrams of hydrogen peroxide concentration as a function of 
acoustic power and dissolved oxygen concentration during sonication of water in 
the ultrasonic cell 
Hydrogen peroxide degradation in nitrogen-, air- and oxygen-saturated water after 
sonication in the ultrasonic cell for 16 h 
Hydrogen peroxide formation in oxygen-saturated water in the ultrasonic cell during 
15 min of sonication at an acoustic power of 57 W, 15 min without sonication and a 
further 15 miD with sonication 
Hydrogen peroxide formation in water during sonication in the ultrasonic cell at an 
acoustic power of 57 W with 15 min of oxygen saturation, 15 min of nitrogen 
saturation and a further 15 min of oxygen saturation 
6_10 : Hydrogen peroxide formation in nitrogen-, air- and oxygen-saturated 0,28 mg L-t 
hydrogen peroxide solutions during sonication in the ultrasonic cell for I h at an 
acoustic power of 57 W 
6. 11 : Hydrogen peroxide formation in nitrogen-, air- and oxygen-saturated 0,43 mg L-t 
hydrogen peroxide solutions during sonication in the ultrasonic cell for I h at an 
acoustic power of 57 W 
6. 12 : Hydrogen peroxide formation in nitrogen-, air- and oxygen-saturated 0,72 mg L'] 
hydrogen peroxide solutions during sonication in the ultrasonic cell for 1 h at an 
acoustic power of 57 W 
6.13 : Rates of hydrogen peroxide formation in 0,28; 0,43 and 0,72 mg L' \ commercial 
hydrogen peroxide solutions saturated with nitrogen, air and oxygen during sonication 
in the ultrasonic cell for I h at an acoustic power of 57 W 
7. 1 
7.2 
Dissolved ozone concentration in water during ozo03tion (0,014 mg s·\), O2003lion 
combined with sonication (57 W), ozonation combined with hydrogen peroxide 
(40 mg L-l) and ozonation combined with sonication and hydrogen peroxide 
Surface response diagrams of dissolved ozone concentration in water during 
ozonation as a function of acoustic power and hydrogen peroxide concentration 
6-10 
6-13 
6-16 
6-21 
6-23 
6-24 
6-24 
6-26 
7-3 
7-7 
Figures conL 
7.3 
7.4 
7.5 
8.1 
8.2 
8.3 
8.4 
8.5 
8.6 
8.7 
8.8 
8.9 
Dissolved ozone concentration in water in the ultrasonic cell after a 20 min saturation 
period during ozonation (0,014 mg sO'), sonication (57 W) and ozonation combined 
with sonication 
Hydrogen peroxide concentration in water during ozonation (0,014 mg 5-1), sonication 
(57 W) and ozonation combined with sonication 
Surface response diagrams of hydrogen peroxide concentration as a function of acoustic 
power and ozone concentration during ozonation of water in the ultrasonic cell 
Alrazine concentration in a 5 mg L-1 atrazine solution in the ultrasonic cell over 3 h in 
the absence of ultrasound without and with oxygen sparging 
Formation of dealkylated degradation products from the reaction of attazine with 
hydroxyl radicals 
Formation of hydrolysed degradation products from the reaction of atrazine with 
hydro:\:yl radicals 
Dissolved ozone concentration in water and a 5 mg L- 1 atrazine solution during 
ozonation (0,014 mg 5-\ ozonation combined with sonication (57 W), ozonation 
combined with hydrogen peroxide (40 mg L-1) and ozonation combined with 
sonication and hydrogen peroxide 
Formation of dealkylated degradation products from the direct reaction of atrazine 
with ozone 
Hydrogen peroxide concentration in water and a 5 mg L-1 atrazine solution in the 
ultrasonic cell during ozonation (0,014 mg 5-\ sonication (57 W) and ozonation 
combined with sonication 
Alrazine concentration in a nominal 5, 10 and 20 mg L-1 atrazine solution during 
sonication in the ultrasonic cell at an acoustic power of 57 W for 3 h 
Effect of oxygen and nitrogen sparging on the degradation of atrazine in a 5 mg L-\ 
atr3Zine solution during sonication in the ultrasonic cell at an acoustic power of 57 W 
over 3 h 
Atrazine concentration in a 5 mg L _1 atrazine solution during ozonation over 3 h at 
ozone production rates of 0,003 ; 0.014 ; 0,030 and 0,047 mg 5·\ 
xvii 
7-9 
7-10 
7-12 
8-2 
8-4 
8-5 
8-{; 
8-7 
8-10 
8-16 
8-18 
8-19 
Fllfures conL 
8.10 First order atraz.ine degradation in a 5 mg L·t atrazine solution during ozonation over 8·21 
3 h at ozone production rates of 0,003; 0,014; 0,030 and 0,047 mg s'\ 
8.11 Atrazine concentration in a 5 mg L'\ atrazine solution during ozonation over 3 hat 8-21 
ozone production rates of 0,003; 0,0]4; 0,030 and 0,047 mg s'\ combined with 
sonication at an acoustic power of 57 W 
8.12 First order atrazine degradation in a 5 mg L·1 atrazine solution during ozonation over 8-23 
3 h at ozone production rates of 0,003; 0,014; 0,030 and 0,047 mg s'\ combined with 
sonication at an acoustic power of 57 W 
8.13 Comparison of atrazine degradation rates in a 5 mg L·t atrazine solution during 
ozonation and ozonation combined with sonication 
8.14 Atrazine concentration in a 5 mg L'\ atrazine solution with 40 mg L' \ hydrogen 
peroxide, hydrogen peroxide combined with sonication (57 W). hydrogen peroxide 
combined with ozonation (0,014 mg s'\) and hydrogen peroxide combined with 
sonication and ozonation 
8-23 
8-24 
8. 15 HPLC chromatograms for the ozonation ofa 5 mg L'\ atrazine solution at an ozone 8-27 
production rate of 0,047 mg L' \ 
8.16 : HPLC chromatograrns for the ozonation ofa 5 mg L· t atrazine solution at an ozone 8-28 
production rate of 0.047 mg L'\ combined with sonication at an acoustic power of 57 W 
8. 17 HPLC chromatograms at 180 min for a 5 mg Lot atrazi ne solution during ozonation 8-28 
(0,0 14 mg L'\) combined with hydrogen peroxide (40 mg L·I ) and ozonation 
combined with both sonication (57 W) and hydrogen peroxide 
8.18 Mass spectrum of a 5 mg l ·1 alrazme solution 
8.19 Molecular fragments generated during the ionisation of atrazine 
8.20 Mass spectrum of deethylatrazine formed during the sonication and ozonation of a 
5 mg L-1 atrazine solution for 3 h 
8.2 1 Molecular fragments generated during the ionisation of deethylatrazine 
8-29 
8-30 
8-31 
8-32 
Figures cont. 
8.22 Mass spectrum of hydroxyatrazine formed during the sonication, ozonation and 
sonication combined with ozonation of a 5 mg L- 1 atrazine solution for 3 h 
8.23 Compounds fonned during the ionisation of hydroxyatral.ine due to the ~-cleavage 
of the alipbatic amine side chains 
xix 
8-32 
8·33 
8.24 : Mass spectrum of deethyldeisopropylatrnzine formed during the sonication, ozonation 8-34 
Al 
A2 
A3 
and sonication combined with ozonation of a 5 mg L-1 atrazine solution for 3 h 
Hydrogen peroxide formation with a new and a machined horn tip 
Schematic diagram of the insulated vessel used in calorimetric experiments 
Temperature measurements in 500 mL of water in an insulated vessel during 
sonication over 40 min at transducer displacement amplitudes of 5. 8 and 11 }UtI 
A4 : Hydrogen peroxide concentration in water volumes of 100 to 1 000 mL during 
sonication with the low intensity ultrasonic horn 
A5 
A6 
Relationship between u1trasonic intensity and water volume during sonication with 
the Iov,' intensity ultrasonic horn 
Relationship between initial rate of hydrogen peroxide fonnation and u1trasonic 
intensity during sonication with the low intensity ultrasonic horn 
A 7 : Comparison of temperature profiles measured during sonication with the low 
i ntensi~' horn in 500 mL of water in the ultrasonic cell and in a beaker 
A.8 
A9 
Mixing of potassium permanganate crystals in 500 mL of water during sonication 
with the high intensity horn: (a) after 2 s and (b) after 5 s 
Mixing of potassium permanganate crystals in 500 mL of water without sonication; 
(a) after 10 min and (b) after 60 min 
A·2 
A·5 
A·5 
A·8 
A-I2 
A-I3 
A-14 
A-16 
A· 16 
A 10 : Rotameter calibration for oxygen at regulator pressures of 200 and 300 kPa A-17 
All : Ozone concentration in a 2.4 mL S·I oX1'gen gas stream generated by a Sorbios ozone A-19 
B.I 
B.2 
8.3 
generator gas stream at voltage settings between 100 and 200 V 
Hydrogen peroxide calibration curve A 
Hydrogen peroxide calibration curve B 
Hydrogen peroxide calibration curve C 
B-2 
B-4 
B-6 
Figures cont. 
8 .4 Solubility of oxygen in water at 101,325 kPa and temperatures between ° and 40 °C 8-8 
8 .5 Alrazine calibration curve 8-9 
E. I Atrazine degradation in a 5 mg L-I auazine solution over 3 h during ozonation E-7 
(0,003 mg S-I), sonication (57 W) and ozonation combined with sonication 
E.2 Atrazine degradation in a 5 mg L'I atrazine solution over 3 h during ozonation E-7 
(0,0 14 mg s'\ sonication (57 W) and ozonation combined with sonication 
E.3 Atrazine degradation in a 5 mg L't atrazine solution over 3 b during ozonation E-8 
(0,030 mg S'I), sonication (57 W) and ozonation combined with sonication 
E.4 : Atrazine degradation in a 5 mg L'l atrazine solution over 3 h during ozonation 
(0,047 mg s'\ sonication (57 W) and ozonation combined with sonication 
E.5 
E.6 
E.7 
E.S 
Fraction ofundegraded auazine in a 5 mg L't atrazine solution after ozonation 
(0,0 14 mg S·I) for 45 min with hydrogen peroxide concentrations of 1, 5, 10, 25 and 
50 mgL,1 
Atrazine degradation in a 5 mg L'I atJazine solution with 40 mg L· I hydrogen 
peroxide. sonication (57 W) and sonication combined with hydrogen peroxide 
Atrazine degradation in a 5 mg L't atrazine solution with 40 mg L'I hydrogen 
peroxide. ozonation (0.01 4 mg S·I) and ozonation combined with hydrogen peroxide 
Atrazine degradation in a 5 mg L'I solution with 40 mg, L·I hydrogen peroxide. 
sonication (57 W) combined with ozonation (0,0 14 mg S'I) and sonication combined 
with ozonation and hydrogen peroxide 
E.9 : HPLC chromatogram ora 5 mg L'I atrazine solution 
E. IO : GC-MS chromatograms ofa 5 mg L'I atrazine solution after sonication (57 W). 
ozonation (0,047 mg S'I) and sonication combined with ozonation for 3 b 
E-S 
E-9 
E-Il 
E-ll 
E-1 2 
E-12 
E-14 
1.I 
1.2 
I.3 
1.4 
2.1 
1'.2 
2.3 
2.4 
2.5 
2.6 
3. 1 
3.2 
3.3 
3.4 
3.5 
4.1 
4.2 
4.3 
4.4 
Classification of the water situation in a country 
Estimates of South African water demand per sector in 1990, 2000 and 2010 
Standard oxidation/reduction potentials at 25 °C and 101,3 kPa of selected oxidising 
agents 
Rate constants for the reaction between hydroxyl radicals and selected compounds 
Effect of gas thennal heat conductivity on actual gas temperature in a cavity and 
on the measured sonoluminescence intensity 
Theoretical maximum temperature in a collapsing cavity calculated from 
mathematical hydrodynamic models 
Applications of ultrasound in water and effluent treatment 
Effect of reaction solution volume and size and shape of reaction vessel on 
sonochemica1 yield 
Parameters that affect cavitation 
Methods of reporting ultrasonic intensity 
Physical properties of ozone 
Classification of kinetic regime of gas absorption reactions according to the Hatta 
Dumber 
Classification of kinetic regime of gas absorption reactions according to diffusion 
and reaction limes 
Some kinetic studies of ozone decomposition in water 
Physical properties of hydrogen peroxide 
Chemical and physica1 properties of atrazine 
Acute toxicity of atrazine 
Goodfarming praclice programme for atrazine 
Second order rate constants for the reaction of atrazine and hydro:>,:)'1 radicals 
TABLES 
I-I 
1-2 
1-7 
1-7 
2- 13 
2-19 
2-48 
2-54 
2-66 
2-67 
3-2 
3-4 
3-5 
3-10 
3-26 
4-2 
4-3 
4-5 
4-22 
Tables conL 
4.5 
4.6 
4.7 
Rate constants for the reaction of hydroxyl radicals with atrazine degradation 
products at 20 QC and pH "'" 7,5 to 8,1 
Rate constants for the direct reaction between ozone and atrazine at pH 2, 7 and 12 
Hatta numbers and kinetic regimes of ozone absorption for the direct reaction 
between ozone and atrazine at pH 2, 7 and 12 
uii 
4-22 
4-26 
4-27 
4.8 Estimated contact times for advanced oxidation with ozone and hydrogen peroxide 4-30 
4.9 Quantum yjeldof atrazine 4-36 
4. 10 Removal of triazine herbicides with mixed liquor with different concentrations of 4-41 
suspended solids 
4.11 Effect of ozone pretreatment on atrazine biodegradation in soil columns after 50 d 4-43 
5.1 Conversion of electrical power to acoustic power 5-4 
5.2 Acoustic jXlwer and intensity of the U1trasonic process system 5-7 
5.3 Comparison of the ultrasonic intensity of the Ultrasonic process system with that of 5-7 
other sonochemical investigations using an ultrasonic horn 
5.4 
5.5 
5.6 
6 .1 
6.2 
6 .3 
6.4 
Comparison of the sample volume that can be sonicated in the ultrasonic cell with 
sample volumes of other sonochemica1 investigations 
Comparison of the ultrasonic cell with vessel types and volumes of other sonochemical 
investigations using an ultrasonic horn 
Average ozone concentration and weight percentage of ozone in a 2,4 mL S· 1 ox)'gen 
gas stream generated by the Sorbios ozone generator at voltages of 100, 130, 150 and 170 V 
E:-..-perimentai programme of the ultrasonic process investigation 
Rate of hydrogen peroxide formation in nitrogen-, a ir- and oxygen-saturated water 
during sonication in the ultrasonic cell for 20 miD at an acoustic power of 57 W 
Rate of hydrogen peroxide fonnation in nitrogen-saturated water and the control 
during sonication in the ultrasonic celJ at acoustic powers of24 and 57 W for 20 min 
Rate ofhydragen peroxide degradation over 3h in nitrogen-, air- and oxygen-saturated 
water after sonication in the ultrasonic cell for 16 h 
5-9 
5-10 
5-17 
6-1 
6-7 
6-11 
6-1 6 
Tables conL 
6 .5 
6.6 
6.1 
6.8 
6.9 
Rate of hydrogen peroxide formation in oxygen-saturated water and a control in the 
ultrasonic cell during 15 min of sonication at an acoustic power of 57 W, 15 min 
without sonication and a further 15 min with sonication 
Regression coefficients for the model of hydrogen peroxide formation in oxygen 
saturated water and a control during 15 min of sonication at an acoustic power of 
57 W, 15 min without sonication and a further 15 min with sonication 
Rate of hydrogen peroxide formation in water during sonication in the ultrasonic cell 
at an acoustic iX'wer of 57 W with 15 m.in of oxygen saturation, 15 min of nitrogen 
saturation and a further 15 min of oxygen saturation 
Regression coefficients for the model of hydrogen peroxide formation in water during 
sonication in the ultrasonic cell at an acoustic power of 57 W with 15 min of oxygen 
saturation, 15 min of nitrogen saturation and a further 15 min of oxygen saturation 
Rate of hydrogen peroxide formation in 0,28; 0,43 and 0,72 mg L" hydrogen peroxide 
solutions saturated with nitrogen, air and oxygen during sonication in the ultrasonic 
cell at an acoustic power of 57 W 
xxiii 
6-19 
6-20 
6-21 
6-22 
6-25 
6.10 Bond dissociation energies 6-28 
7.1 Experimental programme of the ozone process investigation 7-1 
7.2 Rate of hydrogen peroxide formation in water in the ultrasonic cell during ozonation 7-11 
(0,0 14 mg 5.1), sonication (57 W) and O2onation combined with sonication 
1.3 
1.4 
8.1 
8.2 
8.3 
Ozone dccomJX>Sition in water over 45 min during O2003tion, O2onation combined 
\\;th sonication. O2o03tion combined with hydrogen peroxide and ol onation 
combined with sonication and hydrogen peroxide 
Unrcactcd ozone in the solution, gas-phase and KI traps at the termination of the 
45 min experiments as a fraction (%) of the generated ozone during the mass balance 
experiments 
Experimental programme of the atrazine investigation in the ultrasonic cell 
Rate of atrazine degradation in a 5 mg L-' atrazine solution in the ultrasonic cell 
over 3 h in the absence of ultrasound without and with oxygen sparging 
Rate of hydrogen peroxide formation in water and a 5 mg L-I atrazine solution during 
olo03tion (0,014 mg 5-'), sonication (57 W) and ozonation combined with sonication 
1-15 
8-1 
8-3 
Tables conL 
8.4 
8.5 
8.6 
8.7 
8.8 
8.9 
OLone decomposition and atrazine oxidation in a 5 mg L- 1 atrazine solution over 
45 miD during ozonation, ozonation combined with sonication, ozonation combined with 
hydrogen peroxide and ozonation combined with sonication and bydrogen peroxide 
Unreacte<1 ozone in solution, gas-phase and KJ traps at the termination of the 45 min 
experiments as a fraction (0/0) afthe generated ozone during the atrazine mass 
balance experiments 
Rate ofatrazine degradation in a nominal 5, 10 and 20 mg L-1 atrazine solution 
during sonication in the ultrasonic ceU at an acoustic power of 57 W over 3 h 
Rate ofatrazine degradation in a 5 mg LO ' atrazine solution during sonication with 
nitrogen and oxygen sparging in the ultrasonic cell for 3 h at an acoustic power 
of 57 W 
Initial rate of atrazine degradation in a 5 mg L-' atrazine solution during ozonation 
at ozone production rates 0[0,003; 0,014; 0,030 and 0,047 mg S·I 
Initial rate of atrazine degradation in a 5 mg L-1 atrazine solution during ozonation 
over 3 h at ozone production rates ofO,003 ; 0,014; 0,030 and 0,047 mg s' \ combined 
with sonication at an acoustic power of 57 W 
8. 10 Initial rate of atrazine degradation in a 5 mg L'\ atrazine solution with 40 mg L'I 
hydrogen peroxide, hydrogen peroxide combined with sonication (57 W). hydrogen 
AI 
A2 
A.3 
A4 
A5 
A6 
peroxide combined with ozonation (0, 14 mg S' I) and hydrogen peroxide combined 
with sonication and ozonation 
Hydrogen peroxide formation with a new and a machined horn tip 
Rate of hydrogen peroxide formation with a new and a machined horn tip over 20 min 
Calculated acoustic JX)wer during the sonication of 500 mL of water at transducer 
displacement amplitudes of 5, 8 and 1I ~ 
Conversion of electrical power to acoustic power 
Temperature measurements in 500 mL of water in an insulated vessel during sonication 
over 40 min at transducer displacement amplitudes of 5. 8 and 11 J.1m 
Comparison of acoustic JX)wer calculated from calorimeuy with acoustic JXlwer 
calculated from the power drawn by the transducer when loaded and Wl.1oaded 
IDV 
8-13 
8-14 
8-17 
8-18 
8-20 
8-22 
8-25 
A-2 
A-3 
A-3 
A-4 
A-6 
A-7 
Tables cottL 
A 7 Hydrogen peroxide concentration in water volumes of 100 to 1 000 mL during sonication A-S 
with the low intensity ultrasonic born 
AS : Mass of hydrogen peroxide generated in water volumes of lOO to 1 000 mL during 
sonication with the low intensity ultrasonic horn 
A 9 : Rate of hydrogen peroxide formation in water volumes of 100 to I 000 mL during 
sonication with the low intensity ultrasonic horn 
A 10 : Temperature profiles over 10 min in water volumes of 100 to 1 000 mL during 
sonication with the low intensity ultrasonic horn 
A-9 
A-IO 
A-II 
All : Acoustic power and ultrasonic intensity in water volumes of 100 to 1 000 mL during A-12 
sonication with the low intensity ultrasonic horn 
AI2 : Temperatures measured during sonication with the low intensity horn in 500 mL of A-14 
water in the ultrasonic cell and in a beaker 
A 13: Comparison of the gradients of the temperature profiles measured in 500 mL of water A-15 
during sonication with the low intensity horn in the ultrasonic cell and in a beaker 
A 14 : Comparison of acoustic power transferred during sonication with the low intensity horn A-I5 
in 500 mL of water in the ultrasonic cell and in a beaker 
A 15 : Rotameter calibration for oxygen at regulator pressures of 200 and 300 kPa A-17 
A 16 : Rotameter calibration for nitrogen at a regulator pressure of 200 kPa A-IS 
A 17 : Ozone concentration in a 2,4 mL S·I oxygen gas stream generated by a Sorbios ozone A-19 
generator at voltage senings between 100 and 200 V 
A 18 : Weight percentage of ozone in a 2,4 mL S· I o:\)'gen gas stream generated by a Sorbios A-20 
ozone generator at voltage settings between 100 and 200 V 
B. 1 Data for hydrogen peroxide calibration curve A B-3 
8.2 Data for hydrogen peroxide calibration curve 8 B-5 
B.3 Data for hydrogen peroxide calibration curve C 8-7 
B.4 Solubility of ox)'gen in water at 101 ,325 kPa and temperatures between 0 and 40 °C 8 -9 
B.5 Atrazine calibration curve data 8-lO 
Tables cont. 
Cl 
C2 
C3 
C4 
C.5 
C.6 
C7 
Dissolved oxygen concentration in oxygen-sarurated water during sonication, without 
sonication (control) and during nitrogen sparging without sonication 
Dissolved oxygen concentration in water saturated with air 
Dissolved oxygen concentration in nitrogen-, air- and oxygen-saturated water during 
sonication in the ultrasonic cell at an acoustic power of 57 W 
Hydrogen peroxide formation in nitrogen-, air- and oxygen-saturated water during 
sonication in the ultrasonic cell for 16 h at an acoustic power of 57 W 
Hydrogen peroxide formation in nitrogen-saturated water and the control during 
sonication in the ultrasonic cell at an acoustic power of 24 W 
Hydrogen peroxide degradation (normalised data) in nitrogen-, air- and oxygen 
saturated water after sonication in the ultrasonic cell for 16 h 
Hydrogen peroxide degradation in nitrogen-, air- and oxygen-saturated water after 
sonication in the ultrasonic cell for 16 h at an acoustic power of 57 W 
C-I 
C-2 
C-2 
C-3 
C-4 
C-5 
C-6 
C.8 : Hydrogen peroxide formation in oxygen-saturated water and a control in the ultrasonic C-7 
cell during 15 min of sonication at an acoustic power of 57 W, 15 min without 
sonication and a funher 15 min with sonication 
C.9 : Hydrogen peroxide formation in water during sonication in the ultrasonic cell at an 
acoustic power of 57 W with 15 min of oxygen saturation. 15 min of nitrogen 
saturation and a funher 15 min of o)':ygen saturation 
C. IO : Hydrogen peroxide formation in nitrogen-, air- and oxygen-saturated 0,28 mg L· 1 
hydrogen peroxide solutions during sonication in the ultrasonic cell for I h 
Cll Hydrogen peroxide formation in nitrogen-, air- and oxygen-saturated 0,43 mg L· I 
hydrogen peroxide solutions during sonication in the ultrasonic cell for I h 
C.12 : H~'drogen peroxide formation in nitrogen-, air- and oxygen-saturated 0,72 mg L·1 
hydrogen peroxide solutions during sonication in the ultrasonic cell for I h 
D. I Dissolved ozone concentration in water during ozonation (0,0 14 mg 5.1) , ozonation 
combined with sonication (57 W), ozonation combined with hydrogen peroxide 
(40 mg Lol) and ozonation combined with sonication and hydrogen peroxide 
C-s 
C-9 
C-9 
C-IO 
D-I 
Tablt!S cam 
D.2 Dissolved ozone concentration in water in the ultrasonic cell after a 20 min saturation 
period during ozonation (0,014 mg s'\ sonication (57 W) and ozonation combined 
with sonication 
0 .3 : Hydrogen peroxide concentration in water in the ultrasonic cell during ozonation 
(0,014 mg 5,1), sonication (57 W) and ozonation combined with sonication 
D.4 
D.5 
D.6 
E. l 
E.2 
E.3 
E.4 
E.5 
Amount of OZOne generated in 45 min in the ozonation, ozonation combined with 
sonication. ozonation combined with hydrogen peroxide and ozonation combined 
with sonication and hydrogen peroxide mass balance experiments 
Volume of the gas spaces in the ultrasonic cell experimental setup 
Ozone and hydrogen peroxide measurements in water after the 45 min ozonation. 
ozonation combined with sonication, ozonation combined with hydrogen peroxide, and 
ozonation combined with sonication and hydrogen peroxide mass balance experiments 
AlraZine concentration in a 5 mg L'I atrazine solution in the ultrasonic cell over 3 h 
in the absence of ultrasound without and with oxygen sparging 
Dissolved ozone concentration in a 5 mg L'I atrazine solution during ozonation 
(0.0 14 mg 5.1) , ozonation combined with sonication (57 W), ozonation combined 
with hydrogen peroxide (40 mg L- I ) and ozonation combined with sonication and 
hydrogcn peroxide 
Hydrogen peroxide concentration in a 5 mg L'I atrazine solution in the ultrasonic 
cell during ozonation (0,014 mg 5'1), sonication (57 W) and ozonation combined 
v.ith sonication 
Ozone_ hydrogen peroxide and atrazine measurements in a 5 mg L-I atrazine solution 
after the 45 min ozonation. ozonation combined with sonication. ozonation combined 
with hydrogen peroxide and ozonation combined with sonication and hydrogen 
peroxide mass balance experiments 
Atrazine concentration in a 5, 10 and 20 mg L -I atrazine solution during sonication 
in the ultrasonic cell at an acoustic power of 57 W over 3 h 
xxvii 
0-2 
D-3 
D-4 
D-4 
0-5 
E-l 
E-2 
E-3 
E-4 
E-5 
Tables conL 
E.6 Effect of oxygen and nitrogen sparging on the degradation of atrazine in a 5 mg L- 1 
atrazine solution during sonication in the ulttasonic cell at an acoustic power of 
57 W over 3 h 
E .7 : Atrazine concentration in a 5 mg L-1 atrazine solution during ozonation over 3 h 
at ozone production rates of 0,003; 0,014; 0,030 and 0,047 mg S-I 
E.8 : Atrazine concentration in a 5 mg L-' atrazine solution during ozonation over 3 h 
at ozone production rates ofO,003; 0,014; 0,030 and 0,047 mg s'\ combined with 
sonication at an acoustic power of 57 W 
E.9 Fraction ofundegraded atrazine in a 5 mg L· 1 atrazine solution after ozooation 
(0,014 mg S-I) for 45 min with hydrogen peroxide concentrations of I , 5, 10, 25 and 
50 mgL-1 
E.lO Atrazine concentration in a 5 mg L-1 atrazine solution with 40 mg L·I hydrogen 
peroxide, hydrogen peroxide combined with sonication (57 W), hydrogen peroxide 
combined with ozonation (0,014 mg S·I) and hydrogen peroxide combined with 
sonication and ozonation 
F. l 
F.2 
F.3 
FA 
F.S 
Statistical analysis of the formation of hydrogen peroxide during sonication with a 
new and a machined horn tip (Table A I ; Figure A I) 
Statistical analysis of the measured acoustic power of the sonication experiments 
(Table A3) 
Slatistical analysis of the measured temperatures in 500 mL water in an insulated 
vessel during sonication at transducer displacement amplitudes of 5, 8 and 1I ~m 
(Table A.5; Figure A 3) 
Statistical anaJysis of the formation of hydrogen peroxide in water volumes of 
100 to I 000 mL during sonication with the low intensity ultrasonic horn 
(Table A.7; Figure AA) 
Statistical analysis of the temperature measurements in water volumes of 
100 to I 000 mL during sonication with the low intensity ultrasonic horn 
(Table AIO) 
uviii 
E-S 
E-6 
E-6 
E-9 
E-IO 
F-3 
F-3 
F-4 
F-6 
F-8 
Tables conL 
F.6 : Statistical analysis of the comparison of measured temperatures during sonication 
with the low intensity horn in SOO mL of water in the ultrasonic cell and in a beaker 
(Table All; Figure A7) 
F.7 Statistical analysis of the rotameter calibration for oxygen at regulator pressures of 
200 and 300 kPa (Table A 15; Figure A 10) 
F ,8 : Statistical analysis of the rotameter calibration for nitrogen at a regulator pressure 
0[ 200 kPa (Table A16) 
F.9 Statistical analysis of the dissolved ozone concentration in solution during ozonation 
with ozone generated by a Sorbios ozone generator in a 2,4 mL s'] oxygen gas stream 
at voltage settings between 100 and 200 V (Table A 17; Figure All) 
nix 
F-9 
F-IO 
F-Il 
F-Il 
F. lO Statistical analysis of the atrazine calibration curve (Table 8 .5; Figure B.5) F·II 
F . 11 Statistical analysis of the dissolved oxygen concentration in oxygen-saturated water F-12 
during sonication,. without sonication (control) and during nitrogen sparging without 
sonication (Table C.I ; Figure 6 .1) 
F. 12 : Statistical analysis of the dissolved oxygen concentration in water saturated with air F-13 
(fable C.2) 
F.1 3 : Statistical analysis of the dissolved o:o.1'gen concentration in nitrogen-, air- and 
oxygen-saturated water during sonication in the ultrasonic cell at an acoustic power 
of 57 W (fable C.3: Figure 6 .2) 
F-1 3 
F.14 Overall statistical analysis of the dissolved oX1'gen concentration in nitrogen-, air- and F-1 3 
o,,:ygen-saturoted water during sonication in the ultrasonic cell (Table C.3; Figure 6.2) 
F. 15 Statistical analysis of the formation of hydrogen peroxide in nitrogen-, air- and 
ox·ygen- saturated water during sonication in the ultrasonic cell for 16 h at an acoustic 
power of 57 W (fable C.4; Figure 6.3) 
F-14 
F. 16 : Statistical analysis of the formation of hydrogen peroxide in nitrogen-saturated water F-15 
and the control during sonication in the ultrasonic cell at an acoustic power of 24 W 
(Table C.5; Figure 6.4; Figure 6.5) 
F.17 : Statistical analysis of the degradation of hydrogen peroxide (normalised data) in 
nitrogen-, air- and o,,:ygen-saturated water after sonication in the ultrasonic cell for 
16 h at an acoustic power of 57 W (Table C.6: Figure 6.7) 
F-16 
== 
Tables conL 
F. J 8 : Statistical analysis of the formation of hydrogen peroxide in oxygen-sarurated water F-17 
and a control in the ultrasonic cell during 15 min of sonication. 15 min without 
sonication and a further 15 min with sonication (fable C.S; Figure 6.8) 
F. 19 : Statistical analysis of the formation of hydrogen peroxide in water during sonication F-17 
in the ultrasonic cell with 15 min of oxygen saturation, 15 min of nitrogen saturation 
and a further 15 min of oxygen saturation (fable G.9; Figure 6.9) 
F.2D : Statistical analysis of the formation of hydrogen peroxide in nitrogen-, air- and 
oA"Ygen-saturated 0,28 mg L-1 hydrogen peroxide solutions during sonication in the 
u1trasonic cell for I h at an acoustic power of 57 W (fable C. lO; Figure 6.10) 
F.2 1 Statistical analysis oftbe formation of hydrogen peroxide in niuogen-. air- and 
oxygen-sarurated 0,43 mg L-1 hydrogen peroxide solutions during sonication in the 
uluasonic cell for 1 h at an acoustic power of 57 W (Table C. ll ; Figure 6. 11 ) 
F- 18 
F-18 
F.22 : Statistical analysis of !.he formation of hydrogen peroxide in nitrogen·, air- and F·19 
oxygen-saturated 0,72 mg L· I hydrogen peroxide solutions during sonication in the 
ultrasonic cell for 1 h at an acoustic power of 57 W (fable C.12; Figure 6.12) 
F.23 Statistical analysis of the measured dissolved ozone concentration in water during 
ozonation (0,0 14 mg s"). ozonation combined with sonication (57 W). ozonation 
combined with hydrogen peroxide (40 mg L'I) and ozonation combined with sonication 
and hydrogen peroxide (Table 0 .1; Figure 7.1) 
F-20 
F.24 Statistical analysis of the dissolved ozone concentration in water in the ultrasonic cell F-2 1 
after a 20 min saturation period during ozonation (0,014 mg S·I ). sonication (57 W) 
and ozonation combined with sonication (Table 0 .2; figure 7.3) 
F.25 : Statistical analysis of lhe formation of hydrogen peroxide in water in the ultrasonic F-22 
ceD during ozonalion (0.0 14 mg S·I), sonication (57 W) and ozonation combined 
'with sonication (fable 0 ,4; Figure 7,4) 
F.26 : Statistical analysis of the measured ozone input in the 45 min ozonation. ozonation F-22 
combined with sonication. ozonation combined with hydrogen peroxide and ozonation 
combined with sonication and hydrogen peroxide mass balance experiments (fable 0 ,4) 
Tables corrL 
F.27 Statistical analysis of the ozone and hydrogen peroxide measurements in water after F-23 
the 45 min ozonation, ozonation combined with sonication, ozonation combined with 
hydrogen peroxide, and ozonation combined with sonication and hydrogen peroxide 
mass balance experiments (Table 0 .6) 
F .28 Statistical analysis of the atrazine concentration in a 5 mg L-I atrazine solution in 
the ultrasonic cell over 3 b in the absence of ultrasound without and with oxygen 
sparging (fable E. I ; Figure 8. 1) 
F.29 Statistical analysis of the measured dissolved ozone concentration in a 5 mg L- 1 
atrazine solution during ozonation (0,014 mg S-I), ozonation combined with 
sonication (57 W), ozonation combined with hydrogen peroxide (40 mg L- I ) and 
ozonation combined with sonication and hydrogen peroxide (Table E .2; Figure 8.4) 
F.30 Statistical analysis of the formation of hydrogen peroxide in a 5 mg L-1 atrazine 
solution in the ultrasonic ceU during ozonation (0,014 mg s·\ sonication (57 W) 
and ozonation combined with sonication (Table E .3 ; Figure 8.6) 
F .3 1 : Statistical analysis of the ozone, hydrogen peroxide and atrazine measurements in a 
5 mg L-1 atrazine solution after the 45 min ozonation, ozonation combined with 
sonication, ozonation combined with hydrogen peroxide and ozonation combined 
with sonication and hydrogen peroxide mass balance experiments (Table E.4) 
F.32 : Statistical analysis of the atrazine concentration measured in a 5, 10 and 20 mg L-1 
atrazine solution during sonication in the ultrasonic cell at an acoustic power of 
57 W over 3 h (Table E.5: Figure 8.7) 
F.33 Statistical analysis of the initial concentration of the 5 mg L-1 atrazine solutions 
used in the atrazine investigation 
F.34 : Statistical analysis of the effect of oxygen and nitrogen sparging on the degradation 
of atrazine in a 5 mg L-I atrazine solution during sonication in the ultrasonic cell at 
an acoustic power of 57 W over 3 h (Table E.6; Figure 8.8) 
F.35 Statistical analysis ofthe atrazine concentration in a 5 mg L-I atrazine solution 
during ozonation al ozone production rates of 0,003 ; 0,014; 0.030 and 0,047 mg S-I 
(Table E.7; Figure 8.9) 
F·24 
F·25 
F-25 
F·26 
F·27 
F·27 
Tables conL 
F.36 : Statistical analysis oftbe atrazine concentration in a 5 mg L'! atrazine solution 
during ozonation over 3 h at ozone production rates of 0,003; 0,014; 0,030 and 
0,047 mg s·] and sonication at an acoustic power of 57 W (fable E .8; Figure 8. 11) 
F.37 Statistical analysis of the fraction ofundegraded atrazine in a 5 mg L'I atrazine 
solution after ozonation (0,0014 mg S' I) for 45 miD with hydrogen peroxide 
concentrations of 1, 5, 10, 25 and 50 mg L'! (fable E.9; Figure E.5) 
F-28 
F-29 
F.38 Statistical analysis of the atrazine concentration in a 5 mg L'] atrazine solution with F-29 
40 mg L'! hydrogen peroxide, hydrogen peroxide combined with sonication (57 W), 
hydrogen peroxide oombined with ozonation (0,014 mg S'I) and hydrogen peroxide 
combined with sonkation and ozonation (fable E. IO; Figure 8.14) 
A atrazine 
A_ area of the wetted surface of the vessel 
B 
C 
CA 
CB 
C, 
C, 
C, 
C 
Do, 
fA 
fo 
"Gt 
Ml f 
H 
generic chemical compound 
concentration 
atrazine concentration 
concentration of chemical compound B 
concentration of a species; in water 
initial concentration 
liquid heat capacity 
liquid-phase ozone concentration 
ozone diffusivity in water 
fraction of radiation absorbed by atrazine 
ultrasonic resonance frequency 
standard Gibbs free energy of fonnation 
standard heat of formation 
Henry 's Law constant 
Ha dimcnsionlcss Hatta number 
fHO"1ss steady state hydroxyl radical concentration during ozonation 
[HOl hydroxide ion concentration 
hv ultraviolet radiation 
I CtlJ'T(:nt 
I. intensity of adsorbed radiation 
l A atrazinc degradation rate due to direct photolysis 
10 effective intcnsit)' of radiation in water 
k reaction rate constant 
kHQ,B rate constant for the reaction between compound B and HO radicals 
kd rate constant for the direct reaction between ozone and a comp:lUnd 
liquid-phase mass transfer coefficient 
xxxiii 
NOMENCLATURE 
mgL'\ 
mgL'\ 
mgL'\ 
mgL'\ 
mgL'\ 
J kg,l K \ 
mgL'\ 
m2 s'\ 
Hz 
J mOrl 
J morl 
atm m l mor l 
ampere 
Einstein 5'\ 
mol L,I 5'\ 
Einstein L'\ 5' \ 
k\ rate constant for the initiation reaction between solute B and hydroxide ions 
NomencUdure cont 
ks rate constant for the scavenging reaction between solute B and hydroxyl 
radicals 
L 
Le,. 
LD,. 
m 
n 
(o,( 
p 
P 
PA 
Po 
pH 
pKa 
Po, 
q 
Q 
, 
R 
R. 
Ro 
R' 
effoctive path of radiation in a liquid 
lethal chemical concentration that kills 50 % of a group of animals 
lethal chemical dose that kills 50 % of a group of animals 
mass 
sample size 
ozone concentration 
pressure at distance r from bubble centre 
external liquid pressure at infinity 
liquid pressure 
u1trasonic wave amplitude 
negative log of the activity of hydrogen ions 
acid dissociation constant 
gas-phase ozone panial pressure 
bubble gas pressure at maximum radius 
power 
distance from bubble centre 
distance from bubble centre at which maximum liquid pressure occurs 
bubble radius 
maximum bubble radius 
bubble radius at time I = 0 
coefficient of determination 
s standard deviation 
Si scavenger species in water such as inorganic or organic solutes 
SS." variability in data explained by a regression model 
SSra unexplained variability in data, usually attributed to error 
SSID/al tota] variability in data 
ID 
I . 
time 
diffusion time 
reaction time 
cm 
mgm-l 
mgkg-l 
kg 
m 
Pa 
Pa 
W 
mm 
mm 
mm 
mm 
mm 
s 
s 
s 
xniv 
NomenclatMre conL 
T temperature ·c 
maximum temperature in bubble ·c 
To ambient temperature of surrounding liquid ·c 
temperature of the inner vessel wall ·c 
u nuid velocity at distance r from bubble centre 
v volume of irradiated solution m' 
Xi a sample point 
X sample mean 
x., thickness of the inner wall 
Yi a measured dependent variable 
Vi the predicted value of y from a regression model 
y the mean of the dependent variable 
Z volume compression ratio (Rrz/R)3 
Greek letters 
molar extinction coefficient of a species; in water 
e", molar absorptivity at a wavelength of 600 run 
r specific heat ratio of gas in bubble 
7[ constant with an approximate value of 3, 14159 
Cl> A quantum yield of atrazine mol quantum-I 
<D1I,o, quantum yield of hydrogen peroxide I ., mo quantum 
p liquid density 
(J liquid surface tension 
(J,) angular frequency 
Prefixes 
k kilo 10' 
M mega 10' 
m milli 10-3 
" 
micro I O~ 
n nano 10-' 
p pico 10-12 
AEA 
amu 
AR 
ASTM 
ATP 
BOO 
COD 
DM 
DMP 
DOE 
EDTA 
EPA 
ESKOM 
GAC 
HPLC 
LAB 
MSDS 
NADP 
NADPH 
NMR 
NOM 
RAU 
SASOL 
TOC 
UV 
WRC 
Atomic Energy Authority, United Kingdom 
atomic mass unit 
analytical reagent grade 
American Society for the Testing of Materials 
adenosine triphosphate 
biological oxygen demand 
chemical oxygen demand 
Deutscbe mark 
2.9-dimethyJ-l ,lO-phenanthroline 
Department of Energy. United States of America 
ethylenediaminetetraacetic acid 
ABBREVIATIONS 
Environmental Protection Agency, United States of America 
Electricity Supply Commission, South Africa 
granular activated carbon 
high performance liquid chromatography 
laboratory reagent grade 
Materials Safety Data Sheet 
nicotinamide adenine dinucleotide phosphate 
reduced form of NADP 
nuclear magnetic resonance spectroscopy 
natura] organic matter 
Rand Afrikaans University 
Soulh African Coal, Oil & Gas Corporation Lld. 
total organic carbon 
ultraviolet 
Water Research COmmission, South Africa 
aerobic 
adsorption 
advanced oxidation process 
anions 
assimilative capacity 
biological oxygen demand 
biorefractory 
carcinogenic 
cations 
cavitation 
cavitation threshold 
cavitation noise 
chemical oxygen demand 
GLOSSARY 
in the presence of oxygen 
attachment of molecule or ion to a substrate by manipulation of electrical 
charge or pH 
a process that produces sufficient hydroxyl radicals to affect water treatment 
atoms that have gained one or more electrons 
capacity of water to accommodate additional pollutants without a decrease in 
water quality 
oxygen required to meet the metabolic needs of aerobic organisms in water 
rich in organic matter 
resistant to biological degradation 
capable of causing or promoting the development of cancer 
atoms that have donated one or more electrons 
formation., expansion and implosion of bubbles in a liquid 
acoustic pressure amplitude above which cavitation will be induced 
noise caused by the collapsing cavities during sonication 
oxygen required to oxidise all organic matter in a sample that is susceptible 
to oxidation by a strong chemical oxidising agent 
chemiluminescence emission of visible light accompanying a chemical reaction 
coefficient of determination the proportion of total variability explained by a regression model 
confidence interval the range around a calculated statistic that contains the true value of the 
statistic for a panicular level of confidence 
degrees of freedom 
eUlrophicalion 
Fenton ' s reaction 
fluence 
fluorescence 
free radical 
granular activated carbon 
lite number of independent parameters associated with an experiment 
the enrichment of water bodies (such as dams or lakes) with organic or 
mineral nutrients SO that the resultant growth and decay of algae and other 
plants significantly depletes the oJl:ygen content of the water 
reaction between hydrogen peroxide and ferrous salts producing hydrox')'l 
radicals 
the total radiant ultraviolet energy per cross sectional area of target 
emission of visible light from a substance that has absorbed radiation of a 
shorter wavelength from another source 
atom or molecule that contains one or more unpaired electrons 
porous material made from the controll~ combustion of coal, wood or 
coconut shells, used for the adsorption of pollutants during water treatment 
GIoSSDrJI conL 
Hatta number 
heterogeneous 
homogeneous 
hydrodynamic cavitation 
hydrophobicity 
incandescence 
in situ 
in vivo 
magnetostrictive effect 
mean 
minerali.sation 
mutagenic 
oxidant 
oxidation 
pathogenic 
photolysis 
photophosphorylation 
piezoelectric effect 
radiolysis 
recalcitrant 
rectified diffusion 
regiosclective 
salinisation 
sonication 
sonochemistry 
sonolumincscence 
xnviii 
dimensionless group indicating the relative oontribution of reaction kinetics 
to mass transfer effects in gas absorption reactions 
involving more than one phase 
only involving one phase 
cavitation induced in a flowing liquid system by a pressure variation caused 
by a change in velocity 
ratio of hydrophobic groups to hydrophilic groups in a molecule 
emission of visible light due to high temperatures 
in its original place 
in the living body 
a material that undergoes a change in dimension during magnetization 
the average value of a sample 
oomplete degradation of organic oompounds to carbon dioxide and water 
capable of causing mutation 
an oxidising agent 
a chemical reaction in which a oompound loses electrons, a hydrogen atom 
or gains an oxygen atom 
capable of producing disease 
radiation with ultraviolet light 
production of A TP 
a ceramic material that undergoes a dimensional change when an electric 
charge is applied across the opposite faces of the material 
radiation with X-rays or other high~ncrgy sources 
resistant to biodegradation 
mechanism for cavity growth in which more gas diffuses into a cavity when 
it is large in a rarefaction phase than what diffuses out when it is small in a 
compression phase 
reactant orientation so as to produce exclusively_ or nearly so, one of several 
potential isomeric products 
the accumulation of salts in an inland water body (such as a river or dam) 
radiation with sound waves 
the application of ultrasound in chemistry 
emission of visible light during radiation of a liquid with sound waves 
Glossary cont. 
spin trap 
standard deviation 
stripped gas liquor 
synergism 
total organic carbon 
u1trasound 
nxix 
a diamagnetic molecule that reacts with a sbort·lived free radical to produce 
a longer-lived paramagnetic addition product for identification using 
elt:et:ron spin resonance 
a measure of the spread or variation in sample data. the average distance 
between a particular point and the sample mean 
effiuent produced in the quenching of hot gas from coal gasifiers from which 
organic contaminants have been removed and ammonia and acid gas levels 
have been reduced 
a greater combined effect than the summation of the individual effects 
organic carbon other than carbon dioxide (inorganic) 
sound waves with a frequency above the audible range (16 kHz) 
1 
INTRODUCTION 
Water is (he key to development, both present andjurure (Kasrils, 2(00). 
1.1 WATER RESOURCES IN SoUTII AFRICA 
The address delivered by Ronnie Kasrils, the Minister of Water Affairs and Foresuy, on the 17 March 2000 
at the Water Week celebrations emphasised that water is key to development (Kasrils, 2000). Technological 
development will be one of the key drivers ensuring that the 21st centwy becomes the century of the African 
Renaissance as projected by President Thabo Mbeki. Development not only entails industrial and economic 
growth but also the upliftment of the people in the provision of basic services such as potable water, 
sanitation and electricity. All these facets of development, as outlined by the Minister of Water Affairs at 
the Water Week celebrations, are reliant on water. South Africa is one of the leading economies in Africa 
and will thus play a major role in spearheading the upLiftment of the African continent. South Africa, 
however, is not a wateHich country and wiU have to deal efficiently and manage effectively its limited 
available water resources not only to ensure its own development but also its role in facilitating the century 
of the African Renaissance. 
The water situation in a country can be categorised, as presented in Table 1.1, in terms of the total annual 
volume of renewable surface water per head of population (Mullcr, 2000). 
Table 1.1: Classification of the water situation in a country' (Muller, 2000) 
Water availability 
(mJ/personJyear) 
< 500 
500 to 1 000 
1 000 to I 700 
> I 700 
Category of water situation 
absolute scarcity of water 
chronic water scarcity: lack of water begins to hamper 
economic development and human health and welfare 
periodic or regular water stress 
will suffer only occasional or local water problems 
Chapter 1 INTRoDucnoN 1-2 
The total volume of renewable surface water in South Africa is approximately 53 500 million m3 per annum 
(O'Keeffe et al. , 1992). South Africa, based on the population in 1995 and that projected at the present 
growth rate for 2025, was classified according to Table 1.1 for 1995 (1 200 m3 available water per person 
per year) as a country that experiences periodic or regular water shortages and for 2025 (730 m3 available 
water per person per year) as a country that will face chronic water shortages and where economic 
development will be hindered by the lack of water (Muller, 2(00). 
Rivers are the most important water source in South Africa, there are few natural lakes and only 45 % of the 
available groundwater can be utilised. economically (Rabie and Day, 1992). River flow is highly variable 
due to the seasonal changes in rainfall and the uneven distribution across the country. The bighest rainfall 
0CCl.m along the eastern seaboard whereas the greater part of the interior and western portion of the country 
are arid or semi-arid The South African average annual rainfall, 497 mm per year, is below the world 
average of 860 mm (Rabie and Day, 1992). Approximately 65 % of the country receives less than 500 mm 
of rain per year and 21 % below 200 mm per year. The annual evaporation rate across the country varies 
between I 100 and 3 000 mm and is on average greater than the annual rainfall. It is estimated that only 
about 9 % of rainfall reaches the rivers (Muller, 2(00). The variability in flow and high evaporation rate 
results in only approximately 33 000 million m3 (62 %) of the annual runoff of 53 500 million m3 being 
practically exploitable (O'Keeffe et al. , 1992). 
Water demand estimates for 1996 indicated that water use, approximately 20 000 million m3, was almost 
40 % of the total available water and 61 % of the economically exploitable water. In 2030 water use is 
estimated to be approximately 55 % and 80 %, respectively, of the total available and economically 
exploitable water (Muller, 2000). Projections of the expected total water demand per sector in South Africa 
(van der Merwe, 1995) are presented in Table 1.2. 
Table 1.2 : Estimates of South African water demand per sector in 1990, 2000 and 2010 (van der 
Merwe. 1995) 
Se<tor 1990 2000 2010 
million % million % million % 
m
 3/year mJ/year m3/year 
municipal and domestic 2281 12,0 322O 14,4 un 17,3 
industrial 1448 7,6 2043 9,1 2961 11,4 
mining 511 2,7 582 2,6 649 2,5 
power generation 444 2,3 n9 3,5 900 3,5 
irrigation 9695 50,9 10 974 48,9 11 885 45,9 
stock watering 288 1,5 316 1,4 358 I ,' 
nature conservation 182 1,0 187 0,8 191 0,7 
forestry run-otT reduction 1427 7,5 1 570 7,0 1700 6,6 
estuaries and lakes 2767 14,5 2767 12,3 2767 10,7 
Total 19043 100 22438 100 25888 100 
Chapter 1 biTRODUcnON 1-3 
The growing water demand in South Africa has traditionally been addressed by increasing water supply 
with the building of dams and the transfer of water between river catchment basins. Total water storage 
capacity has been increased from 4 400 million m3 in 1956 to the present 29 500 million m3 (Muller, 2000). 
More dams could possibly be built in the eastern pan of the oountry where rainfall is highest, however, good 
dam sites are limited and environmental activists would protest the ecological impact of such a dam. 
lnter-basin transfers have also been used to augment water supplies in areas such as the province of 
Gauteng where water demand is high. The population census of 19% indicated that 18 % of South Africa' s 
population lives in Gauteng (Central Statistical Service, 2000). Gauteng is also home to 60 % of the 
industrial and mining activity in South Africa (van der Merwe, 1995) and 30 % of total electricity 
conswnption (Central Statistical Service, 2000). Water supply to the Gauteng region is supplemented under 
the Tugela-Vaal Scheme where water is pumped from the upper Tugela River across the Drakensberg 
watershed into the Sterkfontein Dam from where it is released on demand into the Vaal River system 
(Labum, 1995; van der Merwe, 1995). An international venture, the Lesotho Highlands Water Project, is 
under construction to help meet the projected water demand in Gauteng till the year 2025. Future water 
supplies could possibly be obtained from the Zambesi River with a pipeline across eastern Botswana or from 
neighbouring Mozambique (Labum, 1995; van der Merwe, 1995). 
Rivers in South Africa are not only under pressure due to the increasing demand for water but also due to 
the reducing water quality caused by inadequate efiluent treatment and illegal waste discharges. Rapid 
urbanisation. poor water management and the lack of funding are leading to organic pollution and 
eutrophication becoming serious problems in many areas (van der Merwe, 1995). Organic pollution in 
rivers also occurs due to chemicals such as pesticides being carried in irrigation runoff from agricultural 
lands. Conventional methods of water treatment are not always successful in the degradation of such 
organic compounds. investigation of new treatment methods, such as advanced oxidation technologies, is 
thus required for the degradation ofbiorefractory organic compounds. 
1.2 WATER LEGISLATION IN SOUTH AFRICA 
Both the efficient use and effective management of available water resources are required to ensure that 
economic development is not hindered due to the lack of water and that the whole nation is provided with 
basic services. Effective management is set in order through the promulgation by government of legislation 
dealing with water issues. Water legislation has changed significantly with the democratically elected 
South African government of 1994. 
Water management has been evident in South Africa since historic times; tribal laws required that people 
washed downstream from where drinking water was collected. Jan van Riebeeck, during early colonisation 
by the Dutch, also decreed that the washing of a soldier' s breeches was not allowed. to take place upstream 
Cbapter 1 b.TRODtJCnOl'l· 1-4 
ofa water offtake point. The Union Health Act of 1919 (Act 36 of 1919) was the first national legislation 
in South Africa that dealt with water issues; liquid effluent had 10 be disposed of on land and was not 
allowed to pollute any surface water (QuibeU et al., 1991). Water reuse bocame inevitable with increasing 
demand and the Water Act of 1956 (Act 54 of 1956) stipulated that all effluent be returned to the water 
body from which it was drawn. The effiuent, however, had to comply with certain standards (the Uniform 
Effluent Standards) so as to prevent a decrease in quality of the receiving waters (Quibell et al., 1997). 
Water legislation, as embodied in the Water Act of 1954, 1II'3$ based on Roman law with the riparian 
principle of water rights being linked 10 land ownership (Kidd, 1997b). The Act distinguished between 
privale and public water where privale water was defined as that which had risen or fallen naturally on any 
piece of land.; the land owner enjoyed exclusive use of the water but could not pollute it. The linking of 
water rights to land ownership in a water-scarce COUDtry such as South Africa resulted in a significant 
imbalance in the aa::ess to water by the majority of the South African population especially during the 
apartheid era (Kidd, 1997b). 
The Constitution of the new government (the lnterim Constitution was enacted with Act 200 of 1993 and 
the Final Constitution with Act J08 of 1996) included a Bill of Human Rights (Kidd, 1 997a). The 
Constitution declared that everyone has the right to access to sufficient water and to an environment that is 
not harmful to their health or well-being (RSA. 1996). The mandate of government is thus to address these 
issues and provide the legal framework within which these rights can be fulfilled 
The right to an adequate supply of safe drinking water is being addressed through the Reconstruction and 
Development Programme wherein government is committed to providing a basic water supply of 25 L per 
person per day to within 200 m from a homestead (van der Merwe. 1995). The fulfilment of this 
constitutional right required the reform of existing water legislation. New water poJiC)' was published in the 
1997 White Paper on National Water Policy, new legislation was enacted with the Water Services Act 
(Act 108 of 1997) and the National Water Act (Act 36 of 1998). The Acts ensure that South Africa ' s water 
resources are protected and used in a sustainable way that is of benefit to everyone. The definition of 
private or public water v.'ilS abolished and all water whether on land.. underground or in surface channels is 
regarded as pan. of Lhe common resource (DW AF. 1997). National government is to act as the custodian of 
the nation ' s water resources. and water management is to be practised per catchment basin. Water 
management is to include an environmental reserve that is recognised as the amount of water required by 
the environment to sustain the integrity of ecosystems (Gels. 1999: Kidd, 1997b). The slogan of the 
Department of Water Affairs and Forestrv Some, For All, Forever summarises the goals of water 
management in South Africa which is to ensure: 
• 
• 
• 
access to a limited resource. 
on an equitable basis, 
in a sustainable manner. now and in the future (DWAF, 1997). 
Cbapter 1 INTRODUCTION 1-5 
The right to an environment that is not barmful to a person's health requires that government prevents 
water pollution so as to protect the quality of water resources in South Africa and ensure the sustainable use 
of available water resources (DW AF, 1997). Water quality is not only defined by the biological, chemical or 
physical attributes but, recognising that water resources are dynamic ecosystems, includes indicators such as 
biotic diversity (GelS, 1999). The quality of South African water resources is declining primarily due to 
salinisation, eutrophication and pollution by trace metals and micro-pollutants (Kidd, 1997b~ 
O'Keeffe et al. , 1992). Water quality may in the future, especially in places in the interior of the country, 
become a more important factor in water management than the increasing demand for water (van der 
Merwe, 1995). 
Industrialisation and economic development inevitably lead to pollution and the deterioration in quality of 
water resources. The allowed environmental impact, however, must be managed so as to ensure sustainable 
development, in that future development is not jeopardised by the present use of available resources, and a 
long-term healthy economy (Walmsley, 1995). Water quality management in South Africa was initiated 
with the Water Act of 1954 that required effluent discharges to comply with the Uniform Effluent 
Standards. The standards consisted of a general standard applicable throughout the country, a special 
standard for certain rivers and, later, a phosphate standard applicable to water sources where eutrophication 
was a problem (Kidd, 1997b; Quibell et al. , 1997). Standards were amended with legislation (and made 
more stringent where necessary) due to the continuing deterioration in water quality. Limitations of the 
Uniform Effluent Standards included that no incentives were provided for pollution reductions to below the 
maximum permitted. level and that the standards were designed for point pollution sources without 
accounting for the cumulative impact of several sources on a waterway (Kidd, 1997b). 
Water quality management changed in the late 1980s with the adoption of the Receiving Water Quality 
Objectives (Kidd, I 997b; O'Keeffe et al. , 1992; Quibell et al., 1997). The policy was based on the 
assimilative capacity of effiuent-receiving water bodies to dilute or degrade pollutants. Site-specific 
standards were formulated to govern effiuent discharge such that qualit), specifications of the receiving 
water, based on the needs of downstream users, were not exceeded. Receiving Water Quality Objectives 
were managed in conjunction with a pollution-prevention approach. Pollution-prevention strategies had to 
be incorporated into the design of plants and companies had to show a commitment to the reduction of 
effluents at source. Administration of the Receiving Water Quality Objectives was limited by the difficult)' 
of establishing the exact needs of downstream users (Kidd, 1997b). 
Water quality management in South Africa, since 1995, has changed with the primary objective being to 
ensure sustainable development. Water resources are to be managed on a catchment basin basis with an 
integrated approach that secures a healthy, stable water resource base to meet ooth current and future water 
needs of South Africa (GelS, 1999). Some impact on the environment is allowed but the impact is to be 
Chapter 1 L"'TRODUCTION 
controlled and managed within a system of waste minimisation technologies, pollution prevention, recycling 
and water reuse sttategies. The use of low-water or zero-eIDuent technologies are to be encouraged with the 
phased-in introduction of effiuent charges and the efficient use of water with voluntary as well as mandatory 
measures for water conservation (GelS, 1999). These measures form. part of the National Strategy 
Framework for Water Conservation and Demand Management that is being developed by the Department of 
Water Affairs and Forestry. This Framework. is to provide a motivation for water conservation in South 
Africa. list the objectives and goals of water conservation and suggest possible water conservation measures 
(DWAF, 2000). The Waste Discharge Charge System is being developed as part of the Water Conservation 
Framework in which a user will pay to be able to discharge effiueDt to a river (DW AF, 2000). 
Water quality management policies are being implemented by government so as to ensure that technological 
development can take place today but not at the expense of water needs of future generations. 
1.3 ADVANCED OXIDATION TECHNOLOGIES 
Government may introduce legislation to effectively manage water qualily and supply, however, daily water 
usage by people must also change to reflect the true value of water. A culture of water conservation focuses 
on minimising wastage and includes water recycling. Pollutants not readily degraded using conventional 
water treatment methods, such as some pesticides and organic compounds, are present in water sources and 
industrial emuents. Research into new or more efficient water treatment technologies is thus ,'ita! to 
combat deteriorating water quality due to pollution. TrihaJomethane by-product formation during 
chlorination has also initiated research into other water treatment technologies such as advanced oxidation 
processes (Kuo and Moo, 1997; Peyton, 1990). 
The focus on waste minimisation and water conservation will result in the production of concentrated or 
toxic residues. Techniques need to be developed to responsibly degrade these streams. Advanced oxidation 
has the potential to degrade (fully or partially) these streams. 
Advanced oxidation processes are defined as processes that generate hydroxyl radicals in sufficient quantit), 
to affect water treatment (Glaze et aI., 1987; Glaze et al. , 1992). These processes include oxidising agents 
such as ozone, ultraviolet radiation,. hydrogen peroxide. Fenton's reagent, ultrasound or combinations 
thereof (Huang et al ., 1993; Sen Gupta et al. , 1995; Vue, 1991). Hydro)':yl radicals are a powerful oxidant 
with an oxidation potential, as shown in Table 1.3, second only to molecular fluorine . Typically, a stronger 
oxidant also exhibits a faster oxidation reaction (Clarke and Knowles, 1982; Huang et al., 1993). 
Chapter 1 br.'TRODUcnON 1-7 
Table 1.3: Standard oxidatioo/reduction potential! at 25°C and 101.3 kPa of selected oxidising 
agenu (Buang et al. 1993; Vanysek, 1998) 
Oxidising agent 
fluorine 
hydroxyl radical 
ozone 
hydrogen peroxide 
chlorine 
Redox reaction 
F2 + 2e- .... 2F-
 HO-+H+ +e- ... H20 
0 3 + 2H+ + 2e- ... 0 2 + H20 
H 20 2 + 2H+ + 2e- ... 2H10 
Ch + 2e- ... 2Cl-
 E'(V) 
2,87 
2,33 
2,08 
1,78 
1,36 
Hydroxyl radicals are not only a stronger oxidant than chlorine, as shown in Table 1.3, but are also less 
selective than molecular ozone (Glaze et al., 1987; Huang et al. , 1993; von Sonntag, 1996). Hydroxyl 
radicals react with most organic and many inorganic solutes with bigh rate constants that approach 
diffusion-controUed limits (Hoigne, 1997; von Sonntag, 1996). Radical scavengers, however, such as 
bicarbonates, carbonates and natural organic matter decrease the effectiveness of hydroxyl radicals (Masten 
and Davies, 1994). Typical rate constants for the second-order reaction with hydroxyl radicals: are listed in 
Table 1.4 (Glaze et al. , 1992). 
Table 1.4: Rate constants for tbe reaction between bydroxyl radicals and Rlected compounds 
(Glaze et al., 1992) 
benzene 
toluene 
I-butanol 
vinyl chloride 
trichlorethylene 
tetrachloroethylene 
chlorobenzene 
nitrobenzene 
, 
3,0 x 10 
, 
4,0 x 10 
2,3 X lOll 
4,5 X lOll 
3,9 X lOll 
Advanced oxidation processes have the potential for complete oxidation (mineralisation) of organic 
pollutants to carbon dioxide. water and salts at ambient temperature and pressure (Glaze et aJ.. 1992; 
Vue, 1992). The processes are also suited for use in combination 'with other technologies as a pretreatment 
or polishing step. Partial oxidation of pollutants where reaction conditions or times may Dot allow for 
complete oxidation has the potential to produce less refractory compounds that can be degraded biologically 
(Glaze et al. , 1992; Vue, 1992). On-site methods for toxic waste degradation eliminate waste transpOrtation 
Chapter 1 INTRoDucnoN 1-8 
and handling and thus comply with waste minimisation practises government is requiring industry to 
implement (Bull and Zeff, 199 1). 
Advanced oxidation processes have been successfully used to degrade compounds such as methyl tert-butyl 
ether from leaking gasoline tanks, chlorinated organics, aromatics and phenolics (Bull and Zeff, 1991 ; 
Glaze, 1987; Kim et al., 1997; Peyton, 1990; Sen Gupta et al., 1995; von Sonntag, 1996). 
1.4 PROJECT BACKGROUND 
Advanced oxidation research at the University of Natal, Durban, under Prof. ehris Buckley began in 199 1 
with a project funded by the Water Research Commission entitled The evaluation of different methods to 
produce free radicals for the ondation of organic molecules in industrial ejJluents and polable waler 'H-ith 
speCial reference to OtV.Q~ (Winship, 1999). The project included an evaluation of the advanced 
oxidation system. CAV"()x..e, that employed DV radiation, hydrogen peroxide and hydrodynamic cavitation 
for the degradation of low-level concentrations of organic pollutants in water (Shah et al., 1999f). Hydroxyl 
radical formation during hydrodynamic cavitation was compared with that during ultrasonic cavitation. 
Ultrasound, an emerging advanced mcidation process, was identified as having potential for the treatment of 
industria] effluents (Shah et al .• 1999f; Winship, 1999). 
UJtrasonic research was initiated in 1992 as the MSc(Eng) project of Schwikkard (Schwikkard, 1995). The 
project, partially funded by SASOL, was entitled An investigarion of the sonochemical degradation of 
hydantoin compounds. Hydantoin compounds are found in the stripped gas liquor at a petrochemical plant 
and increase in concentration as the liquor is recycled for cooling purposes. Degradation of hydantoin 
compounds does not occur during water treatment with the consequence of potential scaling and fouling of 
heat exchangers. The project investigated l-methylhydantoin as a model bioreftactory compound duri ng 
sonication in an ultrasonic bath. The l -methylhydantoin was shown to be degraded during sonication. 
kinetics of degradation were measured and degradation products were identified using various 
chromatrographic t€X:hniques. It was recommended that an ultrasonic cell using a more powerful ultrasonic 
source be designed to allow for the measurement of kinetic data required in the scale-up of ultrasonic 
processes and that ultrasound be evaluated for the treatment of industrial erouen! and the preparation of 
potable water (Schwikkard. 1995). 
Advanced oxidation processes such as ozone, DV radiation and ultrasound are electricity-driven and thus 
attracted interest from ESKOM, the South African electricity supplier. Collaboration with the University of 
Nalal was initiated to investigate and promote the use of ultrasound in potable water and industria] emuent 
treatment (Faul, 1996a; FauJ , 1996b). The project included funding for one of the researchers to attend the 
Fifth Meeting of the European Society of Sonochemistry in Cambridge, July 1996. and various technical 
visi ts to sonochemical research laboratories in the United Kingdom. The visits included that to AEA 
Chapter 1 b.'TRODUCI'ION 1-9 
Technology, Harwell; CoventJy University; the Department of Chemistty at the University of Bath and the 
Department of Physical and Theoretical Chemistry at Oxford University. Research at CoventJy University 
showed that the required dosage of chlorine during potable water disinfection was significantly decreased 
when chlorine was used in conjunction with ultrasoun~ similar results could potentially be expected for the 
combination of ultrasound with ozone (Faul, 19%a). 
A collaborative project between ESKOM and Umgeni Water was undertaken to assess the international 
status and commercial markets of advanced oxidation processes (Gericke, 1999). It was concluded that 
despite the sound scientific basis of the tecbnology certain market forces probibited. full commercialisation 
and that certain fundamental aspects required further research (Gericke, 1999). 
Umgeni Water have also investigated, in a project funded by the Water Research Commission, the potential 
of combining ozone and granular activated carbon as a potential method to treat water from impoundments 
that are tending to become eutrophic (Pryor and Freese, 2000). The project is entitled The treatment of 
eutrophic water using pre- and intermediate ozonation. peroxone and PICA carbon. The nutrient load in 
runoff from informal settlements along river banks and pollutants such as pesticides and herbicides in 
agricultural runoff have contributed to the deterioration in water quality in impoundments in South Africa 
(Pryor and Freese, 2000). Ozone was shown to reduce the concentration of atrazine, a herbicide known to 
persist in ground and surface waters, by over 70 %. Atrazine degradation was enhanced when ozonation 
was followed by absorption on granular activated carbon. Nevertheless. sporadic breakthroughs of atrazine 
above the European Union recommended limit for herbicides (lOO ng L'I) occurred over the 450 d trial 
(Pryor and Freese. 2000). Ozone has been shown to be effective in significantly reducing the concentration 
of herbicides such as atrazine in water, however, to ensure that required limits are consistently achieved. the 
application of ozone in combination with other technologies should be investigated. 
Ultrasound combined with ozone was thus investigated to follow on from the study by Umgeni Water into 
atrazine degradation using different technologies. Atrazine, however. was also a good model compound to 
use in the evaluation of ultrasound in water treatment to degrade organic compounds since atrazine 
chemistry is well studied and it is difficult to degrade. Atrazine is used extensively in South Africa and is 
highly persistent in the environment. Atrazine is also relatively safe to handle and is readily available. 
1-5 PROJECT OBJECTIVES 
The present investigation was initiated to follow-up the recommendations made in the projects described in 
Section 1.4. The objectives of the investigation are to: 
• Design and construct an ultrasonic cell to be used in the evaluation of ultrasound to degrade organic 
pollutants in water treatment 
Chapter 1 INTRoDUCTION 1-10 
• Use such equipment to investigate the process mechanism (the sequence of sub--processes whose 
overall result produces the observed effect) occurring during sonication. 
• Investigate the equipment operation and implications for scale-up. 
• Compare the performance of ultrasound, an emerging technology in water treatment, with that of 
ozone, an already established advanced oxidation process. 
• Use atrazine as a model organic compound and investigate degradation rates during sonication and 
ozonation. 
• Identify degradation products obtained during the sonication and ozonation of atrazine. 
To achieve the objectives, a literature review on ultrasound and ozone was undertaken to gather background 
information so as to assist in designing the ultrasonic cell and to identify experimental techniques for 
performing ultrasonic and ozonation research. Literature on the chemistry of atrazine was reviewed and 
potential degradation mechanisms and products identified.. The ultrasonic cell was designed and 
constructed and a commercially available ozone generator was used for ozone production. Equipment 
characterisation allowed for the necessary experimental techniques to be mastered, to detennine estimates of 
experimentaJ error and to assure that analytical procedures could be followed reproducibly. The 
sub-processes occurring during sonication were investigated by varying acoustic power and ozone input and 
measuring dissolved o:\"ygen, hydrogen peroxide and ozone concentrations. Atrazine degradation kinetics 
during sonication and ozonation were measured and degradation product samples collected for 
identification using mass spectrometry. 
The purJX>Se and function of an ultrasonic cell (or ultrasonic laboratory reactor) within the context of the 
development of large-scale ultrasonic equipment for application in water treatment is demonstrated in the 
reactor design flow diagram shown in Figure 1. 1. 
discovery 
,/' reocto< laboratory ~ 
reactor 
"--.. rate, selectivity 
equations; 
kinetics model 
~ 
reactor 
V model pilot plant 
• "--.. large-sca1e 
reactor 
chemical kineticS; 
catalysis 
f.- G'PPli"'~ ~-.. -.. mathematics 
1'--
 ---G-=-:::> 
I----Essdesv 
Figure 1.1: Reactor design flow diagram (Smith, 1970) 
Chapter 1 (/'I'TRODUCTION \-11 
The scientific disciplines involved between the discovery of a reaction (or demonstration of a process) and 
the large-scale reactor for industrial application are shown on the right in Figure 1.1 . A laboratory reactor 
is built to investigate process parameters (such as reaction conditions, catalyst lifetime) and reaction kinetics 
so as to obtain reaction rate and conversion/selectivity equations (Smith, 1970). This information is 
required for a mathematical model to be developed of the large·scale reactor. The model is evaluated with 
the operation of pilot plant in which questions concerning construction materials. corrosion, operating 
procedures, instrumentation and control are also investigated.. The reactor model can then be used in the 
design and construction of a large·scale reactor for industrial application (Smith. 1970). 
The investigation of the sonochemical degradation of hydantoin compounds in an ultrasonic bath indicated 
the potential of ultrasound to be used in water treatment specifically for the degradation of bioreftactory 
organic pollutants (Schwikkard, 1995). The next step from such a demonstration ofa process, as illustrated 
in Figure 1.1, is the design of an ultrasonic laboratory reactor (an ultrasonic cell) to investigate reaction 
kinetics. process operation and implications for sca1e·up, thus obtaining rate infonnation for the 
development of a model of a large--scale ultrasonic reactor. 
1.6 THESIS OUTLINE 
A flow diagram of the thesis outline is presented in Figure 1.2. 
CHAPTERl I Introduction 
• • • CHAPTER 2 CHAPTER 3 CHAPTER 4 
Ultrasound Ozone and Hydrogen Atrazine Chemistry 
(Literature) Peroxide Chemistry (Literature) 
~ (Literature) • CHAPTERS 
EX~lJeIlt.al Ot:sign 
• • CHAPTER 6 CHAPTER 7 
Ultrasound Process Ozone Process 
Investigation Investigation 
t t 
CHAPTERS 
Atrazine Experiments 
t 
CHAPTER 9 
Concluding Remarks 
Figure 1.2: Flow diagram of thesis outline 
Chapter 1 b-"TRODUCTION 1-12 
The water situation in South Africa specifically the increasing demand, the limited supply. the deteriorating 
water quality and legislation enacted to manage water usage is described in Chapler 1. Advanced oxidation 
processes are introduced and the development of this project from earlier research projects is also described. 
Ultrasound literature is reviewed in Chapter 2. The mechanisms of cavitation and the formation of free 
radicals during sonication are described.. Both chemical and physical effects of ultrasound are discussed and 
examples are listed of ultrasonic applications in the fields of chemistry. industry and medicine. Details are 
also provided. of commercially available sonochemical equipment. Ozone and hydrogen peroxide literature 
are reviewed in Chapter 3. This includes the generation of ozone, the kinetics of ozone decomposition and 
the application of ozone in water treatment. Examples are also given of the application of hydrogen 
peroxide in water treatment. Alrazine literature is reviewed in Chapter 4. The chapter summarises atrazine 
chemistry and the behaviour of atrazine in both soil and water systems in the environment. Potential 
technologies to degrade atrazine are also described. 
Equipment and analytical pnx::edures used in the investigation are presented in Chapter 5. This includes a 
description of the design of the ultrasonic cell and the equipment used for ozone generation. Standard 
procedures were used to measure ozone concentration whereas an HPLC method was developed to measure 
atrazine concentration. Methods to measure hydrogen peroxide and dissolved oxygen concentrations are 
also described. 
Characterisation of the ultrasonic cell is presented in Chapter 6. The process mechanisms occurring during 
sonication were investigated by measuring parameters such as dissolved oxygen concentration, fonnation 
and t;legradation of hydrogen peroxide under saturation with different gases. Implications for scale-up of an 
ultrasonic process are also considered. The investigation of the fundamentals of ozone chemistry and 
equipment characterisation, in tenus of ozone formation and degradation and hydrogen peroxide formation, 
is detailed in Chapter 7. Atrazine experiments are discussed in Chapter 8. This includes the degradation of 
atrazine during sonication and ozonation and the identification of degradation products. 
Concluding remarks and recommendations are presented in Chapter 10. Sources referenced during the 
investigation and Appendices detailing data and calculation procedures are also presented. 
2 
ULTRASOUND 
Ultrasound is defined as sound which is of a frequency above the limit (16 kHz) to which the human ear can 
respond (Mason. 1990). The effects of ultrasound in a medium are due to the chemical and physical effects 
caused by ultrasonic cavitation. The literature presented in this chapter describes the enhancement of 
chemical reactions by ultrasound and the potential for scale-up of industrial applications. 
The history of ultrasound is presented in Section 2.1. The formation and classification of ultrasonic 
cavitation, as well as the phenomena which arise from cavitation, are discussed in Section 2.2. The 
chemical and physical effects of cavitation, and the mechanisms by which ultrasound influences chemical 
reactions, are discussed in Section 2.3 . Applications of ultrasound in the medical. cbemical and industrial 
fields are presented in Section 2.4. Laboratory ultrasonic equipment and ultrasonic reactors for large-scale 
operation are described in Section 2.5. 
2.1 HISTORY OF ULTRASOUND 
The production of ultrasound was made possible with the discovery of the magnetostrictive and piezoelectric 
effects. James Prescon Joule reported in 1847 that the application of a magnetic field to an iron bar caused 
the bar to increase in length (Mason, 1976). The magnetostrictive effect is defined as the production of a 
change in dimension of a magnetic material during magnetisation (Bremner, 1986). Sound waves are thus 
produced in the surrounding medium when an alternating field, by means of a solenoid, is applied across a 
ferromagnetic material. Pierre and Jacques Curie discovered the piezoelectric effect in 1880; a pressure 
applied across certain crystals (such as quartz) was found to produce an electric charge on the surface of the 
crystal (Mason, 1976). The converse of the effect is that a dimensional change is induced. in such a crystal 
when an electric charge is applied across the opposite faces of the crystal. Sound waves are thus produced. 
in the surrounding medium if an alternating voltage is applied across a piezoelectric crystal (Mason, 1976: 
Mason, 1990). 
Paul Langevin, in 1917, performed the earliest investigation of ultrasound (Mason, 1976). Ultrasound was 
investigated. as a method for submarine detection and the first underwater sound transducer (used mainly for 
depth determination) was produoed. The discovery by Langevin that sound waves killed small fish began 
research in the 1920 ' s into the biological effects of ultrasound (Alliger, 1975). At that time pioneering 
Chapter 2 ULTRASOUND 2-2 
work on the chemical effects of u1trasound was performed by Richards and Loontis. Chemical 
transformations accelerated by ultrasound included the explosion of nitrogen tri-iodide, the dispersion of 
mercury, the degassing of liquids, the loy,ering of the boiling temperature of liquids, the hydrolysis of 
dimethyl suJfate and the iodine clock reaction (Rjchards and Loomis, 1921). The emission of light during 
sonication (sonoluminescence) was first reported by Frenzel and Schu1tes in 1934 (Harvey, 1939). 
An investigation of high speed propeUer erosion by Lord Rayleigh (for the British Navy) led to the 
publishing, in 1917, of a paper describing a mathematical model of the collapse of a pre-existing., spherical 
cavitation bubble under a steady pressure (Alliger, 1975; Suslick., 1989). These conditions are different to 
what occurs in an u1trasonic field. however, the concepts presented in the paper formed the basis for the 
development ofu1trasonic cavitation theory (Neppiras, 1984). In 1950, Noltingk and Neppiras published a 
systematic treatment of the growth and collapse of a gas-filled cavitation bubble in a liquid medium 
subjected to a sinusoidal u1trasonic pressure wave (Nollingk and Neppiras, 1950). 
2.2 CAVITATION 
Cavitation is defined. as the formation. expansion and collapse of cavities in a liquid (Sirotyuk, 1963). The 
four types of cavitation are listed below (Shah et al ., 1999a; Young, 1989). 
• 
• 
• 
• 
Hydrodynamic cavitation is produced in a flowing liquid due to the pressure variations caused by 
the system geometry. 
Ultrasonic cavitation is produced in a liquid because of the pressure variations caused by sound 
waves moving through the liquid. 
Optic cavitation occurs when photons of high intensity light (lasers) are focused on a liquid. 
Particle cavitation occurs when elementary panicles (such as protons) are used to rupture a 
superheated liquid in a bubble chamber. 
The formation of u1trasonic cavitation is reviewed in Section 2.2. I: the classification of ultrasonic cavitation 
as either stable or transient in Section 2.2.2: the theoretical equations of cavitation bobble dynamics in 
Section 2.2 .3 and phenomena which arise from u1trasonic cavitation (sonoluminescence and cavitation 
noise) in Section 2.2.4. 
2.2.1 Formation of ultrasonic cavitation 
Sound is transmitted through a liquid by inducing vibrationa1 motion (phases of compression and 
rarefaction) of the liquid molecules (Mason and Cordemans. 1996). Local pressure variations occur as 
liquid molecules are stretched apart during a rarefaction phase of a sound wave and pressed together during 
a compression phase. Cavities arc fonned in a liquid when the negative pressure created by the rarefaction 
phase exceeds the tensile strength of the liquid. the molecules of the liquid are stretched beyond the critical 
Chapter 2 ULTRASOUND 2-3 
molecular distance required to hold the liquid intact (Pandit and Moholkar, 1996; Suslick, 1989). The 
breakdown in a liquid to form cavities should happen simultaneously throughout the liquid but generally it 
occurs only at weak spots or nucleation sites (Sirotyuk, 1963). Microdust particles or dissolved gas 
(gas·filled crevices in suspended particulate matter or microbubbles from prior cavitation events) act as 
nucleation sites in the liquid (Apfel, 1984; Roi, 1957). Cavity formation occurs at these weak spots or 
nucleation sites since the tensile strength of pure liquids is too large for cavitation to be fonned solely by the 
local negative pressure induced by a rarefaction phase of a sound wave (Apfel, 1972; Sirotyuk, 1966; 
Suslick, 1990). 
The cavitation threshold is defined as the sound wave amplitude, depending on frequency, that induces 
sufficient vibration of liquid molecules to cause cavity formation in the liquid (Crawford. 1963; 
Webster, 1963). The following parameters affect the cavitation threshold (Mason. 1990; Pandit and 
Moholkar, 1996; Pestman et al., 1994; Webster, 1963). 
• The presence of dissolved gas in a liquid lowers the cavitation threshold as more nucleation sites 
for cavity formation exist. 
• Physical properties of a liquid, such as surface tension. vapour pressure and viscosity, affect the 
cavitation threshold A reduction in surface tension (such as with the addition of a surfactant) 
decreases the cavitation threshold Liquids with a low vapour pressure do not readily cavitate. 
Liquids with higher viscosities have greater natural cohesive forces and hence a higher cavitation 
threshold. 
• An increase in temperature lowers the cavitation threshold as many liquid properties, such as 
density and viscosity, are temperature dependent. 
• The cavitation threshold is raised as the external applied pressure is increased. A higher external 
pressure requires a greater negative pressure in the rarefaction phase of a sound wave for cavity 
formation to occur. 
• The cavitation threshold is raised with an increase in frequency. The rarefaction and compression 
phases of a sound wave are shortened and insufficient time may be available for a cavity to either 
grow in a rarefaction phase or collapse in a compression phase. 
• Intensity is related to the amplitude of a sound wave, an increase in intensity increases cavitation. 
However, an optimum intensity exists heyond which a further increase will have no effect since the 
fonnation of larger, more stable cavities dampen the passage of sound energy through the liquid 
2.2.2 Classification of cavitation 
Cavitation is classified as either stable or transient (Neppiras, 1984). The experimental distinction between 
stable and transient cavitation is not well defined (Henglein, 1987; Vaughan and Leeman, 1989). However. 
the lifespan of a stable cavity is longer than that of a transient cavity and stable cavities exist over many 
acoustic cycles whereas transient cavities exist for only one or two acoustic cycles. Stable cavitation is 
Cbapter 2 ULTRASOmm 2-4 
produced by ultrasonic intensities in the range of 1 to 3 W cm-2 whereas transient cavitation is produced. by 
intensities greater than 10 W cm·2 (Mason, 1990). 
Stable cavities are bubbles that exist for many acoustic cycles and oscillate, often non linearly, about an 
equilibrium radius that is dependent on the applied ultrasonic field (Henglein, 1987; Riesz et al .. 1990a). 
The lifespan of stable cavities is still short since a cavity which exists for I 000 acoustic cycles in a 20 kHz 
ultrasonic field has a lifespan of 0,05 s (Mason, 1990). Stable cavities grow through a process called 
rectified diffusion in that the amount of gas diffusing in or out a cavity depends on the surface area of the 
cavity; more gas diffuses into a cavity when it is large, in a rarefaction phase, than diffuses out of a cavity 
when it is small, in a compression phase (Suslick. et al., 1990). The cavities grow to a size at which they 
become unstable and implode. The implosion is cushioned by the gas and liquid vapour contained in the 
cavity (lIenglein. 1987; Neppirns, 1984). 
Transient cavities exist for only one or two acoustic cycles and implode violently during a single 
compression phase once the cavities have expanded to a size two to three times the equilibrium radius 
(Riesz et al., I 990a), The lifespan of the cavities is too short for gas dissolved in the liquid to diffuse into 
the cavities. However, the cavities may contain some vapour from the liquid The collapse of the cavities is 
violent, as no gas is present in the cavities to cushion the implosion (Henglein, 1987). Upon implosion, the 
cavities often disintegrate into numerous smaller cavities providing nuclei for fwther cavitation 
(Henglein, 1987). Powerful shock waves are produced and considerable energy is released. into the liquid as 
the cavities implode (Boudjouk. 1986). From theoretical calculations, temperatures and pressures in a gas 
cavity during the final stages of collapse have been estimated to be in the range of 2 000 to 10 000 K and 
100 to I 000 MPa (Suslick et al .• 1986). The high temperatures and pressures in the cavities cause the 
dissociation of water vapour into bydroxyl and. hyclrogen radicals (Bremner, 1986; Makino et al., 1982; 
Riesz et al .. 1990a; Tod.d, 1970; Webster, 1%3). 
Sonochemica1 effects are caused by both types of caviuation despite the earlier belief that only transient 
cavitation enhanced chemical reactions (Ratoarinoro et al ., 1992). Although the temperatures and pressures 
developed in stable cavities are lower then those in collapsing transient cavities (the implosion is cushioned 
by the gas content in the stable cavity), the lifespan of stable cavities is longer and there is thus a greater 
potential for influencing chemical reactivity (Mason, 1990). 
Vaughan and Leeman have proposed a different system for the classification of cavitation (Vaughan and 
Leeman, 1989). The system, unlike stable/transient classification, provides for a hierarchy of thresholds of 
observed cavitation phenomena, such as ca,';tation noise (or subharrnonic emission), sonoluminescence and 
sonocbemical reaction enhancement. The stable/transient mode of classification developed from the 
theoretical interpretation of the implosion of a gas-less cavity, however, ultrasonic cavitation occurs mostly 
in gas-saturated liquids; the classification system proposed. by Vaughan and Leeman is not dependent on a 
Cbapter 2 ULTRASOUND 2-5 
specific bubble dynamic model (Leeman and Vaughan, 1992). Cavitation is defined as the non-linear 
pulsation of bubbles in a sound field and is classified as subsonic, gas-phase or liquid-phase cavitation. The 
classification is performed by comparing the maximum velocity of the bulX>le wall with the speed of sound 
in the gas in the bubble and with the speed. of sound in the liquid surrounding the bubble. The dynamic 
bebaviour of the bubble wall is regarded as one of the most important parameters that characterises the 
interaction of the bubble with the applied sound field (Vaugban and Leeman, 1989). 
Subsonic cavitation occurs when the maximum velocity of the bubble wall is lower than both the speed of 
sound in the gas in the bubble and the speed of sound in the surrounding liquid (Leeman and 
Vaughan, 1992). Subsonic cavitation is characterised by non-linear motion of the bubble wall and is the 
threshold for cavitation noise. Gu-pbue cavitation occurs when the applied acoustic power is increased 
such that the maximum velocity of the bJbble wall becomes greater than the speed of sound in the gas in the 
bubble (but still less than the speed of sound. in the liquid). The gas inside the bubble is shock-excited and 
significant heating occurs. Gas-phase cavitation marks the threshold for sonochemical reactions between 
activated species within the gas bubble as well as for sonoluminescenc:e since light emission is associated 
with the movement of shock waves through a gas. Liquid-pbue cavitation occurs when the applied 
acoustic power is funher increased such that the maximum velocity of the bubble wall is greater than both 
the speed of sound in the gas in the bubble and the speed. of sound in the SWTOUDding liquid.. Shock waves 
are generated in the liquid surrounding the bubble; high temperatures and. pressures accompanying the 
shock waves lead to sonochemical reactions taking place in the liquid.. However. the outward radiating 
nature of the shock waves cause such reactions to take pl.acc: in a relatively thin shell of liquid surrounding 
the bubble (Leeman and Vaughan, 1992). 
2.2.3 Equations of cavitation bubble dynamics 
One of the earliest systematic t:reaUDents of u1trasonic cavities was performed by Noltingk and Neppiras 
(Noltingk and Neppiras, 1950). The uluasonic cavitation of a liquid saturated with gas was mroelled based 
on the analysis performed by Rayleigh of the implosion of a gas-less void.. A differential analyser was used 
to solve a non-linear differential equation of motion for the radial behaviour of a gas-filled bubble in an 
incompressible liquid, acted upon by a sinusoidal varying pressure field. For specific initial conditions. the 
solutions to the equation were plotted as bubble radius versus time graphs. The pressure distribution in the 
liquid surrounding a collapsing, gas-filled bubble was also investigated. The following assumptions were 
made in the model. the liquid was taken to be incompressible, the composition of the gas in the bubble was 
assumed to be constant during the lifespan of the bubble, the applied ultrasonic pressure was assumed to be 
exactly sinusoidal and the diameter of the bubble was assumed to be much less than a wavelength of sound 
(Noltingk and Neppiras. 1950). The hot spot theory for the generation of sonochemical effects developed 
from the analysis by Noltingk and Neppiras. 
Cbapter 2 ULTRASOUND 
The following theoretica1 derivation of the radial behaviour of a gas-filled bubble in an incompressible 
liquid, acted upon by a sinusoidal varying pressure field is summarised from the paper published by 
Noltingk and Neppiras (Noltingk: and Neppiras, 1950). The external liquid pressure P at a distance of 
infinity is equal to the sum. of the hydrostatic and acoustic pressures. The external liquid pressure at time I 
when an ultrasonic wave of amplitude Po and frequency ro/27t, where ro is the angular frequency, is applied 
over a liquid pressure, P A, is 
P = P A - Po sin(wt) [2.1] 
The bubble has an arbitrary radius Ro at time t = 0, and contains gas at the equilibrium pressure of 
(PA + 2aIR o), where a is the surface tension of the liquid The kinetic energy 2%PR{ ~r of the whole 
mass of liquid with density p is equal to the algebraic sum. of the work done by the surface tension, gas 
pressure and liquid pressure at infinity. The energy equation, if the gas changes are isothermal, is 
[2.2] 
The differential equation of motion is found by differentiating with respect to R, giving 
[ . ( k )R! 1 (dR )' cl'R 2R Posm(cot) - P A + PA + R o R 3 = 4a + 3pR t + 2pR2 dt2 [2. 3] 
Noltingk and Neppiras found that the ooundary value of the bubble wall radial velocity (dRldt) at Ro did not 
significantly affect the (c, R) curves produced from numerically solving equation 2.3 (Noltingk and 
Neppiras, 1950). The radial velocity at Ro was therefore assumed to be zero. 
12.4] 
The solution of the differential equation for radial motion of a bubble (equation 2.3) over four cycles of 
pressure is shown in Figure 2. l. The equation was solved numerically for a specific set ofinitial conditions 
(P A, Po and Ro) at frequencies of 5 and 15 MHz. 
Chapter 2 ULTRASOUND 
·~.--------,~,,-----.--.y.-------~u. 
Time (x 10 sec) 
" 
.. 
• !!l 
~ 
• 
" ~ 
" ~ 
oil 
u 
• 
\ 
\ 
\ 
\ 
\ 
\ 
\ 
\ 
\ 
I 
I 
I 
I 
I 
I 
I , 
01" -4 U 
Time (1: 10 1«) 
2-7 
Figure 2.] : Radius versus time curves for a gas-fiUed bubble in water in an ultrasonic field 
with a frequency of 5 MHz (dashed line) or 15 MHz (solid line); PA = 100 kPa, Po = 400 kPa 
and R o "" 0,8 f.Lm (Noltingk and Neppiras, 1950) 
The radius of a bubble in an ultrasonic field increases to a maximum Rm at which the bubble becomes 
unstable and begins to collapse. After the bubble radius has reached a maximum, as shown by the dashed 
line in Figure 2.1. it shrinks with increasing rapidity. The slope of the graph eventually becoming too steep 
to be traced with the differential analyser as used by Noltingk and Neppiras in solving equation 2.3 
(Noltingk and Neppiras, 1950). Noltingk and Neppiras regarded bubble collapse as a prerequisite for 
cavitation and did not classify a bubble oscillating according to the solid line in Figure 2.1 as cavitating. 
The system of classifYing cavitation as either transient or stable, where stable cavitation was defined as the 
non-linear oscillation of a bubble, was developed later (Flynn, 1964, in Vaughan and Leeman, 1989). A 
review published by Neppiras in 1980 used the diagram shown in Figure 2.1 to show the difference in radial 
motion between stable and transient cavitation (Neppiras, 1980). 
Chapter 2 ULTRASOliND 2-8 
Noltingk and Neppiras also investigated the pressure distribution in the liquid surrounding a collapsing 
bubble (Noltingk and Neppiras, 1950). The energy equation for the collapse of a gas-filled bubble, 
assuming adiabatic conditions, is 
3P(dR)' = 1'(Z _ 1) _ q(Z - Z') 
2 dl 1 Y 12.5] 
where r is the specific heat ratio of the gas in the bubble and q the gas pressure in the bubble at the 
maximum radius R = Rm. The symrol Z, the volume compression ratio, is defined as (Rn/R)). The externa] 
liquid pressure P is assumed to be constant because bubble collapse is very fast. Surface tension forces are 
ignored, as they would be insignificant during the collapse as compared with the two opposing pressure 
terms (Noltingk and Neppiras, 1950). 
The pressure distribution in the liquid surrounding a collapsing bubble is obtained by application of the 
general hydrodynamic equation of motion 
12.61 
where u is the fluid velocity at a distance r from the centre of a bubble and p is the pressure at r . The partial 
derivatives, ". and -f, are found as functions of rand R from the energy equation (equation 2.6) and the 
continuity equation, 
12.71 
Noltingk and Neppiras found this to give 
dp = -.iL[ qZ'(3y - 4) + ~ _ (Z _ 4 )P j + 4R' [1'(Z _ 
dr 3r2 1 - Y I - Y 3r~ 
q(Z-Z')j 1) - ] 
- y 12.8] 
which up:m integration gives 
_ P _ _ lL [ qZ'(3y - 4) 
p - 3r 1 - Y + -'1£ _ (Z - 4)P j- K [1'(Z _ 1) _ q(Z - Z') j l -y 3,... l y 12.9] 
The function in equation 2.9 was solved by Noltingk and Neppiras for various values of Z and is plotted in 
Figure 2.2. 
Chapter 2 ULTRASOU/lo'l) 
.. 
r 
,.~ 
, .. 
__________ _ _ )':<l _ 
., 
QO R. lR. J 2. 
Distance from bubble centre 
Figure 2.2: Pressure distribution in tbe liquid surrounding a collapsing, gas-itlled bubble 
wbere Z is the volume compression ratio (R. IR)l (Noltingk and Neppiras, 1950) 
The value of Z (defined as the ratio (RrdR)3) becomes greater during bubble collapses since the radius R is 
decreasing whereas Rm is constant. The four lines drawn in Figure 2.2 show the pressure distribution in 
and around a collapsing gas-filled tNbble. The line for Z = 1 represents the situation just before the bubble 
begins to coUapse. The horizontal part of the lines indicate the constant gas pressure within the gas bubble; 
pressure then increases from the bubble swface into the surrounding liquid. The amplitude of the liquid 
pressure wave increases steeply as the bubble radius decreases and inward velocity increases. 
The maximum liquid pressure occurs at a distance rm from the bubble centre and is obtained by solving for 
dp/dr = 0 in equation 2.8 and then substituting for r = rm in equation 2.9. This pressure occurs at the bubble 
surface al minimum radius and for adiabatic collapse is equal to q(P13q)Y (Naltingk and Neppiras, 1950). 
The maximum liquid pressure was calculated for various ultrasonic frequencies and is plotted in Figure 2.3. 
Chapter 2 ULTRASOUND 
" 
- ,.-.----7""-----------, e 
f UMO 
i 
· ,. 
"-
 :: lDO 
• g 
e 
• e 
•• 
• 
" 
:;: ".~------_c,c--------7." 
Frequency (MHz) 
2-10 
Figure 2.3: Maximum liquid pressures calculated for different ultrasonic frequencies during 
bubble collapse under adiabatic conditions; Po = 400 kPa and & = 3,2 .... m (Noltingk and 
Neppiras, 1950) 
As shown in Figure 2.3 high pressures are produced during the collapse of a gas-filled bubble. These 
pressures, however, exist for only a fraction of the total collapse time and over a small volwne of fluid near 
the bubble surface. Isotherma1 gas changes were assumed for bubble growth and adiabatic for bubble 
collapse (Noltingk and Neppiras, 1950). The assumption of adiabatic conditions for the collapse period 
implies that gas temperature increases to a high value as bubble radius approaches a minimum. The gas 
temperature at radius R for adiabatic compression is given by 
( Rm )"' - " T= To R 12.10) 
where To is the ambient temperature of the surrounding liquid The maximum temperature in the bubble 
occurs when the bubble radius is a minimum (Noltingk and Neppiras, 1950). The maximum temperature 
can therefore be calculated from equation 2.10 when R is equal to the minimum radius. At minimum radius 
equation 2.10 becomes 
( 
p )xr-" 
T = To 3q 12.11) 
The maximum gas temperature is calculated from equation 2. 11 to be 10 000 K when To is 300 K, P is 
100 kPa and q is I kPa (Noltingk and Neppiras, 1950). The temperature of the bubble surface would 
effectively be the same as that of the surrounding liquid (To) and a steep temperature gradient would exist in 
the gas adjacent to the bubble wall. The high temperatures confined to the gas in the bubble was proposed 
to lead. to incandescence, accounting for the luminescence observed during ultrasonic cavitation (Noltingk 
and Neppiras, 1950). 
Cbapter 2 ULTRASOUND 2-11 
Most of the theory of ultrasonic cavitation has developed from the theoretical analysis performed by 
Noltingk and Neppiras. Various authors, to model experimentaJ conditions more realistically. have made 
adaptions to include terms accounting for parameters such as viscosity and compressibility of the liquid 
medium. The review published by Neppims in 1980 summarises the ultrasonic cavitation theory 
(Neppiras. 1980). Whereas the early development of u1trasonic cavitation theory focused on what is now 
defined as transient cavitation, equations have been published for both stable and transient cavitation 
(Neppiras. 1980: Shah et al ., 1999b; Tbompson and Doraiswamy, 1999). 
2.2.4 Cavitation phenomena 
Sonochemical n:action enhancemenL sonoluminescence and cavitation noise are phenomena caused by 
ultrasonic cavitation (Webster, 1963). Sonoluminescence and cavitation noise are reviewed in 
Section 2.2.4.1 and Section 2.2.4.2 respectively; sonochemistr)' is reviewed in Section 2.3. 
2.2.4.1 Sonoluminescence 
Sonoluminesccnce is a weak emission of light observed during ultrasonic cavitation in liquids such as \\'3ter 
containing dissolved gas (Verrall and Sehgal, 1988). One of the earliest observations of sonoluminescence 
was by Frenzel and Schultes in 1934 (Harvey, 1939). Sonolwninescence occurs periodically with the 
frequency of the sound field, one flash per cycle of sound (Sak.sena and Nyborg, 1970). The occurrence of 
sonoluminescence has been reported for both continuous (Sehgal et al. , 1980a) and pulsed ultrasound 
(Henglein and Gutierrez, 1993). The amplitude of an ultrasonic wave must be greater than a certain 
threshold value before sonoluminescence will occur (Vaug.han and Leeman. 1989). Various thermal and 
electrical theories have been proposed to ex "Plain the origin of sonolurninescence. 
• 
• 
The electrical microdischarge theory proposed by Frenkel in 1940 postUlates that lens-shaped 
cavities are formed when a liquid is ruptured by a sound field (Frenkel. 1940. in Webster. 1963). 
Ions of opp:tSite charge gather on the opp:tSing cavil)' walls and produce an electric field within the 
cavity. The electric field becomes stronger as the cavity develops a spherical shape. Emission of 
light occurs during the growth of the cavity. 
The mechanochemical theOJ)' proposed by Weyl and Marboe in 1949 postulates that light emission 
occurs upon photochemical recombination of ions formed by the mechanical fracture of molecules 
at the surface of an expanding cavity. Emission of light is independent of dissolved gas and occurs 
during the growth of a cavity (Weyl and Marboe, 1949, in Verrall and Sehgal. 1988). 
The anion discharge theol)' was proposed by Degrois and Baldo (Degrois and BaJdo. 1974). This 
theol)' postulates that light emission occurs when the internal charge of a gaseous bubble is rapidly 
discharged during bubble compression. Electrons are produced in the bubble when the gas 
molecules at the gas-liquid interface neutralise the anions in the liquid which are adsorbed onto the 
external bubble surface. If the total internal charge acquired by the bubble is considered to be 
Chapter 2 ULTRASOUND 2-14 
(Sehgal et al., 1980b). The actual temperatures in the gas cavities were calculated (not asswning adiabatic 
conditions) to be I 350 K for nitric oxide and 860 K for nitrogen dioxide. The sonoluminescence spectra of 
nitric oxide- and nitrogen dioxide-saturated water were found to be similar to the spectra from fluorescence 
and thennal emission of nitrogen dioxide and from NO + ~ chemiluminescence (Sehgal et ai. , 1980b). It 
was concluded that sonoluminescence arose from a similar mechanism; emission occurring from the 
gas-phase, electronic-excited state of N~ (SehgaI et al., 1980b). An investigation of the sonoluminescence 
spectra from argon-satuJated alkali metal salt solutions (sodium and potassium) also concluded that light 
emission occurred from the de-excitation of excited chemical species in the gas cavities 
(SehgaI et al., 1979). Light emission was found to occur from the de-excitation of the excited alkali metal 
atoms (Sehgal et al., 1979). The conditions in the gas cavities excite molecules into high-energy states and 
light is emitted when the molecules return to their ground states (Suslick, 1989). 
It was originally reponed that sonoluminescence was on1y characteristic of transient cavitation and did not 
occur during stable cavitation (Neppiras and Fill, 1969). The work published by Saksena and Nyborg in 
1970 and Crum and Reynolds in 1985 shov.'ed that sonoluminescence also occurs during stable cavitation 
(Saksena and Nyborg, 1970; Crum and Reynolds, 1985). 
In another investigation by Sehgal and co-workers, sonoluminescence was observed from water saturated 
with helium, neon, argon, krypton, oxygen, nitrogen and air (SehgaJ et al. , 1980a). The solutions were 
sonicated at ultrasonic frequencies of 333 and 459 kHz. The emission spectra were measured by using 
single-photon counting techniques and showed a continuwn that extended from a wavelength of 240 om to 
the near-infrared region (Sehgal et al., 1980a). The spectrum of argon-saturated water sonicated at 333 kHz 
is shown in Figure 2. 4. 
",---~--~-------, 
I 
. , 
~ 
• 
... .. 
XI""') 
Figure 2.4: SonolumineKence spectrum of argon-saturated water sonicated at 333 kHz; 
I, emission from HzO*; n, emission from OH*; and rn, emission from 0"+ 00" reaction 
(Sebgal et al., 1980a) 
Chapter 2 ULTRASOUND 2-15 
The emission spectrum of a solution arises due to the sonolwninesc:ence from 'ooth stable and transient 
cavitation (Sehgal et al., 1980a). The higher temperatures in transient cavities cause electronic excitation. 
The formation of excited water molecules is followed by either radiative relaxation or dissociation into 
hydrogen and hydroxyl radicals. The emission in the near-UV region with a peak maximum in the region 
of 270 nm to 290 nm, shown by curve I in Figure 2.4, is the emission from the transitions from the triplet 
38, state of water (Sehgal et al. , 1980a). Emission below 400 om, shown by cwve II, is due to the 
electronically exited OH(~") formed by the fragmentation of water in the excited state. The lower 
temperatures in stable cavities cause vibrational excitation leading to the dissociation of water in the ground 
state. The radicals formed combine radiately to give a continuum with a broad maximum around 400 nm, 
shown by curve ill in Figure 2.4 (Sehgal et al., 1980a). Selective quenching of hydroxyl radical emission 
by the addition of nitric acid has also provided chemical evidence for the formation of the excited water 
molecule, H~ (Sehgal et al., 198Od). The nitric anion, N(X, preferentially scavenges the hydro,,",),' 
radicals and uncovers the emission peak. at about 280 run, from the excited water molecule. In the absence 
of a hydroxyl radical scavenger, the emission peak: for the excited water molecule is hidden by the emission 
peaks for the excited hydroxyl radicals (Sehgal et al., 198Od). 
Various factors influence the intensity of sonoluminescence. An increase in temperature of the bulk 
solution causes the intensity of sonoluminescence to decrease (Didenko et al. , 1994; Sehgal et al., 198Oc). 
By comparing the spectra from neon-saturated water at various temperatures, Sehgal and co-workers 
concluded that the chemica1 reactions producing sono1uminescence were not dependent upon ambient 
temperature, however, the decrease in sonoluminescence was mostly due to a decrease in the degree of 
cavitation (Sehgal et al. , 198Oc). Didenko and co-workers investigated the sonoluminescence from argon 
saturated water and reasoned that an increase in bulk solution temperature increased the vapour/gas ratio 
inside the gas cavity (Didenko et al., 1994). This was due to the decrease in gas SOlubility and hence the 
water vapour pressure in the gas cavity was raised. The specific heat ratio (y) of water vapour is lower than 
that of the noble gases (argon) and hence a lower temperature would be reached in the cavity during cavity 
collapse; and a lower intensity of sonoluminescence would be produced (Didenko et al. , 1994). 
For a fixed ultrasonic intensity, sonoluminescence decreases with an increase in frequency (Verrall and 
Sehgal, 1988). The spectral distribution of the light is also changed 'With a change in frequency. Sehgal 
and c;:o..workers measured the spectra fOT argon-saturated water at frequencies of 333 and 459 kHz 
(Sehgal et al. , 1980a). It was reported that the 310 nm band ascribed to the OH. emission was more 
pronounced and defined at the lower frequency, and that the continuum band from the radiative 
recombination of free radicals dominated the spectrum at the higher frequency. It was reasoned that the 
differences in the spectra 'were related to the relative change. with frequency, in the distribution of transient 
and stable cavities (Sehgal et al. , 1980a). Chendke and Fogler investigated the intensity of 
Chapter 2 ULTRASOUND 2-16 
sonoluminescence from nitrogen..saturated water for static pressures ranging from 0, 1 to 1,46 MPa 
(Chendke and Fogler, 1983b). The intensity of sonolwninescence was found to increase with static pressure 
up to a maximum at about 0,6 MPa. A further increase in static pressure resulted in a decrease in 
sonoluminescence intensity; at a pressure of 1,4 MPa sonoiuminescence was totally suppressed. The initial 
increase in sonoluminescence was attributed. to an increase in the number of gas cavities. The decrease in 
sonoluminescence at higher static pressures was due to the acoustic pressure being lower than the static 
pressure and thus 100 low to initiate cavitation (Chendke and Fogler, 1983b). The spectral distribution of 
the light did not change over the pressure range investigated (Chendke and Fogler, 1983b). Investigation of 
the sonoluminescence of aqueous carbon tetrachloride solutions by Chendke and Fogler also showed that the 
intensity of sonoluminescence was dependent on static pressure whereas the spectral distribution of light 
was independent of pressure (Chendke and Fogler, 1983a). 
Sonoluminescence has been recorded in non-aqueous liquids. The sonoluminescence intensity of cartxm 
tetrachloride and water mixtures was measured by Chendke and Fogler (Cbendke and Fogler, 1983a). The 
intensity was found to increase with increasing amount of carbon tetrachloride in water; a fully-saturated 
solution luminesced with an intensity 2,8 times that of pure water. It was ooncluded that light emission 
occurred because of chemiluminescence (Chendke and Fogler, 1983a). The sonoluminescence from various 
hydrocarbons saturated with different gases was measured by Flint and Suslick (Flint and Suslick., 1989). 
Similar to sonoluminescence from aqueous solutions, sonoluminescence from non-aqueous solutions was 
found to occur from ex.cited chemical species. Spectral studies showed that light emission from 
argon-saturated alkanes occurred from excited Cl, C2H and CH molecules. For nitrogen-saturation, light 
emission occurred from the excited state of the CN molecule; for oxygen-saturation, light emission occurred 
from the excited states of the COz, CH and OH molecules (Flint and Suslick, 1989). For linear alkanes such 
as dodccane. decane and octane. sonoluminescence was found to decrease with increasing vapour pressure. 
This is due to the reduced cavity temperature reached during cavity collapse (Flint and Suslick, 1989). 
For both aqueous and non-aqueous liquids sonolwninescence arises from the chemical reactions of high 
energy species formed during cavity collapse. It is a form of chemilwninescence and is not due to either 
blackbody radiation or electrical discharge (Suslick, 1990). 
2.2.4.2 Cavitation noise 
U1uasound is accompanied by an intense noise. The collapsing cavities act as secondary sources of sound 
emitting spherical waves (Neppiras, 1980). The spectral oomposition of this cavitation noise oonsists of 
harmonics and subharmonics of the fundamental uluasonic frequency and a oontinuous background or 
continuum of white noise (Roi, 1957). The noise spectra for ultrasonic frequencies varying from 3 kHz to 
3,3 MHz was measured by Esche (Esche, 1952, in Webster, 1963). The noise spectrum for an ultrasonic 
frequency of 15 kHz is shown in Figure 2.5. 
Chapter 2 ULTRASOUND 2-17 
f 7 
6 
005 23 S 
, 
"1 IOB 
, 
10 '" I. 
Frequency (kHz.) 
Figure 2.5: Cavitation noise spectrum for an ultrasonic resonance frequency «0) of 1!5 kHz 
(Esche, 1952, in Webster, 1963) 
The half-order subharmonic line and the higher order harmonic lines of the driving frequency, as well as 
the continuous background of the noise spectrum. are shown in Figure 2.5. The hannonic lines are caused 
by the non-linear radial and surface oscillations of the cavities and the continuous portion of the spectrum 
arises from the violent collapse as well as oscillation of the cavities (Rooney, 1988; Webster, 1963), The 
subhannonic lines are possibly caused by the parametric amplification of a low-level signal already present 
in the liquid (Tucker, 1965, in Room~y, 1988), or possibly by the large amplitude radial oscillations of 
cavities that are unstable to subhannonic response when the resonant frequency of the cavity is half the 
applied frequency (Room~y, 1988). 
The cavitation noise spectra depend on the state of gasification of the liquid, ambient pressure and 
temperarure and ultrasonic intensity and frequency (Neppiras, 1980). Cavitation noise increases with 
increasing ultrasonic intensity and is greater for transient than stable cavitation (Neppiras, 1980). 
2_3 SONOCHEMISTRY 
Sonochemistf)' is the application of ulttasound in chemistry. Sonochemical benefits are caused by the 
chemical and physical effects that arise from ultrasonic cavitation; chemical effects are reviewed in 
Section 2.3.1 and physical effects in Section 2.3.2. 
2.3.1 Chemical effects 
The chemical effects of ultrasonic cavitation are caused by the fonnation of hydroxyl and hydrogen radicals 
in the collapsing cavities (Makino et al., 1982). The radicals react with each other and with solutes present 
Chapter 2 ULTRASOUND 2-18 
in the liquid medium. The site of these sonochemical reactions (in relation to the collapsing cavity) is 
reviewed in Section 2.3. 1.1 and the effect of dissolved gas on the reactions is reviewed in Section 2.3. 1.2. 
2.3.1.1 Site ofsonocbemical reactions 
Sonochemical reactions in a cavitating liquid occur in three regions, the gas inside a cavity, the interface 
between a gas cavity and a liquid and the bulk liquid (El'piner, 1959; Suslick and HammertoIl, 1986). 
Margulis disagreed with such a system of classification as it was reported that sonochemical reactions were 
the result of different processes occurring in both the gas and liquid phases, the role of each depending on 
various factors that changed throughout the process, and thus could not be classified as reactions occurring 
in separate regions (Margulis, 1969). A diagram of the three regions in a cavitating liquid in which 
sonochemical reactions take place and the processes that occur in each region are shown in Figure 2.6. 
into cavity 
GAS CAVITY 
5 200 ~ 100 MPs 
thermal dissociation processes 
HP -+ Ho + HO 
radical reactions 
HO- + HO· --+ HP! 
Ho+H°--+H, 
Ho + He) --+ HP 
diffusion of volatile 
solutes into bubble 
GASILIQUID 
INTERFACE 
steep lemper8tw"e and 
pressure gradients 
about 200 run width; 
2 J..IS lifespan; 
I 900 K average 
tempernture 
outward diffusion of 
H· HO- HI HP! 
=!'::~om,,~ ) ) 
hydrophobic SOl/ / 
BULK 
SOLUTION 
ambient conditions of 
temperature and pressure 
species 
H· HO- HI Ht): 
react with 
hydrophilic solutes 
reaction kinetics similar to 
radiation chemistry 
shock waves 
fragment polymers 
Figure 2.6: Schematic diagram of tbe three regions in a cavitating liquid in which chemical 
reactions take place 
The gas inside the collapsing gas cavities is the region where according to the hot spot theory, high 
temperatures and pressures occur (Noltingk and Neppiras. 1950). Noltingk and Neppiras calculated the 
maximum temperature in a collapsing cavity to be approximately IQ 000 K (Noltingk. and Neppiras, 1950). 
The formula for temperature calculation uncler adiabatic conditions (equation 2.11) was approximated by 
Neppiras to be 
T~ T"P(y - I ) = q 12.12) 
where To is the ambient temperature of the liquid, q is the gas pressure in the cavity at its maximum radius. 
y is the specific heat ratio of the gas in the cavity and P is the external pressure (sum of the hydrostatic and 
Chapter 2 ULTRASOUND 2-19 
acoustic pressures) of the liquid (Neppiras, 1980). This equation indicates the direct relationship between 
the maximum temperature reached in the collapsing cavities and the specific heat ratio of the gas; the 
higher the specific heat ratio, the higher the temperature (Riesz et al., 1990a). The maximum temperatures 
calculated for specific heat ratios of 1,67 (for a monatomic gas) and 1,30 (for the triatomic gas N-zO) were 
3 015 and I 350 K, respectively, when To was 300 K and the ratio Plq was 15 (Riesz et al. , 1990a). 
The work published by Noltingk and Neppiras in 1950 has been followed by the publishing of different 
equations to calculate the maximum temperature in a collapsing cavity. These theoretical studies begin 
with the adiabatic collapse of a gas-filled cavity and then apply various corrections for parameters that affect 
the maximum temperature, such as the gas thermal conductivity, vapour condensation, gas reactions and 
radial temperature distribution inside the cavity (Suslick et al., 1990). A comparison of calculated 
maximum temperatures from different hydrodynamic models is shown in Table 2.2. 
Table 2.2; Theoretical muimwn temperature iD a coUapsing cavity cakulated from mathematical 
hydrodynamic models (SusUck et aL. 1990) 
Author'! Date Dissolved gas Maximum temperature 
(K) 
Noltingk and Neppiras 1950 N, 9500 
Hickling 1963 N, 3500 
Fogler 1969 Co, 1900 
Young 1976 AI 1650 
Fujikawa and Akamatsu 1980 air 6700 
Margulis and Dmitieva 1982 air 5300 
The theoretical temperature values reported in Table 2.2 cannot be compared directly as different initial 
conditions were assumed in each model and the relative imponance of the various corrections are difficult 
to establish (Suslick et al. , 1990). However, the values of the maximum temperature (> I 500 K) indicate 
that high temperatures are reached during cavity collapse. 
Young showed that the sonoluminescence from a range of noble gases. with the same specific heat ratio, 
increased with decreasing thermal conductivity (Young, 1976). This indicates that cavity collapse is not 
completely adiabatic, as assumed by Notingk and Neppiras (1950). and that heat is lost to the surrounding 
liquid (Riesz et al., 199Oa). The amount of heat lost during cavity collapse is dependent on the thermal 
conductivity of the gas. As reported in Table 2.1 , the estimated temperatures during cavity collapse for 
water saturated with heliwn. neon, argon, krypton and zenon were 815; I 420: 1 650; 1 890 and 2 000 K 
respectively (Young, 1976). 
Chapter 2 ULTRASOUND 2-20 
MBik and co-workers used the temperature dependence of the kinetic deuterium isotope effect (for the 
cleavage of o-H and o-D bonds, respectively, in H~ and IM) molecules) to investigate the cavitation 
temperature (Misik et al., 1995). The rates of formation of HO and DO radicals, caused by the cleavage of 
O-H and 0-0 bonds in a mixture of water (H~) and IhO, were measured during sonication by using 
various spin traps. The ratio of the rates of formation of the HO and D° radicals (kpjko) was compared \\itb 
the ratio of kPlko calculated, at various temperatures, using the semi-classical model of the kinetic isotope 
effect for HO and D° radical formation. The three spin traps gave different cavitation temperatures in the 
range of2 000 to 4 000 K (Misik et al ., 1995). 
flint and Suslick investigated the maximum cavitation temperature by measuring the sonolwninescence 
spectra from silicone oil (Flint and Suslick, 1991). Light emission occurred from the excited state of the Cz 
molecule. Synthetic spectra were generated, as a function of rotational and vibration temperatures, using 
diatomic-molecule emission theory and compared with the observed sonoluminescence spectra. From this 
comparison, the effective cavitation temperature was found. to be 5 075 ± 156 K (Flint and Suslick, 1991 ). 
In an investigation of the sonochemical ligand substitution rates of volatile metal carbonyls, Suslick and 
co-workers used comparative rate thermometry to determine that the temperature in a collapsing cavity was 
approximately 5 200 ± 650 K (Suslick et al, 1986). The actual temperature within a collapsing cavity may 
be lower than that predicted by the original hot spot theory yet experimental evidence indicates that 
significantly high temperatures are reached in the cavities. 
The high temperatures and pressures in the collapsing cavities cause the thermal dissociation of water 
vapour. as shown in Scheme 2.1, into hydro" .. yl HO° and hydrogen radicals H" (hydrogen and oJ\.')'gen atoms 
are referred to as radicals since they contain an unpaired electron). These radicals have been identified by 
reaction with spin trapping compounds (Mak,:ino et aI. , 1982: Makino et aI., 1983). 
H,O ~ HO· + H· 
Scheme 2.1 
Hydrogen peroxide formation during sonication was first proposed by Schmin and co-workers in 1929 to 
explain the sonochemical formation of iodine from a IXltassium iodide solution (Schrnin et al.. 1929). The 
decolourisation of potassium permanganate either during sonication or in water that had been sonicated was 
early proof that hydrogen peroxide was formed during sonication (Flosdorf et al ., 1936; Liu and Wu, 1934). 
This lead to the hypothesis that the action of ultrasonic waves was an indirecI ejfocr, in that the chemical 
changes were due to reactive compounds (such as hydrmq'l radicals) that were formed from the dissociation 
of water molecules and not from any direct ultrasonic attack on the solute (Miller, 1950). The 
sonochemical formation of hydrogen peroxide was proposed to be analogous to the formation of hydrogen 
peroxide from the radical reactions occurring during radiation chemistry (Weissler, 1959). Hydrogen 
Cbapter 2 ULTRASOUND 2-21 
abstraction from sodium formate during sonication was used to prove the formation of hydrogen radicals 
during sonication (Anbar and Pechl, 1964b); similarly, the oxidative deamination of ethylenediamine 
during sonication was reported as evidence for the formation of hydroxyl radicals during sonication (Anbar 
and Peeht, 1967). The formation of hydroxyl and hydrogen radicals during sonication has been proposed, 
and accepted, for many years, however, the electron spin resonance measurements by Makino and 
co-workers provided the first direct evidence for such formation (Makino et al., 1982~ Makino et al., 1983). 
Riesz and co-workers showed that the formation of hydroxyl radicals is directly linked to the temperature 
within a collapsing cavity (Riesz et al., 1990a). The different heat conductivity's of the noble gases was 
used to obtain different temperatures in the collapsing cavities. The production of hydroxyl radicals in 
water saturated with the different noble gases increased in the order for decreasing heat conductivity of the 
gases and thus increasing cavity temperatures (Riesz et al ., 199Oa). Hydroxyl and hydrogen radicals have 
been detected in pulsed (Christman et al., 1987) and continuous ultrasound (Makino et al., 1982). The 
occurrence of sonoluminescence from both stable and transient cavitation indicates that hydroxyl and 
hydrogen radicals are fonned in both types of cavitation., since sonoluminescence arises from radical 
recombination reactions and excited molecules (Crurn and Reynolds, 1985; Verrall and Sehgal, 1988). 
Hydroxyl and hydrogen radicals in the gas cavities react either with other radicals pnxlucing water, 
hydrogen peroxide and hydrogen gas, as shown in Scheme 2.2, or with other gaseous components such as 
volatile organic solutes (Sehgal et al., 1982; TOOd, 1970). Fischer and co-workers estimated that 
approximately 80 % of the radicals formed react to produce water (reaction la] of Scheme 2.2) in the 
absence of any volatile solutes (Fischer et al., 1986a). Hart and Henglein also reported that more radicals 
react to produce water, according to reaction la] of Sch.eme 2.2. than hydrogen peroxide, according to 
reaction [bJ of Scheme 2.2 (Hart and Henglein, J987). 
H- + HO' - H,O 
HO- + HO- - H, O, 
H· + H" .... H2 
Scbeme 2.2 
[at 
[h] 
[cl 
Conditions in a collapsing cavity are not only determined by the specific heat ratio and heat conductivity of 
the gas but also by the vapour pressure of any volatile solute in the cavity (Susiick et aI., 1983). Suslick and 
co-workers studied the sonochemicalligand substitution rates of volatile metal carbonyls, such as Fe(COh . 
Cr(CO~ Mo(CO)6 and W(CO~ to investigate the site of sonochemical reactions (Suslick and 
Hammenon, 1986; Suslick et al., 1986). The rate constants were found to increase linearly as the vapour 
pressure of the metal cartxmyls increased This is expected for gas phase reactions. since an increase in 
vapour pressure would cause an increase in solute concentration in the gas cavity and hence an increase in 
Chapter 2 ULTRASOVl'I"D 2-22 
the rate coefficient (Suslick and Hammerton, 1986: Suslick et aL, 1986). Suslick and co-workers also 
investigated the sonochemistry of non-aqueous liquids (Suslick et al. , 1983; Suslick et aI. , 1984). Two 
chemical dosimeters were used to study the sonochemical yields from alkane compounds, radical trapping 
by ctiphenylpicrylhydrazyl and decomposition of Fe(CO),. Both dosimeters indicated a decreasing linear 
relationship between the log of the sonocbemical rate and the solvent vapour pressure (Suslick et al., 1983; 
Suslick: et al., 1984). An increase in vapour pressure increases the solvent vapour in the gas cavity and thus 
decreases the intensity of the cavity collapse (the collapse is cushioned by the solvent vapour). This reduces 
the temperature reached during cavity collapse and thus lowers the rate of reaction (Suslick et al. , 1983; 
Suslick et al ., 1984). An investigation of the sonochemistry of alcohol-water mixtures by Krishna and 
co-workers also showed the effect of solvent vapour pressure on cavitation conditions and hence on 
sonochemical yields (Krishna et al., 1987; Krishna et al ., 1989). The degradation of alcohols (such as 
methanol, ethanol, I-propanol and 2-propanol) occur through bond cleavage and radical formation caused 
by thermal dissociation in the gas cavities (Krishna et al., 1987; Krishna et al., 1989). The investigation of 
the sonochemistry of non-aqueous solvents by Suslick and co-workers and the investigation of alcohol-water 
mixtures by Krishna and co-workers indicate that the radica1 reactions occurring inside the collapsing gas 
cavities are similar to radical reactions occurring during high temperature pyrolysis (Krishna et al., 1987; 
Krishna et al., 1989; Suslick et al., 1983; Suslick et aL , 1984). 
The second region where sonochemical reactions take place, as shown in Figure 2.6, is at the interface 
between the hot gas cavities and the bulk liquid Large temperature and pressure gradients exist in this 
region (Riesz. and Kondo. 1992). The investigation by Suslick and co-worken: of the sonochemical ligand 
substitution rates of volatile metal carbonyls found that the linear relationship between the sonocbemical 
rate constant and the vapour pressure had a non-zero intercept . This indicated that part of the reaction was 
not dependent on the vapour pressure and occurred in the liquid, probably at the interfacial region heated b)' 
the collapsing gas cavities (Suslick and Hammerton, 1986~ Suslick et al., 1986). A conduction model was 
used to calculate that the interfacial region extended for about 200 nm from the cavity surface, existed. for 
less than 2 }.LS after the cavity collapsed and had an approximate temperature of I 900 K (Suslick and 
Hammcnon. 1986; Suslick et aI. , 1986). 
The fonnation of hydrogen peroxide, water and hydrogen gas. according to the radical reactions listed in 
Scheme 2.2 also occur at the interface between the hot gas cavity and the bulk liquid. Weissler reponed 
that the yield of hydrogen peroxide decreased, for oxygen- and argon-saturated water, when volatile solutes 
such as aUylthiourea, formic acid or acrylamide were added to the water (Weissler, 1959). Hydrogen 
peroxide formation was reduced because the reactions listed in Scheme 2.2 are inhibited by the solutes 
reacting with the hydroxyl and hydrogen radicals. The most volatile solute would produce the largest 
reduction in hydrogen peroxide formation if the reactions only occurred in the gas cavities. However. the 
greatest reduction in hydrogen peroxide formation was not recorded for the addition of formic acid the 
Cbapter 2 ULTRASOUND 2-23 
most volatile solute. Thus the radical reactions must also occur outside of the gas cavities. at the liquid 
interface (Weissler, 1959). 
The concentration of a solute that decreases the hydrogen peroxide yield to 50 % of the yield observed in the 
absence of the solute is defined as the C~ concentration (Heng,lein and Konnann. 1985). This concentration 
was found to be dependent on the hydropbobicity of the solute and not on the rate constant for the reaction 
between the solute and bydroxyl radicals. Hydrophobicity is defined as the ratio of the number of 
hydrophobic groups (such as CH" CH, and CH) to hydrophilic groups (such as OH, COOH. CO and NH,) 
in a solute molecule (Henglein and Kormann, 1985). Riesz and co-workers investigated the sonochemiSU"y 
of volatile and nODMvolatile solutes, the interfacial region was found to have a low polarity and a high 
concentration of hydrophobic, nODMvolatile solutes (Rjesz et al., I99Ob; Riesz and Kondo, 1992). 
Organic radicals were generated in the interfacial region from the thermal decomposition of nonMvolatile 
solutes. The formation of these radicals was dependent on the ability of the solutes to accumulate at the 
cavity interface and on the activation energy (detennining ease of bond scission); the higher the solute 
concentration and the lower the activation energy. the higher the yield of thermal decomposition products 
(Krishna et al ., 1989). Products from reactions between solute molecules and hydrogen and hydroxyl 
radicals predominate when only low concentrations of nonMvolatile solutes are present at the interface 
(Kondo et al., 1993; Riesz et aI., I99Ob). 
The fonnation of supercritical water has been proposed to occur in the interfacial region since temperatures 
and pressures greater than the critical temperature (647 K) and critical pressure (22,4 MPa) of water are 
reached (Hua et al. , 1995b; Riesz and Kondo, 1992). The physicochemical properties of supercritical water, 
such as viscosity. density and heat capacity, are different to that of water, supercritica1 water acts as a more 
nonMpolar solvent because of its lower dielectic constant (Franck, 1987). This ex-planation has been used to 
account for the preferential accumulation of hydrophobic solutes in the interfacial region (Riesz and 
Kondo. 1992). Hua and c:o--workers investigated the formation of supercritica.l water during sonication by 
measuring the ultrasonic hydrolysis of p--nitrophenyl acetate (Hua et al. , 1995b). The reaction was 
acc:clerated by ultrasound by two orders of magnitude over the pH range of 3 to 8 (Hua et aI., 1995b). It wa s 
proposed that the reaction took place in a supercritical region at the interface because p-nitrophenyl acetate 
is nonMvolatile and relatively hydrophobic; hydroxyl radicals are not involved in the reaction mechanism. 
The changes in the tbermoctynamic activation parameters (AY, !J,.(Jl and MP) of the p-nitrophenyl acetate 
hydrolysis in sonicated and unsonicated solutions were explained in terms of the physical properties of 
supercritical water (Hua et al. , 1995b). A heat transfer model was used to estimate that after to ms. the 
radius of the supercritical region around a collapsed cavity extended about 40 % further into the bulk 
solution than the original cavity; also, approximately 0. 15 % of the water would be in the supercriticaJ state 
at any time during sonication (Hua et al .. 1995b). 
Chapter 2 ULTRASOUND 2-24 
The third region where sonochemical reactions take place, as shown in Figure 2.6, is in the bulk liquid at 
ambient temperature (Riesz et al., 1990b). Radicals produced. in the coUapsing cavities and at the cavity 
interface diffuse into this region and react with non·volatile solutes with kinetics analogous to that observed 
in aqueous radiation chemistry (Riesz et al., 1990b; Sehgal et al., 1982). Gutierrez and co-workers 
estimated that less than 10 % of the hydroxyl and hydrogen radicals formed in the gas cavity reach the bulk: 
solution (Gutierrez et al. , 1991). Products, such as hydrogen peroxide, from radical reactions occurring in 
the first two regions (as listed in Scheme 2.2) also di1fuse into this region and undergo secondar)' reactions. 
Anbar and Pecht showed that the radical reactions producing hydrogen peroxide do not occur in the bulk 
liquid (Anbar and Pecht, 1964a). Shock waves are generated in the bulk liquid when the liquid moves back 
into the volume occupied by the collapsing cavities. These shock waves generate shear forces that fragment 
polymers dissolved in the liquid (Webster, 1963). 
The relative contributions of reactions in the collapsing cavities, interfacial region and in the bulk liquid 
depend on the volatility and hydrophobicity of solutes (Riesz et al., 1990b). 
2.3.1.2 Effects of disiOlved gas 
Early ultrasonic investigations showed that the nature of a gas present during sonication affected the 
sonochemical reactions. Weissler and co-workers found that the amount of iodine liberated from a 
potassium iodide solution was dependent on the gas (air, oxygen, nitrogen, helium or carbon dioxide) that 
was present during sonication (Weissier et al., 1950). Parke and Taylor found that hydrogen peroxide 
formation differed when different gases (oxygen, argon or nitrogen) were present during sonication (Parke 
and Taylor, 1956). 
The investigation by Weissler into the effect of volatile solutes (allylthiourea, formic acid or acrylamidc) on 
the yield of hydrogen peroxide found that it decreased in the presence of the solutes, however, for each 
solute, the yield was higher when the water was saturated with o:\"ygen than with argon (Weissler, (959). 
This was attributed to the formation of additional hydrogen peroxicle from the reactions listed in 
Scheme 2.3. Hydrogen radicals, in the presence of oxygen, react with the oxygen molecules, as shown in 
reaction [a] of Scheme 2.3, to form perhydroxyl radicals CHOz). Hydrogen peroxide is formed from the 
reaction between a hydrogen radical and perhydroxyl radical (reaction [b] of Scheme 2.3) and from the 
reaction (reaction [c] of Scheme 2.3) between two perhyciroxyi radicals (E1'piner, 1959; Parke and 
Taylor. 1956: Weissler, 1959). 
H" + O2 .... HOi 
HOi + H" .... H10 2 
HO; + HO; .... H20 2 + O2 
Scheme 2.3 
la) 
[b) 
Ic) 
Chapter 2 ULTRASOUND 2-25 
The investigation of the sonochemical formation of hydrogen peroxide by Del Duca and co-workers using 
isotopic techniques estimated that one third of hydrogen peroxide in oxygen-saturated water was produced 
via the perhydroxyl radical reactions listed in Scbeme 2.3 and the remaining bydrogen peroxide from the 
hydroxyl radical reactions listed in Scheme 2.2 (Del Duca et al., 1958). It was found that the Q.() txmd in 
an oxygen molecule (Oz) was not broken and that the molecule was incorporated as a unit into a peroxide 
molecule (H2Dz). Other radical reactions that Del Duca and oo-workers proposed may occur during 
sonolysis, based on the standard free energy change in the gaseous state at 25°C, are listed in Scheme 2.4 
(Del Duca et al .. 1958). 
HO· + HOi - H20 + O2 
HO' + HO' - H,O + O' 
O· + H20 - H20 2 
HO· + HO· - H2 + O2 
Scbeme 2.4 
laJ 
[bJ 
IcJ 
Id] 
Mead and oo-workers compared the formation of bydrogen peroxide in water saturated with different gases 
(Mead et al., 1976). It was found that the hydrogen peroxide yield was a .maximum for oxygen saturation 
and decreased in the order for air, argon and nitrogen saturation. Hydrogen peroxide fonnation does not 
occur in water saturated with carbon dioxide (Henglein, 1985) or hydrogen (Gutierrez et aI., 1987; Hart and 
Henglein, 1981). Hydroxyl radicals are scavenged. during hydrogen saturation, according to the reaction 
listed in Scheme 2.5 and are thus unable to react with other radicals to produce hydrogen peroxide 
(Gutierrez et aI. , 1987; Hart and Henglein, 1987). 
HO· + H2 - H2 0 + H· 
Scheme 2.5 
A1though the yield of hydrogen peroxide in water saturated with pure gases is a maximum for o;.. .. ygen, 
higher yields have been observed for mix"tureS of oxygen with argon (Fischer et aI., 1986b; Han and 
Henglein, 1985) and with hydrogen (Hart and Henglein, 1987). The maximum hydrogen peroxide yield for 
gas mixtures of oxygen and argon occurs at a 30 to 70 % oxygen to argon ratio. An increasing o:-.:ygen 
concentration initially increases the formation of hydrogen peroxide because of perhydroxyl radical 
formation (according to Scheme 2.3). The temperature in the gas cavities decreases with increasing oX'ygen 
concentration (oxygen has a lower specific heat ratio than argon) causing fewer hydrogen and hydroxyl 
radicals to be fonned. The formation of hydrogen peroxide thus reaches a maximum and then decreases 
(Fischer et al., 1986b; Hart and Henglein, 1985). The investigation by Fischer and oo-workers using 11·1I0z 
showed that oxygen molecules dissociate in the gas cavities and are not included in a peroxide molecule as a 
whole unit as proposed by Del Duca in 1958 (Fischer et aI. , 1986b). 
Chapter 2 ULTRASOUND 2·26 
The maximum hydrogen peroxide yield for gas mixtures of oxygen and hydrogen occurs at a 80 to 20 % 
oxygen to hydrogen ratio (Han and Henglein, 1987). The hydrogen molecules dissociate into hydrogen 
radicals. as shown in Scheme 2.6. 
Hz - H" + H" 
Scbeme 2.6 
The yield of hydrogen peroxide is initially increased by the reaction of hydrogen radicals with oxygen to 
form perhydroxyl radicals (Scheme 2.3). Hydrogen peroxide formation decreased at higher hydrogen 
concentrations because the cavity temperature was lowered by the higher beat conductivity of hydrogen 
(Hart and Hengleio. 1987). No hydrogen peroxide is formed in water saturated. with gas mixtures of 
hydrogen and argon if the proportion of hydrogen is greater than 20 % (Gutierrez et al., 1987). This is due 
to the scavenging of hydroxyl radicals by hydrogen. as shown in Scheme 2.5 (Gutierrez et al., 1987). 
Hydrogen peroxide is the major product fonned during the sonication of water (Mead et al ., 1976). Other 
products. such as nitrous and nitric acid, are formed when a nitrogen--containing gas (N2 or NzO) is present 
during sonication (Hart et al., 1986; Hart and Henglein, 1986). Mead and co--workers found that the pH of 
water saturated with oxygen, air or argon decreased during sonication whereas pH during 
nitrogen-satwation decreased initially before remaining constant (Mead et al., 1976). 
2.3.2 Physical effects 
Sonochemica1 applications, such as catalysis, cleaning, emulsification and depolymerization, are possible 
because of the physical effects of ultrasound These effects. like the chemica1 effects of ultrasound are 
caused by the collapse of cavities in a liquid under the action of a sound wave. The two mechanisms 
responsible for the effects of cavitation in solid-liquid mixtures. as shown in Figure 2.7. are microjet impact 
and shock wave damage (Suslick. 1990). 
BULK 
SOLUTION 
shock wave 
fonnation 
ASYMETRlC 
CAVITY COLLAPSE 
microjet 
fonnation 
Figure 2.7: Scbematic diagram of tbe origin of ca ... itational pbysical effects 
SOLID 
SURFACE 
Cbapter 2 ULTRASOUND 2-27 
The changes in a sound field near a solid-liquid interface lead. to asymmetric cavity collapse. The cavity 
deformation occurring during the collapse is self-reinforcing and a stream of fast-moving liquid directed 
towards the solid surface, called a microjet, is generated (Olson and Hammin. 1969; Suslick et al., 1990). 
The velocities of1he microjets are estimated as being as high as lOO m s·\ (Suslick et al ., 1990). The impact 
of microjets on a solid surface causes loc:alised erosion or pitting. Ultrasonic pitting has been recorded 
photographically (Numachi, 1965; Olson and Hammin, 1969). The surface area of a solid is increased and 
new surface material is exposed tluough pitting. Ultrasound. can thus be used to activate solid catalysts 
(Mason and Cordemans, 1996). Cavity collapse is only distorted by a solid surface if the surface is 
oonsiderably larger than the resonance size of the cavity, for example, microjet formation does not occur for 
particles smaller than 200 J.UD in an ultrasonic field with frequency of 20 kHz (Suslick et al., 1990). Alex 
and co-v."Orkers also found that microjet formation was dependent on the size of solid particles. Reaction 
yields were greater for copper turnings than copper powder « 0,63 J.UD), and similarly for lead foil pieces 
and lead. powder (A1ex et al ., 1995). 
Shock wave formation also occurs during cavity collapse (Boudjouk. 1986). Shock waves break apart 
loosely aggregated panicles along existing cracks (increasing the surface area) and remove loosely adhering 
particles to a solid surface (Crawford, 1963). Reaction products are thus removed from a solid surface. 
leaving the surface clean and available for further reaction (Mason., 1990). The turbulent flow and the 
shock waves produced by intense ultrasound cause small metal particles to oollide with sufficiently high 
speed so as to induce melting at the point of oollision (Suslick et al ., 1990). Suslick and co-workers 
published scanning electron micrographs of zinc powders, before and after sonication., to show the fusion of 
particles due to interpanicle collisions (Suslick et al., 1990). 
Shock wave formation also results in high microntixing in the bulk liquid surrounding solid particles 
(Contantine el al.. 1994). Davidson and co-workers reported that ultrasonic mixing increases the product 
yield for the reaction of an amine with an alkyl halide in toluene solution in the presence of potassium 
hydroxide and a phase transfer catalyst (Davidson et al .. 1987). An 80 % yield was obtained for the 
reaction between indole and benzyl bromide with stirring at 20 DC for 8 h, the yield was increased 10 95 % 
for 2 h of sonication without any ex"temal stirring (Davidson et al. , 1987). Ultrasonic mLxing increases heat 
and mass transfer as well as the diffusion of chemical species inside the pores of the solid particles 
(Contamine et al .. 1994). 
The rapid coUapse of cavities also produces shear forces in the surrounding bulk liquid (Mason and 
Cordemans, 1996; Webster, 1963). This causes the breakage of chemical bonds in any polymeric materials 
dissolved in the liquid A pair of macromolecular radicals is formed from the break in the polymer-chain, 
these radicals recombine randomJy and reduce the molecular weight of the polymer (Mason and 
Conlernans, 19%). 
Cbapter 2 ULTRASOUND 2-28 
In a solution of two immiscible liquids, ultrasonic cavitation leads to the formation of an emulsion . An 
emulsion oc:cun when microscopic droplets of one liquid are suspended in another liquid (Suslick, 1989). 
The cohesive forces that hold a droplet together are overcome by the conditions created during cavity 
collapse, this causes a droplet to burst into smaller droplets and the liquids become emulsified 
(Suslick, 1989). Ultrasonic reaction enhancement between immiscible liquids is due to the large increase in 
the interfaciaJ contact area during emulsification (Pestman et al., 1994). 
2.4 APPLICATIONS OF ULTRASOUND 
The different applications of ultrasound are reviewed in the following sections, physical applications in 
Section 2.4. 1, chemical applications in Section 2.4.2, medical applications in Section 2.4.3 and industrial 
applications in Section 2.4.4. 
2.4.1 Physical applications 
The first application of ultrasound, underwater depth determination, was made possible by Paul Langevin 
building the first underwater sound transducer (based on the piezoelectric effect in quartZ crystals) . The 
underwater transducer was initially developed for detection of submarines during World War I, however, it 
was not completed in time and was used for depth determination after the war (Mason, 1976). During 
World War n, ultrasound was used for both the detection and destruction of submarines. Ultrasonic 
listening devices mounted on torpedoes allowed the sound from the target submarine or ship to control the 
direction of the torpedo (Mason, 1976). 
The ultrasonic flaw detection system to detect flaws in materials and to measure the thickness of materials 
"ith only one surface exposed developed independently in America and Britain between 1939 and 1945 
(Mason, 1976). In 1945. ferroelectric materials such as barium titanate became available in ceramic form 
(Mason. 1976). These ceramics behave as piezoelectric cT)'stals when polarised by means of a high voltage 
applied at a high tem~rature . Ferroelectric materials have a natural polarisation. yet in the unpolarized 
state the polarisation occurs in a random direction. The availability of ferroelectric ceramics increased the 
production of ultrasonic transducers and a1lowed for the diversification of ultrasonic applications 
(Mason. 1976). 
Ultrasound is widely used in industry for non-destructive testing and fluid flow measurement (Hoyle and 
Luke. 1994). Application in fluid flow measurement has lead to ultrasound being used in chemical analysis 
(Asher, 1987). The measurement of ultrasonic velocity can be used for liquid identification (for example, 
the grade of petrol passing along a pipeline has been identified using ultrasound), delennination of solution 
concentration and composition of liquid mh.:tures (Asher, 1987). Advantages of ultrasonic meters are that 
the response is rapid, radioactivity is not used. mQving parts (that could lead to mechanical failure) are not 
employed and the system is non-invasive (Ashcr, 1987), 
Chapter 2 ULTRASOUND 2-29 
2.4.2 Chemical applications 
The earliest chemical application of ultrasound was perfonned by Richards and Loomis in 1927. Reactions 
such as the hydrolysis of dimethyl sulfate and the iodine clock reaction were enhanced by ultrasound 
(Richards and Loomis, 1927). Early application of ultrasound in organic chemistry was demonsttated by 
Zechmeister and co-workers in the 1950's. Ring cleavage in compounds containing aromatic or 
heterocyclic rings. such as bromobenzene, phenol and pyridine, occurred during sonication (Currell and 
Zechmeister, 1958; Zechmeister and Wallcave, 1955). 
Ultrasound has the following benefits when applied to chemica1 reactions (Mason, 1990). 
• Ultrasound can be used to accelerate reaction rates. Mason and co-workers found that the 
homogeneous solvolysis of 2-<:bloro-2-methylpropane in aqueous ethanolic mixtures showed a 
20-fold rate increase at 10 DC in 50 % ethanol (Mason et al., 1985). Kristol and co-workers found 
that ultrasound increased the rate of hydrolysis of a series of nitrophenyl esters by up to 15 % 
(Kristol et al .. 1981). 
• 
• 
Ultrasound can be used to avoid forcing conditions. The production of transition metal carbonyl 
anions from the metal halide, sodium and tetrahydrofuran requires a high temperature and 
pressure. Suslick and Johnson found that the use of ultrasound for the reaction listed in Scheme 2.7 
decreased the required temperature from 160 to 10 DC, and the required pressure from 20.3 to 
0.45 MPa, while still giving a comparable yield of the end product (Suslick and Jobnson, 1984). 
VCJ, .(THF), ~ V(COk 
Scheme 2.7 
Ultrasound can allow for the use of simpler and safer procedures. The preparation of highly 
reactive Rieke metal powders involves refluxing the metal halide with potaSsium (the reducing 
agent) in tetrahydrofuran for extended periods of time (Rieke and Rhyne, 1979). Metal powders 
",ith similar reactivity can be prepared using a less hazardous procedure if the p.:ltassiunl is 
replaced by lithium and the reaction mixture is sonicated.. For example, the formation of Rieke 
copper powder. that usually requires refluxing the metal halide (CuBr} or Cul2) in tetrahydrofuran 
for 8 h. was complete within 40 min (as shown in Scheme 2.8) if the reaction was sonicated in an 
ultrasonic bath (Boudjouk et a1.. 1986). Similarly. zinc powder was produced from ZnBr2 within 
40 min with sonication instead of refluxing in tetrahydrofuran for 4 h (Boudjouk et al .. 1986). 
CuBr} + 2Li ..... Cu· + 2LiBr 
Scheme 2.8 
Cbapter 2 ULTRASOUND 2-30 
• The use of ultrasound reduces particle size and allows for the continuous activation of surfaces. 
The rate of reaction of the Ullmann coupling reaction of 2-iodonitrobenzene, in the presence of a 
4-fold excess of copper powder (as shown in Scheme 2.9), increased over 50-fold with ultrasound 
(Lindley et al., 1986; Lindley et al. , 1981). Scanning electron micrographs showed that ultrasonic 
pretreatment increased the reaction rate because of a reduction in size of the copper particles 
(Lindley et al ., 1981). The reaction rate was further increased ifultrasound was applied continually 
through the reaction. The ultrasonic-induced over SO-fold increase in reaction rate was thus not 
only due to a reduction in panicle size but also due to the continual activation of the copper surface 
(Lindley et al., 1986; Lindley et al ., 1987). 
Q-I 
NO, 
ell, DMF, N2 
• 60 oc. 60 h 
Scbeme 2.9 
NO, 
NO, 
• Ultrasound can be used to accelerate phase transfer catalyst reactions. Phase transfer catalysts act 
as a bridge between aqueous and organic phases in a reaction~ reactants are carried one way and 
products returned the other way (Bremner, 1986). The addition of diethylmalonate to chalcone in 
toluene, using potassium bydroxide as a solid catalyst and trimethylbenzylammonium chloride as a 
phase transfer catalyst has been used as a model reaction to investigate the effect of ultrasound on a 
phase transfer catalyst reaction (Contamine et al .. 1994; Ratoarinoro et al.. 1995). The reaction 
does not occur (with stirring) in the absence of the phase transfer catalyst within a reaction time of 
2 min. Without the phase transfer catalyst but 'With sonication, the yield increased to 40 % after a 
2 min period. In the presence of the phase transfer catalyst, the yield was 5 % after stirring for 
2 min and 91 % for sonication (Contamine et al.. 1994: Ratoarinoro et al .. 1995). 
• Ultrasound can be used to change a reaction pathway. Ando and Kimura have proposed that 
ultrasound switches the pathway for the reaction of benzyl bromide, potassium cyanide and alumina 
stirred in toluene from the Friedel-Crafts reaction to nucleophilic substitution (Ando and 
Kimura. 1990) . A mixture of 0- and p-benzyltoluene (75 %) was produced in the absence of 
ultrasound but benzyl cyanide (71 %) was formed when the reaction mixture was sonicated 
(AnOO et al .. 1984). The change in pathway was caused by the ultrasonic acceleration of the 
poisoning of the acidic sites on the alum.ina surface by the potassium cyanide so that the reaction 
was catalysed instead by the basic sites on the alumina surface (Ando and Kimura, 1990). 
Sonochemical switching has also been observed in the Kornblum-Russell reaction (Luche. 1992). 
The alkylation of nitronate anions by 4-nitrobenzyl bromide follow either a polar two-electron 
Chapter 2 ULTRASOUND 2-3\ 
pathway leading to the O-alkylation product, or a single electron transfer leading to the 
C~alkylation product. Luche used the reaction of an ethanolic solution of 4.nitrobenzyl bromide 
and 2·lithio--2·nitropropane to show that the ratio of the CIO alkylated products was almost 
reversed with the application of ultrasound.. The single electron transfer pathway was the dominant 
mechanism during sonication (Luche, 1992). 
The benefits as described above indicate that ultrasound enhances roth homogeneous and heterogeneous 
reactions. The chemical and physical effects of ultrasound that causes such reaction enhancement are 
described in Section 2.3.1 and Section 2.3.2. U1trasonic reaction enhancement is greater in heterogeneous 
reactions and the industrial application of ultrasonic chemical reactions is more feasible for reactions 
incorporating a solid phase or catalyst (Bremner, 1986). Liquid·liquid reactions, another heterogeneous 
application, are also enhanced. by ultrasound because of the fonnation of emulsions. Davidson and 
co.workers investigated the saponification of naturally occurring and commercially imponant waxes, such 
as wool .... 'axes (Davidson et al., 1987). Sonication led to the fonnation of an emulsion of the wool wax 
dissolved in ethanol and the aqueous sodium hydroxide, this decreased the rtaetion time and temperature at 
which saponification occurred. Another advantage of ultrasound was that the products fonned during 
sonication exhibited less colour than that produced in the absence of ultrasound (Davidson et aI., 1987). 
Ultrasound has also been used to improve the liquid·liquid extraction of acetic acid with water from a 
methyl isobutyl ketone·acetic acid mixture (Flisak and Pema, 1977). An increase of I1 % in the rate of 
acetic acid extraction was observed when an ultrasonic transducer. operating at 60 W, was attached to the 
side of the ex1raction column so as to cause the ultrasonic energy to be propagated at right angles to the 
direction of liquid flow (Flisak and Pema, 1977). 
The follo .... ing chemical applications of ultrasound have been reviewed separately. organic synthesis in 
Section 2.4.2.1, sonocatalysis in Section 2.4.2.2, polymer chemistry in Section 2.4.2.3, sonoelectrochemistry 
in Section 2.4.2.4 and sonocrystallization in Section 2.4.2.5. 
2.4.2.1 Organic syntbesis 
In organic synthesis, the use of ultrasound can reduce reaction time and by·product formation. and increase 
the reaction yield (HengJein. 1987). Luche and Damiano investigated the application of ultrasound in the 
fonnation of organometallic compounds, frequently·used chemical reagents (Luche and Damiano, 1980). 
Activating agents are used to shorten the induction period during which no reaction between the organic 
halide and the metal occurs. Ultrasound caused the reaction to occur almost immediately without requiring 
an activating agent. Reaction yields of 90: 61 and 95 % were obtained, respectively, in the preparation of 
n·propyl·, n-butyl. and phenyllithium from the bromide compounds with a sonication period of I h (Luchc 
and Damiano, 1980). 
Chapter 2 ULTRASOUND 2-32 
The Barbier reaction, a one-step coupling of an organic balide with a carbonyl compound in the presence of 
lithium or magnesium, was also found to be enhanced by ultrasound (Luche and Damiano, 1980). The use 
of ultrasound allows for such reactions to be performed in ""'Cl technical grade tetrahydrofuran instead of in 
dry tetrahydrofuran. The use of wet solvents, at room temperature, has potential for large-scale production 
(Luche and Damiano, 1980). Under ultrasonic conditions, yields ofbetv.'een 76 and 100 % were obtained. 
for the reaction of organic balides (alkyl, aryl, benzyl, allyl and vinyl halides) with ketones and aldehydes 
(Luche and Damiano, 1980). A 100 % yield was obtained after a reaction time of 10 min for the reaction, 
in the presence of lithium wire, shown in Scheme 2.10 (Luche and Damiano, 1980). 
~ U. lliF A~O/-CHO + CH,! -=--"-" ... 
IOmin ~ OH 
Scheme 2.10 
Scanning electron microscopy has been used to show that the ultrasonic enhancement of these 
organometallic reactions is at least partly due to the effect of ultrasound on the lithium surface 
lde Souza-Barboza et al ., 1988). Electron micrographs show the significant corrosion pits (thus increasing 
surface area) that formed in the lithium during sonication (de Souza-BarOOza et al., 1988). Luche and 
co-\\"Orkers, however, showed that ultrasound also has a direct chemical effect on the BaJbier reaction 
(Luche et aL, 1990). The lithium metal was preactivated under highly energetic conditions (in the absence 
of the reagents) so that the mechanical effects of ultrasound would not be dominant. The reaction was 
carried out at both a low and high ultrasonic intensity; the product concentration was found to be greater at 
the higher intensity (Luche et al., 1990). The difference in product concentration was attributed to the 
direct effect of ultrasound on the rate determining step of the reaction sequence. This step is the electron 
transfer from the metal to the carbon-haJogen bond producing the radical anion, the first reactive species of 
the Barbier reaction (Luche et aL. 1990). 
Boudjouk and co-workers investigated the use of ultrasound in the preparation of symmetrical organic and 
bimetallic compounds (Boudjouk, 1986). The Wurtz-type coupling reactions of organic halides (shown in 
Scheme 2.11 ) and organometallic chlorides (shown in Scheme 2.12) using lithium wire and dry 
tetrahydrofuran al room temperature takes place only in the presence of ultrasound The yield of the 
reaction shown in Scheme 2.11 is 70 % for a reaction of 10 h (Ran and Boudjouk. 1981). The yield of the 
reaction shown in Scheme 2. 12 is 68 % for a reaction time of 10 h (Boudjouk and Han. 1981). 
2CJi~Br ..... H~C6-CJi~ + Br2 
Scheme 2.11 
Chapter 2 ULTRASOUND 2-33 
2Ph,HSiCl ~ Ph,HSi-SiHPh, + Cl, 
Scbeme 2.12 
The coupling reaction to produce tetramesityldisilene, a compound discovered by West and co-workers 
where the silicon-silicon double bond is stabilised by four mesityl groups (West et al., 1981), was prepared 
ultrasonically by Boud.jouk and. co-workers (Boucljouk et al ., 1982). The reaction was carried out with 
lithlum wire in tetrahydrofuran. Tetramesitylclisilene was prepared ultrasonically in one step, as shown in 
Scheme 2.13; the yield of the reaction is 90 % for a reaction time of 20 min {Boucljouk et al. , 1982). The 
symbol))) denotes ultrasonic radiation. 
Mes... ,Mes 
Si=Si 
Mes/ 'Mes 
Scheme 2.13 
Boudjouk and co-workers also investigated other metal reactions such as the Refonnatsky reaction (Han and 
Boudjouk., 1982b) and the generation of o-xylylene (Han and Boudjouk., 1982a). The Refonnatsky reaction 
is the reaction of ethyl bromoacetate with activated zinc dust and aldehydes or ketones to give the 
!3-hydro,,')·t ester (Han and Boudjouk. 1982b). The yield of the reaction of cyclopentanone (as shown in 
Scheme 2. 14) in the presence of ultrasound is 98 % after 30 min at a reaction temperature of 25 to 30°C. 
The yield in the absence of ultrasound is 50 % after 12 h at a reaction temperature of 80 °C (Han and 
Bou<Uouk. 1982b). 
0=0 Zn B,cH,co,C,H, 
25 • 30 °C, 30 min 
Scheme 2.14 
O<0H CH,CO,C,H, 
The reaction of Q..Q. '~bromo-o-xylene, zinc and a dienophile does not proceed in the absence of ultrasound 
(Han and Boudjouk., 1982a). A high yield of cycloaddition products is obtained with ultrasound: the 
reaction proceeds via the reactive intermediate, o-xylylene. The yield of the reaction shown in Scheme 2. 15 
is 89 % for a reaction time of 15 h at a reaction temperature of 20 to 25 DC (Han and Boudjouk. 1982a). 
Chapter 2 ULTRASOUIII'l) 2-34 
o 
cc::~ [tcJ +Go-
 o 
Scheme 2.15 
Ultrasound can be used to accelerate the reduction of aromatic halides (Han and Boudjouk, 1982c). These 
compounds do not readily undergo nucleophilic substitution and electron-donating groups on the ring 
fwther impede the displacement of the halogen. The yield of the reduction of p-bromoanisole to anisole in 
the presence of lithium aluminium hydride (as shown in Scheme 2.16) is 70 % after 7 b of sonication (Ban 
and Boudjouk, 1982c). A yield of 35 % is obtained without sonication after refluxing with an excess of 
lithium aluminium hydride in diglyme for 24 h at 100 QC (HaD and Boudjouk. 1982c). 
CHJO-o-' Br LWH.,D~ CHJO-Q' 
_ 35OC, 7b _ 
Scheme 2.16 
2.4.2.2 Sonocatalysis 
Catalysts are classified as either homogeneous or heterogeneous. A homogeneous system exists if the 
catalyst is a molecular or ionic species dissolved in a liqui~ a heterogeneous system exists if the catalyst is a 
solid and the reactants are either in a percolating gas or liquid (Suslick, 1990). Both types of catalysts may 
be difficult to activate or to keep active (Suslick, 1990). The advantages of the application of ultrasound to 
catalysis are the use of ambient temperatures to preserve thermally-sensitive substrates and to enhance 
selectivity, the ability to generate high-energy species that are difficult to obtain from photolysis or 
pyrolysis, and the production of high temperatures and pressures during cavitaLion that generate autoclave 
reaction conditions on a microscale (Suslick. 1990). 
Organometallic compounds are used as homogeneous catalysts. these compounds arc often catal)1icaUy 
inactive until the loss of metal-bonded ligands such as carbon monoxide (Suslick, \990). Suslick and 
co-workers investigated the sonolysis of iron carbonyl compounds for the catalysis of olefin isomerizaLion 
reactions (Suslick et al .. 1981 : Suslick et al., 1983). Ultrasound induces ligand dissociaLion of iron carbonyl 
Fe(COh producing Fe3(CO)12 and finely divided iron in the absence of additional ligands (Suslick et al .• 
1981). The primary process Fe(COh undergoes during sonication is shown in reaction la] of Scheme 2.17. 
The comp:>und F~(CO:l9 is fonned from Fe(CO),. according to reaction £bl of Scheme 2.17; and FC3(CO)1 2 
is fonned from Fe(COh according to reactions le} and [d] of Scheme 2.17 (Suslick et aI., 1981 ). 
Chapter 2 ULTRASOUND 
Fe(CO), - Fe(CO),_. + nCO (n = I to 5) 
Fe(CO), + Fe(CO), - F.,(CO), 
Fe(CO), + Fe(CO), - F.,(CO). 
Fe,(CO). + Fe(CO), - Fe,(CO)" + CO 
Scbeme 2.17 
2-35 
la] 
[b] 
Ic] 
Id] 
Sonication of l -pentene in a decane solution with either Fe(COh, F~(CO» or F~(CO)\2 produces trans 
and cis-2-pentene in approximately a 3: I ratio, the thermodynamic ratio (Suslick et al., 1981). Sonolysis of 
l-pentene with Fe(COh. Fe2(CO), or F~(CO)12 enhances the rate of isomerization of J -pentene by factors 
of approximately 1O~ , 102 and 102, respectively (Suslick et al ., 1981). 
The sonolysis of Fe(COh in the presence of additionalligands (such as phosphine or phosphite compounds) 
produces Fe(CO»)l.z and Fe(CO).L, where L is a phosphine or phosphite compound (Suslick et al., 1983). 
Similar substitution patterns have been observed for other metal carbonyl compounds, such as F~(CO)I2. 
Mn2(CO)\G. Cr(CO)6. Mo(CO)6 and W(CO)6 (Suslick et al., 1983), Similarities between the sonocatalysis 
and photocatalysis of metal carbonyl compounds exist. however, different relative efficiencies and 
selectivities are observed (Suslick et al .. 1983). 
Ultrasound also enhances the preparation of the transition-metal carbonyl complexes. Suslick and Johnson 
found that where high pressures of CO and high temperatures are usually required to produce the 
transition-metal carbonyl anions, the use of ultrasound allows for the reaction to be carried out at low 
temperatures and pressures (Suslick and Johnson, 1984). For the reaction listed in Scheme 2.7. sonication 
allowed for the required temperature to be decreased from 160 to 10 °C. and the required pressure from 
20,3 to 0,45 MP.a, while still giving a comparable yield of the end product (Suslick and Johnson, 1984). 
Heterogeneous catalysts play an imponant role in industry. for example the bulk of the petroleum industry is 
based on a series of catalytic transformations (Suslick. 1990). These catalysts are often produced from rare 
and expensive metals. the use of ultrasound could allow for the activation of less reactive (and less 
expensive) metals (Suslick., 1990). Suslick and Casadonte found that the sonication of nickel powder 
increased its activity as a hydrogenation catalyst by a factor greater than 10' (Suslick and Casadonte. 1987). 
Nickel ~wder is usually an inactive catalyst for the hydrogenation of alkenes. ultrasonic pretreatment for 
1 h activated the ~wder and caused the hydrogenation of l-nonene to proceed with an initial rate of 
1,5 nunol L·l min·l (Suslick and Casadonte, 1987). Sonication did not reduce the panicle size of the 
powder, thus the surface area was not increased and the sonocatalytic activity was attributed to the changes 
Cbapter 2 ULTRASOUl\I'D 2-36 
in the panicle aggregation, surface morpbology and thickness of surface oxide coating. Electron 
micrographs revea1ed that sonication smoothed the initial highly crystalline structure of nickel and 
i.ncreased the aggregation of the powder. Auger electron spectra depth profiles sbowed that the initial oxide 
coat that extended for 25 x 10" m with a surface NilO ratio of J was decreased to a thickness of less than 
5 x 10.\"1 m with a surface NilO ratio of 2. The changes in the nickel powder that led to the increased 
sonocatalytic activity were caused. by interpartic1e collisions driven by the turbulent conditions and shock 
waves created by the ultrasonic field (Suslick and Casadonte, 1987; Suslick et al., 1989). 
Suslick and co-workers investigated the ultrasonic-induced changes of copper powder (Suslick et al. , 1989). 
Similar to nickel powder, ultrasound did not reduce the copper powder particle size (and bence increase the 
surface area); electron micrographs showed an increased aggregation and surface smoothing with 
sonication. The partial oxide coating, approximately 30 x 10.\"1 m deep, was completely removed during 
sonication and was replaced by a thin, approximately 10 x 10-9 In, but heavy (greater than 60 % atomic 
composition) surface dep)sition of carbon. The origin of the carbon and its influence on the reactivity of 
the copper powder are not known (Suslick et al., 1989). The changes in the copper powder were also 
produced by the ultrasonic-induced interpartic1e collisions. Direct particle collisions produce agglomeration 
and collisions at an angle produce the surface smoothing and removal of the surface oxide coating 
(Suslick et aI., 1989). 
2.4.2,3 Polymer cbemistry 
Research into the effects of ultrasound on polymer chemistry began in the 1930' s (Price and Smith, 1993). 
Most of the early work, such as the investigation by Thomas and de Vries of the heterolytic cleavage of 
polymethylsiloxane (Thomas and de Vries, 1959), focused on the degradation of polymer chains during 
sonication. Ultrasound however. can also be used to initiate polymerisation (Kruus et al.. 1988). 
Bonds of polymers are broken because of the hydrodynamic shear forces produced in the liquid surrounding 
the vibrating and collapsing cavities (Henglein and Gutierrez.. 1988: Webster. 1%3). The cleavage of a 
polymer does not occur randomly but statistically near the middle of the chain (Price and Smith. 1993). 
Polymer cleavage produces a pair of macromolecular radicals that react with other radicals and reduce the 
molecular weight of the polymer (Mason and Cordemans. 19%). A limiting molecular weight exists below 
which no further cleavage occurs. The sonication of a 1 % solution of polystyrene in toluene at 25 DC 
showed that polymer degradation was more efficient at higher molecular weights and that no further 
degradation occurred below the limiting value of approximately 30 ()()() fOT this system (Price. 1993). 
Price and Smith investigated the role of parameters such as ultrasonic intensity, temperature and dissolved 
gas on the degradation of polystyrene in toluene (Price and Smith. 1993). The rate of degradation was 
found to increase. although not indefinitely, ",ith increasing intensit)' and an optimum rate was reached at 
Chapter 2 ULTRASOUND 2-37 
approximately 145 W cm·l . Polymer degradation increased with decreasing temperature over the range of 
·10 to 61 °C (Price and Smith. 1993). Polymer degradation was also affected by the presence of a dissolved 
gas but because bond cleavage was caused by the hydrodynamic shear forces (and not of thermal origin) the 
degradation rate could not be directly correlated to gas properties such as thermal conductivity and specific 
heat ratio. Polymer degradation rate, however, tended to be greater for gases with a lower solubility 
(Henglein and Gutierrez, 1988; Price and Smith. 1993). The rate of polymer degradation can also be 
enhanced by using a solvent \\-ith a low vapour pressure, reducing the ultrasonic frequency and decreasing 
the solution concentration (Lorirner, 1990). 
The non· random nature of the polymer cleavage allows for the modeUing of the process so that the polymer 
molecular weight during sonication can be predicted (Price, 1993). Manipulation of the experimental 
conditions allows for control of the limiting molecular weight. Price and Smith found that the limiting 
molecular weight increased linearly with increasing temperature, decreased linearly with increasing 
ultrasonic intensity and was lower for gases with a lower solubility (Price and Smith. 1993). 
Macromolecular radicals produced from two polymers sonicated in a common solvent react with each other 
to produce the cross product called a block ~polymer (Price, 1993; Price and Smith. 1993). This p rocess 
is not easily manageable since both the recovery of the ~polymer and the control of the ~polymer length 
arc difficult. This technique, however, is of benefit if a polymer is sonicated together with a monomer such 
lhat the macromolecule radicals from the polymer initiate the polymerisation of the monomer (Price, 1993). 
The example shown in Figure 2.8 is of the formation of the block co-polymer, poly(styrene-b-methyl 
methacrylate), from the sonication of polystyrene and methyl methacrylate (Price. 1993). 
») RH ooR - -
 '" I 
M~ ~o Me Me 
0 , 
m 
0 0 
Me 0 0 
\ \ 
Me Me 
Figure 2.8: Scbematic diagram of the production of end capped polyst~'rene and 
poly(styrene-b-metbyl methacrylate) block co-polymers (Price, 1993) 
A radical initiator is required to start a polymerisation reaction. These compoWlds are usually activated by 
ultraviolet irradiation or by heating to above 60 QC (Price ct al., 1991) . .Price and co-workers used the 
polymerisation of methyl methacrylate "ith az~is(isobut)'ronitrile). a common radical initiator. to show 
Chapter 2 ULTRASOUND 2-38 
that ultrasound accelerates the formation of radicals from initiators (Price et al. , 1991). Ultrasound has also 
been used to initiate polymerisation reactions in the absence of radical initiators. Kruus and co-workers 
showed that pure, dry, vinyl monomers (such as styrene) can be polymerised solely by the application of 
ultrasound {Kruus et al., 1988). Price and co-workers polymerised methyl methacrylate "\\;th ultrasound in 
the absence of a radical initiator (Price et al., 1991). The methyl methacrylate conversion (fraction of the 
monomer converted. to polymer) after 4 h of sonication was 12 % for the solution containing azo-bis(iso 
but)TOnitrile) and. 6 % without the radical initiator. The conversion did not proceed further than 12 % as 
the solution became viscous and cavitation was suppressed (Price et al., 1991). The suppression of 
cavitation with increasing viscosity has led to conversion rates of ultrasonic polymerisation reactions to not 
ex.ceed 25 % (Price and Lenz, 1993). 
Price and co-workers monitored the poly(methyl methacrylate) molecular weight and molecular \-\'Ciglu 
distribution for 8 h of sonication; long polymer chains were initially formed but these were broken down 
after further sonication (Price et al ., 1991). Polymerisation and polymer degradation were found to occur 
concurrently. In the investigation of the effect of argon on the polymerisation of styrene, Kruus and 
co-workers found that polymerisation took place when argon was butj:)led through the solution and that 
degradation occurred when the argon flow was turned off (Kruus et al., 1988). The molecular weight of a 
polymer produced from ultrasonic polymerisation is thus determined by parameters such as reaction time, 
ultrasonic intensity and gas present during sonication {Kruus et al. , 1988; Price et al., 1991). 
2.4.2.4 Sonoeiectrocbemistry 
Electrochemical methods are used to produce reactive intermediates; organic electrosynthesis can generate 
compounds that may be difficult 10 prepare by other methods (Mason et al., 1990). The scale·up of 
laboratory electrolytic processes is hindered by restrictions to the maximum attainable current (such as the 
mass transfer to the electrode surface, fouling of the electrode surface and evolution of gases dwing 
reaction) and thus the maximum possible rate of reaction (Mason et al ., 1990). Ultrasound produces the 
follo";ng benefits when applied to an electrochentical process (Mason et al ., 1990; Reisse et al. , 1994: 
Walton et al ., 19%). 
• 
• 
• 
• 
Limited gas bubble accumulation at the electrode. 
Disturbance of the diffusion layer, thus the depletion of the electroactive species is stopped. 
More even transport of ions across the electrode double layer. 
Continual cleaning and activation of the electrode surface. 
The potential for ultrasound to enhance electroplating by increasing the de~ition rate has been 
investigated since the 1950's (Mason et al .. 1990). Ultrasound is introduced into the electroplating process 
by sonicating the whole plating bath. Reisse and co-workers designed a $Onoelectrochemica.l reactor in 
which the titanium ultrasonie horn acts as the worldng electrode (Reisse et at., 1994). The planar circular 
Chapter 2 ULTRASOUND 2-39 
surface at the bottom of the horn acts as the electroactive part of the sonoelectrode and the immersed 
cylindrical part of the horn is covered by an isolating plastic jacket. The cathodic current during the 
investigation of the electrodeposition of copper (at ~,20 V relative to the standard hydrogen electrode) was 
found to be approximately 30 times greater with ultrasound than without (Reisse et al. , 1994). Scanning 
electron micrographs showed that ultrasound also affected the morphology of the copper deposits, finer 
JXlwders with a higher density of submicronic particles were produced. Nucleation of copper on the 
titanium sonoelectrode occurred at a higher overpotential than that occurring in the absence of ultrasound 
(Reisse et al ., 1994). 
The electrochemical synthesis of selenium and tellurium anions (Sel-, Se2-, Tel and Tel "), intenned.iates in 
the formation of organoselenium and organotellurium compounds, have been enhanced by ultrasound 
(Gautheron et al. , 1985)_ Mason and co-workers showed that ultrasound switches the reaction mechanism 
of the Kolbe reaction,. the formation of alkanes from the electrodecarboxylation of carix:lxylate salts such as 
RCOO~a"', from a single electron-transfer process to a process involving two electron transfers 
(Mason et al. , 1990). The change in product distribution because of a change in reaction mechanism is 
shown in the sonoelectrolysis of cyclohexanecarboxylate (Chyla et al. , 1989). Ultrasound decreased the 
yield ofbicyclohexyl from 49 to 7,7 % and increased the yields of cyclohexene and cyclohexyl methyl ether 
from 4,5 and 2,1 %, respectively, to 32,4 and 34,3 % (Chyla et al. , 1989). The reaction efficiency was also 
increased by ultrasound, the average cell potential for the required current density of2 000 A m-l was 8,3 V 
in the absence and 7,3 V in the presence of ultrasound (Chyla et al., 1989). Walton and co-workers used 
ultrasound to enhance electTOchemical reactions producing hydrogen and chlorine gas (WaIton et al. , 1996). 
The reactions are reversible and are limited by the removal of gaseous products from the electrode. 
Ultrasound did not enhance the ox-ygen-producing reaction for the e:\:perimental conditions used in the 
investigation (Walton et al. , 19%). The application of ultrasound to such gas-producing reactions has 
implications for the operation of chloralkali cells (Walton et al., 1996). 
2.4.2.5 Sonocrystallization 
Ultrasound can be used to precipitate solids from a supersaturated solution because of the physical effects, 
such as panicle fragmentation, produced by cavitation (Mason. 1990). Finely divided and unifonn panicles 
are produced by ultrasound because of the cycle of formation,. growth and break-up of seed particles 
producing funher seeding throughout the mediwn (Mason, 1990). Srinh'asan and co-workers found that 
the application of ultrasound to the crystallisation of diphenyl oxide and dimethyl phenyl carbinol 
(compounds used to perfume soaps) improved crystal purity (Srinivasan et al., 1995). Ultrasound caused. 
the crystal melting point for both diphenyl oxide and dimethyl phenyl carbinol to match., more closely, the 
melting )Xlint of the pure com)Xlunds. Ultrasound also improved the shape of the diphenyl oxide cl)'stals_ a 
higher proportion of need1e-shaped crystals (the crystal structure of pure diphenyl oxide) was obtained 
under sonication (Srinivasan et al., 1995). 
Chapter 2 ULTRASOUND 2-40 
2.4.3 Medical applications 
Ultrasound in medical applications is used for either interaction with a system producing a desired 
biological effect such as in physical therapy and surgery or for the purpose of extracting information without 
causing any biological effects such as in diagnosis (Dunn, 1991). Diagnostic applications of ultrasound are 
well established in medical fields such as obstetrics, cardiology, gynaecology, neurology and ophthalmology 
(Dunn. 1991). Medical applications are usually performed with pulsed ultrasound The duration of the 
pulses is typically in the I InS range; intervals between the pulses are in the 1 ms range for diagnostic 
ultrasound and in the 1 s range for disintegration applications. sucb as the disintegration of kidney stones 
(Henglein et al ., 1989). Continuous·wave ultrasound is used for physical therapy (ler Haar, 1988). 
The mechanisms by which ultrasound can induce changes in biological systems are thermal, cavitation and 
mechanical effects such as particle movement (ter Haar, 1988). The energy of an ultrasonic wave moving 
through tissue is attenuated due to either absorption or scatter. The absorption of energy leads to tissue 
heating (Ier Haar, J 988). The selective beating of tissue allows for ultrasound to be used in physical therapy 
for the relief of pain, reduction of soft tissue stiffness and the acceleration of healing (NCRP 74, 1983). 
Heating increases the blood flow to the area thus bringing metabolites needed for tissue repair and removing 
waste products; heating also increases the rate of many biochemical reactions and accelerates the diffusion 
rate across biologic membranes (NCRP 74, 1983). 
The formation of free radicals could promote chemical reactivity in vivo if cavitation occurs during any 
medical application of ultrasound (Carmichael et aI., 1986). Investigation of the chemical effects of pulsed 
ultrasound has shown that a memory effecr exists, the shorter the time interval between pulses the more 
efficient the ultrasonic pulse is at producing a chemical effect. The memory for what has happened before 
fades as the length of interval increases (Henglein, 1987: Henglein et aL 1989). The low time ratio of pulse 
length to pulse interval used in ultrasonic diagnosis (over I to I CXXJ) indicates that the memory effect is 
unlikely to be evident and that the supplied time-averaged. acoustic energy is low (Henglein et al .. 1989). 
The high instantaneous intensity of the pulses. for example 150 W cm·2 for diagnosis and 10000 W cm·2 for 
disintegration, could, however, induce cavitation to occur during a pulse (Henglein et al ., 1989). 
Theoretical studies of the cavitation threshold indicate that cavitation could possibly occur during pulsed 
ultrasound (F1ynn, 1982; Flynn and Cburch, 1984). Pulsed ultrasound has been shown to produce free 
radicals in aqueous solution under conditions similar. although not identical. to diagnostic exposure 
conditions (Cannichael et aI., 1986: Christman et aI., 1987). The experimental conditions used in the 
investigation represented worst case values that would be more sensitive than typical diagnostic conditions 
(Carmichael et al ., 1986; Chrisnnan et al., 1987). The extensive use of ultrasonic diagnosis has required 
that the safety of such techniques and instruments be continually investigated (Crum et al .. 1992). There is 
little supponing evidence that diagnostic ultrasound instruments pose even a measurable risk to a patient 
Chapter 2 ULTRASOUND 2-41 
(Crum et al., 1992; Dunn. 1991). Reports published by the World Health Or-ganisation and the National 
Council on Radiation Protection and Measurements have also indicated that no adverse health effects iD 
human beings exposed to diagnostic ultrasound have been reported (NCRP 74, 1983; EHC 22, 1982). 
2.4.4 Industrial applications 
The applications of ultrasound in industry include cleaning, drying, degassing, defoaming, soldering, 
drilling, plastic welding, extraction, filtration, sterilisation., emulsification., homogenisation, flotation, 
dissolution, deaggregation of powders. biological cell disruption, crystallisation and chemical reaction 
enhancement (Mason, 1991). 
The application of ultrasound for cleaning is a long~l.ished and efficient technology (Mason, 1993). 
Large cleaning systems are commercially available. An ultrasonic cleaner (volume, 7 500 L; power. 
40 kW) has been used by the US Navy since ]968 to facilitate the disassembling of jet engines where heat 
and corrosion have made the unscrewing of fastening bolts difficult (Hunicke, 1990). The uniformity of the 
ultrasonic field, supplied by two banks of immersible transducers, was within 10 % throughout the tank 
(Hurucke, 1990), 
Ultrasound can be applied in environmental remediation for the removal of airborne particles (Grinthal and 
Ondrey, 1992). Ultrasonic agglomeration was used to remove pollutant panicles from air that were less 
than 5 }ll11 in diameter and were too small to be removed by electrostatic precipitators. Airborne soot in a 
IS m3 chamber \\'as reduced by three orders of magnitude with an ultrasonic transducer (frequency, 21 kHz) 
in experiments performed at the Institute of Transuranium Elements in Gennany; experiments were also 
performed in a 170 m) chamber (Grintbal and Ondrey, 1992). 
The recovery of precious metals from ore has been improved with the cleaning of 3 L s' ) water slurries of 
crushed ore in a reverbatory ultrasonic mixer ofpo\\'er 16 kW (Hunicke. 1990). An ultrasonic soldering pot 
(power. 14 kW) operating at 400 °C has been in operation for over IS years with negligible ultrasonic 
equipment downtime (Hunicke, 1990). Ultrasonic defoaming has been used when the addition of chemical 
defoarning agents will contaminate the liquid to be defoamed (Chendke and Fogler, 1975). Ultrasonic 
defoaming has been used at a Coca-Cola bonting plant in Madrid to prevent overflow waste and rejections 
on a 72 000 bottle per hour line (Grintbal and Ondrey. 1992). 
The application of ultrasound in various industries has been reviewed separately in the following sections. 
water and effluent treatment in Section 2.4.4.1, the food industry in Section 2.4.4 .2, the textile industry in 
Section 2.4.4.3, the petroleum industry in Section 2.4.4.4 and membrane processes in Section 2.4.4.5 
Cbapter 2 ULTRASOUND 242 
2.4.4.1 Water and eft1uent treatment 
The combination of the physical effects (production of shear forces and shock waves) and chemical effects 
(radical formation) of cavitation allow for the application of ultrasound in water and effiuent treatment 
(Mason et al., 1993). Ultrasound, however, is more effective when used in combination with other 
conventional treatment processes than as a stand-alone process (Mason et al., 1993; Mason et al., 1994). 
The biocidal activity of chlorine is enhanced by ultrasound (Mason et al., 1994). The number of bacteria 
surviving a 5 min chlorine contact time was reduced by 30 % with 2 miD of ultrasound. The break-up and 
dispersion of bacteria clumps, caused by ultrasound, allows for greater penetration of the biocide and 
renders bacteria more susceptible to chemical attack (Mason et al. , 1994). llltrasound could thus be used to 
reduce the biocide concentration required in water treatment (Mason et al. , 1994). 
The biological effects of ultrasound have been studied since the 1920's, ultrasound has been shown to 
disrupt cells (Alliger, 1975; Frizzell, 1988). Bacteria cells can be destroyed by ultrasound, however, high 
intensities are required for effective disinfection and a 100 % kill (Mason et al., 1994). Cellular disruption 
is due to the physical effects (such as shear forces and shock waves) of ultrasonic cavitation (FrizzeU, 1988; 
Riesz and Kondo, 1992). Cells that are small and round in shape are more resistant to disruption by 
ultrasound (Alliger, 1975). 
Ultrasound has been used to inactivate Cryplosporidium parvum oocysts that cause cryptosporidiosis (severe 
diarrhoea) in humans and domestic animals. (Anonymous. 1995). A 750 L h' \ pilot plant developed by 
Biwater and the University of Leeds has shown that complete destruction of the oocysts is achieved with 
20 s of sonication. 99,8 % destruction is achieved with 15 s of sonication (Anonymous. 1995). 
Ultrasound has also been used to ensure that no bacterial species are present after water treatment (Clasen 
and Sobona, 1994). Rapid sand filters and flocculants are used to remove the bulk of planktonic organisms 
during water treatment, however, species such as NOlho/ca caudata can persist through these processes. 
Bacteria1 regro\\th can occur in water distribution networks if all bacteria] species are not inactivated during 
water treatment. Clasen and Sobotta have shown that 99 % inactivation rates were achieved for the two 
plankton species Artemia salina and Cyclops nauptli within 30 s of sonication at an ultrasonic intensity of 
0.8 W cm·2 and a frequency between 20 and 40 kHz (Clasen and Sobotta, 1994). A 95 % inactivation rate 
was achieved in an investigation of a pilot flow reactor (volume, 80 L) with a water throughPUt of 
8 000 L h- I and a specific output of 12 W L·1 (Clasen and Sobotta. I 994). The 1994 estimation of capital 
spending and operating costs were OM 0,026 per ml, this compared favourably with the savings in 
flocculant chemicals and the reduction in sludge (Clasen and Soboua, 1994). 
Chapter 2 ULTRASO\l1"I'D 2-43 
Ultrasound has been used in combination with technologies such as ozone and electrostatics to improve 
water and effiuenl treatment. The use of ultrasound in combination with ozone enhances emuenl treatmenl 
by producing an emulsion from suspended particles and dispersed microorganisms and by preventing the 
coa1escence of ozone bubbles that ensures maximum bubble surface area is available for oxidation (Chendke 
and Fogler, 1975). An investigation of a pilot plant processing 76 kL of sewage per day indicated that 60 s 
of treatment with ultrasound and ozone destroyed 100 % of the faecal bacteria and viruses (Chendke and 
Fogler, 1975). Dahi found that ultrasonic tteatment (2 min) of effiuenl from a biological sewage plant 
reduced the sterilization dose of ozone by 50 % (Dahi, 1976). Ultrasound. combined with ozone has been 
used to enhance the degradation ofnatura1 organic matter (Dlson and 8arbier, 1994), cyclohexene (Weavers 
and Hoffmann, 1998) and 4-nitrophenol (Barbier and Petrier, 1996). The 4-nitrophenol degradation 
increased with increasing ultrasonic frequency, (20 versus 500 kHz) because of the greater quantity of 
hydroxyl radicals escaping from the cavitation bubbles (Barbier and Petrier, 1996). 
Reimers and co-workers investigated the combination of ultrasound and electrostatics as a method to reduce 
the addition of chemicals during water treatment (Reimers et al. , 1979). Benefits of ultrasound were the 
increased agitation and mixing of water, minerals and organics and the generation of kinetic energy to 
excile atoms and molecules. Application of this method in the soft drink bottling industry was found to 
reduce operation and maintenance problems and to reduce possible noxious effiuents (Reimers et a1.. 1979). 
Industrial effiuent can also be treated with biological metbexls such as the activated sludge process 
(Balasubramaniam et al. , 1995). Ultrasound has been used in conjuction with conventional techniques to 
enhance sludge processing. Bien investigated the use of ultrasound as a non-chemical method for the 
preparation of sludge for dewatering (Bieu. 1988). Sludge is usually dewatered by filtration with the 
addition of polyelectrolytes as flocculants. The initial hydration of 92.8 % of a mineral sludge was reduced 
to 84.2 % after vacuum filtration. 77.7 % with the use of ultrasound.. 69.2 % with the addition of a 
polyelectrol~1e and 61 ,2 % y,ith the oombination of ultrasound and polyelectrolyte addition (Bieu. 1988). 
The required polyelectrolyte ooncentration was halved y,ith the simultaneous application of ultrasound.. A 
similar pattern in final hydration values was also found for an organic sludge. The effect of ultrasound was 
greater on sludges of higher hydration and rigid structure than on amorphous sludges of lower initial 
hydration (Bien. 1988). The investigation by Bien and Wolny of two digested and difficult dewatered 
sludges has also shown the advantageous use of ultrasound.. in oombination with polyelectrolyte addition. on 
sludge dewatering and reduction in sludge volume (Bien and Wolny. 1997). 
Ultrasound was found 10 affect sludge and polyelectrolyte structure. In oontrast to unprepared sludge. the 
addition of a polyelectrolyte produced oompact but separate clusters of sludge floes ; the simultaneous 
application of ultrasound increased the spaces between the conglomerates (Bien and Wolny, 1997). The 
polymer chain shape and crystal outline of Zetag-50, the polyelectrolyte used: in the investigation, was found 
to change during sonication (Bien and Wolny. 1997). It was concluded that the sonication time required to 
Chapter 2 ULTRASOUND 2-44 
enhance sludge dewatering may depend on the type of polyelectrolyte used and should be determined 
experimentally (Bien and Woiny, 1997). 
King and Foster investigated the effect of ultrasound on the settling properties of activated sludge (King and 
Foster, 1990). Ultrasound was found to improve the sctt1eability of the sludge, however, the filterability of 
the solids deteriorated and the supernatant. after settlement, had a high concentration of suspended solids. 
U1trasound degraded the sludge Does and increased the number of small panicles. Floc strength, which can 
vary widely in activated sludge floes, determines the susceptibility of Does to ultrasonic shear forces (King 
and Foster, 1990). U1tra.sound also caused the release of extracellular polymeric material such as soluble 
carbohydrate and protein from the sludge (King and Foster, 1990). 
Low intensities of ultrasound have been used to increase the activity of activated sludge microorganisms 
(Balasubramaniam et al., 1995; Tkachuk et al., 1990). The treatment of sludge with ultrasound with a 
frequency of 2,6 MHz and intensity between 2 and 6 W cm-2 caused the dehydrogenase, proteolytic and 
glycolytic activity to increase by a factor of 1,9; 1, 1 and 4 respectively (Tkachuk et aI. , 1990). The 
increased. biological activity was attributed to the partial loosening of the sludge structure. aUowing greater 
contact between the bacterial cells and the substrate (I1cachuk et al., 1990). The investigation performed by 
Balasubramaniam and co-workers found that ultrasound increased substrate removal and oxygen uptake, 
thus indicating an increase in microorganism activity with ultrasound (Balasubramaniam et aI., 1995). The 
suspended solid content in the effluent increased with ultrasound but decreased with refloculation of the floc 
particles. Ultrasound also improved the settleability of solid panicles (such as Kaolin panicles) in 
agglomeration experiments (Balasubramaniam et al., 1995). 
Ultrasound has also been used to accelerate the anaerobic digestion of sewage sludge (Chiu et at.. 1997; 
Tiehm et at .. 1997). Sewage sludge is stabilised by anaerobic digestion, however, the slow degradation rate 
of the sludge (sludge hydrolysis) requires large digestors and long residence times of approximately 20 d 
(Tiehm et aI., 1997). Tiehm and co-workers used a flow-through ultrasonic reactor (frequency, 31 kHz: 
power, 3,6 kW; treatment time, 64 s) that was specially adapted for sludge treatment (Tiehm et al. , 1997). 
The reduction in particle size and the increase in chemical oxygen demand of the supernatant (organic 
substances were transferred from the sludge solids into the aqueous phase) indicated that ultrasound caused 
the deagglomcration of aggregates and microbia1 cell disruption. Biogas production was greater during 
batch fermentation experiments (ISO L fermenter vessels, 22 d residence time) in the fermentation vessel 
that had been fed with the ultrasonic-disintegrated sludge. This indicated that the organic compounds 
transferred from the sludge solids into the aqueous phase were readil)' biodegradable (Tiehm et aI., 1997). 
The degree of sludge degradation in semi-<X)ntinuous fermentation e"':periments was calculated with respect 
to the reduction of volatile solids. In fermentation vessels with a 22 d residence time. the reduction of 
volatile solids \1,'35 45,8 % for untreated sludge and 50,3 % for ultrasonic-disintegrated sludge. The 
reduction of volatile solids in a fermentation vessel with a 12 d residence time was 47,3 %. It was 
Chapter 2 ULTRASOUND 2-45 
concluded that the ulU'3S0ruc pretreatment of sewage sludge was a potential method to increase the rate of 
fermentation and reduce the volume of sludge digestors (fiehm et aI., 1997). 
Chiu and co-workers investigated the combination of alkaline (NaOH dosage) and ultrasonic pretreatment 
of sludge before anaerobic digestion (Chiu et al., 1997). Experiments were performed in 1 L fermentation 
vessels; an ultrasonic horn rated at 120 W with a frequency of 20 kHz was used for ultrasonic pretreatment 
of the sludge. The soluble chemical oxygen demand, soluble organic nitrogen release and production of 
total volatile fatty acids were monitored for alkaline treatment, alkaline treatment followed by ultrasound 
and thc simultaneous application of alkaline treatment and uluasound (Chiu et aI., 1997). The highest 
initial rate of sludge hydrolysis (211 ,9 mg L'1 min·1) and production of total volatile fatty acids (84 %) were 
recorded for the simultaneous application of ultrasound and alkaline treatment. It was concluded that 
alkaline and ultrasonic pretreatment of sludge could significantly reduce the treatment time of anaerobic 
digestion (Chiu et al., 1997). Ch1orobeozene is not usually affected by conventional biological treatment 
due to the presence of other more readily metabolised organic compounds~ ultrasonic pretreatment resulted 
in a significant reduction in chemical oxygen demand during biodegradation with activated sludge under 
aerobic conditions (Neis, 2000). 
The removal or degradation of chemical pollutants, such as herbicides and pesticides of agriculruraJ origin 
or hydrocarbons from industrial effiuent and storm water run-off, is required during water treatment prior to 
the water being reused or discharged into the environment (Mason, 1993; Mason et al., 1994). Ultrasound 
has been used to enhance the degradation of various pollutants. Agricultural herbicides such as atrazine 
and alach10r degrade during sonication according to first~rder kinetics with rate constants of 0,0021 and 
0.0080 min'\ respectively, degradation mechanisms were not investigated (Koskinen et at.. 1994). 
Sonication of water saturated with the pesticide parathion indicated that parathion was totally degraded 
within 2 h of sonication at 30 °C with an ultrasonic horn operating at a frequency of 20 kHz and an 
intensity of 75 W cm·2 (Kolronarou et al. , 1992a). The primary degradation mechanism of parathion was 
thermal decomposition in the hot interfacial region of the collapsing cavities, reaction with hydroA')'1 
radicals acted as a secondary mechanism. Degradation products such as p -nitrophenol and phosphate, 
sulfate. nitrite, nitrate and oxalate ions were identified (Kotronarou et aI., 1992a). 
Monochlorophenol compounds, water pollutants that exhibit moderate toxicity to mammalian and aquatic 
life. have been degraded with ultrasound (Gondrexon et aI., 1993; Lin et aI., 19%; Serpone et aL , 1994). 
These compounds are formed during the degradation of chlorinated herbicides. the ch1orination of phenolic 
compounds in industrial effluent and chlorine treatment of drinking water (Serpone et aL 1994). The 
application of ultrasound degraded 2-cblorophenol and 3-chlorophenol (initial concentration, 80 !lmol L· I ) 
into dechlorinated intermediate products 'within 10 h; 4-chlorophenol degradation occurred within 15 h. 
Degradation followed first order kinetics. the disappearance rate constants were 0,0048; 0.0044 and 
Cbapter 2 ULTRASOUND 2-46 
0,0033 min'\ respectively (Serpene et al., 1994), The ultrasonic degradation products \\-"ere similar to those 
fonned during ultraviolet radiation with TiOz semiconductor particles. Sonication was perfonned with an 
ultrasonic horn (power, 50 W; frequency, 20 kHz) in a glass sonication cell (Serpone et aL, 1994). Tbe 
degradation of 2-<:hloropbenol was greater with the addition of hydrogen peroxide (200 mg L'I) during 
sonication than without the hydrogen peroxide (Lin et al,. 1996). Lin and co-workers investigated how 
parameters such as pH, ionic strength. the presence of a catalyst and initial concentration affected the 
ultrasonic degradation of 2<hlorophenoJ in the presence of hydrogen peroxide (Lin et al" 1996). The 
degradation was found to follow pseudo first order reaction kinetics (Lin et al. , 1996). 
Ultrasound has been used to degrade chlorinated hydrocarbons that are typically the most common 
pollutants found at hazardous waste sites (Cheung et al., 1991; Cyr et al" 1998; Orzechowska et al., 1995). 
The use of these compounds in products such as solvents, dyes, pesticides and ink.. have led to the 
contamination of water resources (Bhatnagar and Cheung. 1994). Cheung and co-workers found that the 
ultrasonic degradation of methylene chloride, carbon tetrachloride, 1, I ,1-trichloroethane and 
trichloroethylene (for a concentration range of 100 to 1 000 mg L-I) followed first order reaction Idnetics 
(Cheung et al. , 1991). The decrease in solution pH during sonication indicated that hydrochloric acid was 
produced, no other chlorinated products were identified (Cheung et al., 1991). The investigation by 
Bhatnagar and Cheung also recorded a decrease in solution pH, first order reaction kinetics and the 
formation of hydrochloric acid during the ultrasonic degradation of methylene chloride, 1,2-dichloroethane, 
carbon tetrachloride, chlorofonn, 1,1, I-trichloroethane, trichloroethylene and perthloroethylene (Bhatnagar 
and Cheung, 1994). Ultrasonic degradation varied between 72 and 99,9 % for solutions of initial 
concentration bet\\'een 50 and 350 mg L-1 for a 40 min sonication period; the rate of degradation was not 
affected if mixtures of compounds were sonicated (Bhatnagar and Cheung, 1994). 
C)T and co-workers found that a 20 mg L'] trichloroethylene solution degraded more than 90 % wilhin a 
10 min sonication period (Cyr et al., 1998). Sonication was perfonned with an ultrasonic horn (power. 
120 W: frequency, 20 kHz) in a 50 mL reaction vessel. Parameters such as initial pH value of solution 
(between 3 and 11), bicarbonate concentration (below 10 nunol L· l ) , presence of a metal or metal oxide 
(0.0 1 to lOO g L-1 concentration) and the combination of a metal or metal oxide (0.01 to 0.1 g L·1 
concentration) with 34 mg L-1 hydrogen peroxide did not affect the ultrasonic degradation of 
trichloroethylene (Cyr et al.. 1998), Orzechowska and co-workers monitored chloride fonnation (Cr) to 
show the degradation of chlorofonn, cartx:m tetrachloride and trichloroethylene: chloride fonnation could be 
detected within I min of sonication of solutions with initial concentration between 3 and 80 mg L-1 
(Orzechowska et al., 1995), 
Cbapter 2 ULTRASOUND 2-47 
A range of other chlorinated compounds have been degraded with ultrasound Cheung and Kurup 
investigated the sonochemical destruction of chJorofluorocarbon compounds (Cheung and Kurup, 1994). 
The degradation of fluorotrichloromethane and trifluorotricbloroethane followed first order reaction 
kinetics; 50 mg L'] solutions were degraded within a 10 min sonication period. The decrease in solution pH 
during sonication was due 10 the formation of HCI and HF (Cbeung and Kurup, 1994). The rate of 
ultrasonic degradation of chloral hydrate was found to depend on parameters sucb as initial concentration. 
oxygen concentration and the ultrasound frequency and intensity (Sakai et al ., 1977). Degradation occurred 
through a radjcal mechanism producing hydrochloric acid (Sakai et al., 1977). Pettier and co-workers 
investigated the effect of gases such as oxygen, air and argon on the ultrasonic degradation of 
pentachlorophenate (Pettier et al., 1992). The reduction in toxicity of the solution during sonication 
(quickest for argon saturation) inwcaled that complete mineralization of pentachlorophenate occurred 
(Petrier et al. , 1992). The ultrasonic degradation of chlorobenzene and 1,4-<licborobenzene followed first 
order reaction kinetics; the rate of reaction depended linearly on the ultrasonic intensity (Price et al.. 1994). 
Complete degradation of 1,4-<lichJorobenzene occurred within 40 to 50 min with sonication at an intensity 
of 39 W cm·2; the first step in the degradation was the removal of a chlorine atom from the aromatic ring 
and reduction to chloride (Price et al., 1994). Total degradation of polychlorinated biphenyl compounds in 
water treatment is not essential since the removal of chloride atoms reduces toxicity and accelerates 
biodegradation (Price et al., 1994). The investigation of effiuent contaminated with benzene and toluene 
recommended that ultrasound be used as a pretreatment stage for relatively high target concentrations rather 
than as a polishing stage for meeting effluent guidelines (Iboma et al., 1998). 
Ultrasound has been used to degrade a range of phenolic compounds. The rate of phenol degradation at 
different ultrasonic frequencies was directly related to the rate of hydrogen peroxide formation during 
sonication (Petrier and Francony, 1997: Petrier et al.. 1994). The first step in the degradation mechanism 
of phenol was reaction with hydroxyl radicals; the primary intermediate products were hydroquinone. 
catechol and benzoquinone (Petrier and Francony, 1997; Petrier et al.. 1994). The zero order reaction rate 
increased with increasing phenol concentration reaching a limit at a concentration of 188 mg L·1 for a 
frequency of 20 kHz and 470 mg L'] at 487 kHz (Pettier et a1.. 1994). The ultrasonic degradation of 
p-nitrophenol occurred primarily through thermal decomposition at the cavity interface and then due to 
reaction with hydroxyl radicals (Kotronarou et al. , 1991). Degradation followed first order reaction kinetics 
producing nitrate and nitrite ions, benzoquinone. hydroquinone, 4·nitrocatechol. formate and oxalate 
(Kotronarou et al.. 1991). The rate of degradation of p-nitrophenol in solutions containing particulate 
maner. phosphate, bicari:xmate and humic acid (to simulate natural waters) was the same as in pure water 
(Cost et al ., 1993). Chemical components in natural waters that are radical scavengers thus do not affect 
the degradation of pollutants (such as p--nitrophenol) that degrade predominately due to high temperature 
reactions (Cost ct al., 1993). Inhibition, however, occurs in waters containing high concentrations of 
bicarbonate and phosphate compounds (Cost et al .• 1993). 
Chapter 2 ULTRASOUND 2-48 
Monocyclic aromatic compounds (such as benzene, ethylbenzene, styrene and o-<:blorotoluene) and 
polyaromatic compounds (such as napthalene, anthracene and pyrene) have been degraded with ultrasound 
(de Visscher et al., 1996; Price et al., 1994). Degradation followed first order reaction kinetics: reaction 
products were not identified (de Visscher et al., 1996; Price et al., 1994). Hydrocarbon compounds such as 
methane, ethane, propane, butane, ethylene, acetylene, methanol. ethanol. n-propanol and isopropanol have 
also been degraded with ultrasound (Biittner et al., 1991 ; Han et al., 1990a; Han et al ., 1mb; Koike, 
1992). 
Ultrasound has also been used to degrade inorganic compounds. Kotranarou and co-workers investigated 
the u1trasonic degradation of hydrogen sulfide dissolved in the effiuent from fossil fuel refining processes 
(Kotronarou et al., 1992b). Degradation followed zero order reaction kinetics at a pH greater than 10 and 
was due to reaction with hydro).)'1 radicals; degradation at a pH less than 8,5 was due to tbennaJ 
decomposition and followed first order reaction kinetics (Kotronarou et al., 1992b). De Visscher and 
co-workers concluded that mechanisms for ultrasonic degradation that are primarily due to reaction with 
hydroxyl radicals follow zero order reaction kinetics whereas degradation due to thermal decomposition 
(pyrolysis reactions) follow first order reaction kinetics (de Visscher et al ., 1996). 
Potential applications of ultrasound in water and effluent treatment have been summarised by Mason and 
co-workers and are listed in Table 2.3 (Mason et al., 1993). 
Table 2.3: Applications of ultrasound in water and effluent treatment (Moon et al •• 1993) 
Effect of ultrasound 
degassing 
gasification 
dispersion 
biocidal action 
filtration 
destruction of chemicals 
2.4.4.2 Textile indust~· 
Application 
removal of excess chlorine 
enhance oxygenation and ozonolysis 
homogenise mixtures before treatment and dispersal of chemicals 
in treatment 
improve efficiency of biocide 
improved dewatering 
remo\'3lofpollutants 
Large quantities of water and energy are required in the wet processing of textiles (Thakore. 1 99Oa). 
Conventional ctyeing techniques could be enhanced by ultrasound because of the u1trasonic-induced 
dispersion and break-up of dye aggregates, the expulsion of entrapped air from fibre capillaries and the 
acceleration of the rate of diffusion of dye inside the fibre (Ahmad and Lomas. 1996). The effects of 
ultrasound are due to cavitation and heating, the temperature in the dye bath increases when ultrasonic 
waves pass through the liquid (Thakore. 199Oa: lbakore, 1990b). 
Cbapter 2 ULTRASOUND 2-49 
Shimizu and co-workers investigated the effect of ultrasound on the dyeing of nylon 6 films by dyeing 
amorphous., unoriented films with four dye typeS, disperse, reactive, acid and acid mordant at temperatures 
of 20, 40 and 60 DC (Shimizu et al ., 1989). The uptake of the four dyes in nylon 6 increased (at all 
temperatures) with the application of ultrasound (Shimizu et al., 1989). 
Shukla and Mathur investigated the low-temperature ultrasonic dyeing of silk (Shukla and Matbur, 1995). 
Silk was dyed with cationic, acid and metal-GOmplex dyes at temperatures of 45 and 50 DC. The uptake of 
cationic dyes, in comparison with conventional dyeing for 60 min at 85 DC, increased by between 71 and 
104 % when dyeing for 15 min at 45DC with ultrasound Under similar ultrasonic conditions, dye uptake 
for acid and metal-complex dyes increased by 108 and 76 % respectively (Shukla and Mathur, 1995). Fibre 
degradation was not apparent and the level of wash fastness, for all the dyeings, was equivalent to that for 
samples dyed according to the conventional procedure (Shulda and Mathur, 1995). 
Ah.mad and Lomas investigated the low-temperature ultrasonic dyeing of polyester fabric with disperse dyes 
for possible application in batik printing (Ahmad and Lomas, 1996). Elevated dyeing temperatures would 
cause the wax used in batik printing to melt. The dye uptake was similar for both the conventional and 
ultrasonic dyeing processes at 50 DC and was lower than the dye uptake for the conventional dyeing process 
at 100 DC (Ahmad and Lomas, 1996). Shuklar and Mathur also found that the dye uptake in polyester 
fibres, even after pre-swel1ing, could not be enhanced by ultrasound (Shukla and Mathur, 1995). Polyester 
fibres are highly compact whereas fibres, such as silk, that have a less crystalline structure are more 
accessible for dyes and chemicals (ShukJa and Mathur, 1995). 
Oner and co-workers investigated the use of ultrasound in the dyeing of cotton fabrics with reactive dyes 
(Oner et al.. 1995). lfllrasound was found to improve dye fixation and increase the colour yield without 
affecting the fastness properties (wash fastness and staining) of the dyed fabric (Oner et al., 1995). It was 
concluded that the shorter dyeing times and lower quantities of dye required to produce a panicular colour. 
and the reduced \o ... ater required to remove unfixed dye. would result in both energy and water savings with 
the use of ultrasound in reactive dyeing of cotton fabrics (Oner et al .. 1995). 
2.4.4.3 Food induSl~' 
Ultrasound is applied in both the analysis and processing of food (Floros and Liang, 1994). Ultrasound can 
be used for treating heat-sensitive materials such as in the extraction of apple juice from apple puJp and the 
decolouration of soyabean oil without causing a loss in flavour or other heat-related damage (Chendke and 
Fogler, 1975). Lillard investigated the application of wtrasound for the microbial decontamination of 
poultry skin (LiUard I 994). A combination of ultrasound and chlorine, for application in the final 
processing stage of broilers. was more effective in reducing the Salmonella count in poultry skin than either 
ultrasound or chlorine alone (LilIard. 1994). lfltrasound can affect enzyme activity (Robens. 1993). 
Chapter 2 ULTRASOUND 2-50 
Ultrasound usually inactivates oxidase enzymes but only affects low concentrations of catalase enzymes. 
Reductase and amylase enzymes are resistant to sonication (Robens, 1993). Ultrasound has been used to 
improve the rate of the hydrogenation reaction for converting unsaturated fatty acids (such as yegetable oil) 
to saturated fatty acids that are used as shortening agents in the baking industry (Robens, 1993). 
Ultrasound has been used to induce nucleation in the crystallisation of aqueous salt solutions and sugar 
solutions (Roberts, 1993). Dehydration rates of products, such as gelatine, yeast cake and orange powder 
have been enhanced with the application of ultrasound (Floros and Liang, 1994). The use of ultrasound to 
measure anima1 back fat thickness is the major application of low·intensity ultrasound in the food industry 
(Povey, 1989), 
2.4.4.4 Petroleum industry 
An ultrasonic vibrating tray (power, 2 kW) has been used to treat ore slurries with a flow rate of 6 kg 5'\ 
(Hunicke. 1990). Junk coal from waste-<:Oal ponds, after screening and mixing with water to produce a 
slurry, can be cleaned on the vibrating tJay to yield a coal with a low ash., sulphur and residual water 
content (Hunicke, 1990). The removal of sulphur from coal in which the sulphur is predominantly in the 
pyrite form improves the susceptibility of coal to spontaneous combustion (Zaicti, 1993). The sulphur 
content in a high sulphur Pakistani lignitic coal was reduced by 34 % with sonication at 30 QC in a 
0,975 g L'\ sodium hydroxide solution (Zaidi, 1993). The ash content, during ultrasonic treatment, 
increased between 2 and 18 % for different coals; no enhancement of sulphur removal was achieved for 
sonication of coal in water (Zaidi, 1993). 
Sadeghi and oo-workers have developed an ultrasonic method to refine fossil fuel-containing materials 
(Sadeghi et al.. 1994). Materials such as tar sand, asphalt, heavy oil, coal liquids and oil shales were 
sonicated in an alkaline solution at ambient temperature and pressure to recover upgraded lighter fuels. 
The recovered fuel was lower in heterocyclic elements. For the treatment of a coal liquid sample, the coal 
asphaltene content was reduced to 58 % within 30 min of sonication while the oil and resin content 
increased to 42 % (Sadeghi et al., 1994). 
The characterisation of coal and coal tar pitches. performed by studying fractions ex1racted with organic 
solvents. is hindered by the insolubility of coal (GuiDen et al.. 1991 ; Matturro et al .. 1990). Guillen and 
co·workers studied 27 organic solvents to determine which was the most effective for the ultrasonic 
extraction of coal tar pitches (Guillen et al., 1991). Solvent ex1:raction was performed at 25 QC for 2 h in an 
ultrasonic bath. Yields were more reproducible for solvents with a high density and low viscosity. Solvents 
such as carbon disulphide. pyridine. 1·methyl-2·pyrrolidinone. tetrahydrofuran. quinoline and nitrobenzene 
were the most effective in the extraction of six coal tar pitches (Guillen et al.. 1991). MatturTo and 
oo-workers extracted coa.1 \\ith tetra-n.ootylammonium hydroxide during sonication in a solvent mixture 
containing pyridine and methanol (Matturro et al ., 1990). The sonication of an Illinois No. 6 coal for 3 hat 
Chapter 2 ULTRASOUND 2-51 
34 °C with an ultrasonic hom (frequen~', 23 kHz; intensity, 85 W cm-l ) produced an extraction yield of 
77 0/0. the yield for the stirred control was 27 % (Matturro et al., 1990). 
tfltrasound can also be used to detennine the concentration of coal extracts (Knesinska and Pajak, 1988). 
The ultrasonic velocity in a letrahydrofuran solution was found to be linearly dependent upon the 
concentration of coal in solution. Such a system would first have to be calibrated by measuring the 
ultrasonic velocity in solutions of known concentration for the specific coal to be monitored (Knesinska and 
Pajak, 1988). 
2.4.4.5 Membrane processes 
Lenart and Ausl~der investigated the effect of ultrasound on the di.ffiJsion of chloride oompounds (sodium. 
potaSSium and ca1cium chloride) through cellophane membranes (Lenart and AusUinder, 1980). tfltrasound 
enhanced diffusion but the increase in the amount diffused was detennined by the direction of diffusion 
flow. The greatest increase occurred. when the ultrasonic field and diffusion flow where in the same 
direction. A lower increase occurred when the ultrasonic field was acting at 90° to the diffusion flow and 
the smallest increase occurred when the ultrasonic field and diffusion flow where acting in opposite 
directions (Lenan and Ausl~der, 1980). The ultrasonic-enhanced diffusion was attributed to the acoustic 
microcurrents that increased the velocity of the particles in solution (Lenart and Ausliinder. 1980). 
Howlcins investigated the application of ultrasound to enhance dialysis in artificial kidney machines 
(Howkins. 1969). The enhanced rate of dialysis was attributed to the efficient stirring of fluid layers near 
the membrane surface (Howkins, 1969). tfltrasound. also reduces the diffusion boundary layer and may 
increase the diffusion coefficient (Floros and Liang. 1994). 
Wakeman and Tarleton investigated the effects of ultrasound on crossflow microfiltration (Wakeman and 
Tarleton, 1991). Crossflow filtration is used for slurry de\>.'atering. Fouling of the membrane occurs if the 
shear rates generated by the crossflow velocities are not sufficiently high to prevent cake formation 
(Wakeman and Tarleton. 1991). Ultrasound reduced membrane fouling and increased the filtrate flu .. ~ . 
Membrane fouling was further reduced and filtrate flux increased when ultrasound was used in conjunction 
with an electric field It was ooncluded (hat lower crossflow velocities could be used when filtration is 
assisted by electric and ultrasonic fields (Wakeman and Tarleton. 1991 ). 
2.5 SONOCHEMICAL EQUIPMENT 
Sonochemical experiments are performed in equipment such as ultrasonic baths, horns and reactors. 
Equipment suitable for routine experiments only became available in the 1950·5 (Bremner, 1986). 
Early ultrasonic research indicated that eX"PCrimentai conditions had to be carefully oontrolled to ensure 
reproducible results. Weissler and co-workers found that the tightness of clamping of a tesHube above an 
Cbapter 2 ULTRASOUND 2-52 
ultrasonic transducer affected. the reaction yield of potassium iodide and carbon tetrachloride in the test-tube 
by 10 % (Weissler et al., 1950). Jennings and Townsend found that the deviation in reaction yield of 
carbon tetrachloride and chloroform in aqueous solution in an inert atmosphere was as large as 30 % 
(Jennings and Townsend, 1961). The large deviation in results was due to the ultrasonic equipment rather 
than the analysis procedure. Three possible sources of error were suggested: the ultrasonic intensity in the 
reaction solution would change unless the coupling of the transducer to the reaction vessel was identical for 
all experiments, the electrical output of the high frequency generator was not constant and thus caused 
variations in the ultrasonic intensity; and the amount of gas dissolved in the liquid at the start of reaction 
could be different for different experiments (Jennings and Townsend, 196 1). 
Recent literature also indicates the importance of careful control of experimental conditions in ultrasonic 
experiments. For example, a reaction vessel in an ultrasonic bath has to be placed in the exact position in 
the bath to ensure consistent and reproducible results (Goodwin. 1990). The range of parameters that have 
to be controUed in ultrasonic experiments lead to the results being specific to the particular piece of 
equipment and experimental setup (Mason. 1991). 
Transducers used for ultrasonic equipment are described in Section 2.5. L Ultrasonic equipment is reviewed 
in the following sections, ultrasonic baths in Section 2.5.2, ultrasonic horns in Section 2.5.3 and ultrasonic 
reactors in Section 2.5.4 . Reactor modelling and equipment scale-up recommendations are reviewed in 
Section 2.5.5. 
2.5.1 Transducers 
Piezoelectric and magnetostrictive transducers are used in ultrasonic equipment; piezoelectric transducers 
are reviewed in Section 2.5. 1. 1 and magnetostrictive transducers in Section 2.5.1.2. 
2.5.1 .1 Piezoelectric transducers 
Materials used as piezoelectric transducers undergo a change in dimension when an electrical potential is 
applied across the surface (Payne, 1994). Examples of such materials are quartz and ceramics such as 
barium titanate and lead zirconate titanate (a solid solution of lead zirconate and lead titanate). Lead 
zirconate titanate is the most common1y used material in piezoelectric transducers (Payne, 1994). The 
frequency at which a transducer operates is determined by the width of the piezoelectric material, typically, 
two or four layers of the material are used in a transducer. The piezoelectic material is prepolarised and 
bolted between two metal blocks for protection, prevention of overheating and to improve the coupling of 
the ultrasonic vibrations to the liquid (Crawford, 1963; Mason. 1991 : Payne. 1994). 
Piezoelectric transducers are relatively small and inexpensive. Damage can be caused by impact. operation 
at temperatures above 150 to 200 °C or if operated without liquid (Hunicke. 1990). Generally, piezoelectric 
transducers are not suitable for industrial equipment if continuous usage at high temperatures is required 
Cbapter 2 ULTRASOUND 2-53 
since the polar orientation of the ceramic materia] changes graduail}' over time (Mason, 1991). These 
transducers have a relatively high gain but operate efficiently over a narrow band width. The narrow band 
width makes it difficult to operate a number of transducers in synchronism as required in large i.ndustrial 
applications (Hunicke. 1990). 
Piezoelectric transducers have a potential efficiency of 98 % and can handle power transfers in the region of 
500 10 I 000 W when used In continuous operation, tbe maximum amplitude of vibration at the radiating 
face of a transducer is between 15 and 20 JllTl for a frequency of 20 kHz (Perkins. 1990). Most ultrasonic 
baths and horns contain piezoelectric transducers (Mason. 1991). 
2.5.1.2 Magnetostrictive transducers 
A magnetostrictive transducer is based on the changes in dimensi.on induced. in a ferromagnetic materia] b}' 
an alternating magnetic fi eld, such a transducer consists of a ferromagnetic rod or bar within a solenoid 
(Crawford, 1963; Mason, 1991). Nickel or a nickel alloy rod are used most commonly in magnetostrictive 
transducers. Such transducers are bigger and more expensive than piezoelectric transducers; operation over 
a larger temperature range (easily withstanding temperatures up to 180 QC) is also possible (Hunicke. 1990). 
A magnetostrictive transducer is resistant to mechanical damage because the nickel core is silver~razed to 
a stainless steel plate to couple the ultrasonic vibrations to the liquid (Mason, 1991). No damage is caused 
to the transducer if operated without liquid (Hunicke, 1990). Since magnetostrictive transducers have no 
mode for degrading or failing, they are not only robust but are also suited for long·term continuous 
operation. Various large commercial systems have been recorded as operating successfully for over twenty 
years (Hunicke, 1990). Magnetostrictive transducers have a relatively low gain which can be an advantage 
in commercial applications requiring more than one transducer since all transducers will operate close to 
peak perfonnance (Hunicke, 1990). 
2.5.2 Ultrasonic baths 
An ultrasonic bath or cleaner is the most readily available and inexpensive source of ultrasound for 
sonochemical experiments. An ultrasonic bath consists of a stainless steel tank of rectangular cross-section 
with transducers bonded externally to the base. The transducers arc usually piezoelectric transducers and 
operate with an intensity of betv.'een 1 and 5 W cm·1 and within a frequency range of between 20 and 
40 kHz (Goodwin, 1990). intensit)' is defined as the ratio of the electrical power into a transducer to the 
radiating surface area of the transducer (Perkins. 1990). The intensity and frequency of an ultrasonic bath 
depends on the t)'pe and number of transducers the bath contains: operating power also varies bctv.-een 
ultrasonic baths produced by different manufacturers (Goodwin, 1990). Experimental results are thus 
characteristic to a particular model of ultrasonic bath. The volume of ultrasonic baths vary between 1.5 to 
approximately 50 000 L (Mason, 1991). Ultrasound is transmitted through a liquid medium, usually water, 
to a reaction vessel containing the reaction solution, as shown in Figure 2.9. 
Chapter 2 ULTRASOUND 
waler -----i~ 
reaction mi'dw'e 
in conical flask 
/ _______ bath 
transducers 
2-54 
Figure 2.9: Schematic diagram. of an ultrasonic batb (Goodwin, 1990) 
Ultrasound is attenuated at the water/glass interface of a reaction vessel. The decreased intensity within a 
reaction vessel may be below the cavitation threshold and cavitation will not occur (Goodwin, 1990). The 
amount of energy available in a reaction solution is also dependent on the size and shape of the reaction 
vessel (Mason and Serlan, 1992). Mason and Serlan defined the sonochemical yield as the ratio of a 
measured effect to the u1trasonic power entering the reaction vessel (Mason and Serlan, 1992). The 
sonocbemical yield was calculated using iodine liberation as a model reaction. The amount of iodine 
liberated from a 4 % potassium iodide solution (containing 10 % carbon tetrachloride) was measured for a 
5 min sonication period; the u1trasorric power entering the reaction vessel was determined by calorimetry. 
The sonochemical yields for reaction solutions (volume, 55 and 110 mL) in a 250 mL conical flask, a 
100 mL conical flask and a 100 mL round bottomed flask. are listed in Table 2.4 . 
Table 2.4: Effect of reaction solution volume and size and shape of reaction vessel on sonochemical 
yield (Mason and 8ertan, 1992) 
Reaction "essel 
lOO mL round·bottomed fl.ask 
100 mL conical flask 
250 mL conical flask. 
Sonochemical yield 
(mg of h per W of ultrasonic power) 
SS mL reaction solution 
0,01 
0.02 
0,10 
t 10 mL reaction solutioD 
0.00 
0,Q2 
0.0 1 
As shown in Table 2.4, the sonochemical yield was the highest for a 55 mL reaction solution in a 250 mL 
conical flask. The larger flat base area of the 250 mL conical flask allowed for a greater transfer of energy 
(Mason and Berlan. 1992). Pugin also measured a greater u1trasonic intensity in a reaction vessel with a 
flat base. a 100 mL conical flask. than a round base, a 100 mL round--bottomed flask (Pugin. 1987). Less 
energy is lost through attenuation at a flat water/glass interface than at a rounded interface. 
Cbapter 2 ULTRASOUND 2-55 
A non-unifonn ultrasonic field is produced within the water in an ultrasonic bath. ultrasonic intensity varies 
with distance from a transducer (Goodwin, 1990). Maximum intensity occurs vertically at levels equal to 
multiples of the half-wavelength of sound in water (Niemczewski, 1980; Pugin. 1987); and horizontally 
directly above a transducer if an ultrasonic bath is fitted with a single transducer or midway between two 
transducers (Pugin. 1987). A reaction vessel has thus to be placed in the identical position in an ultrasonic 
bath for consistent and reproducible results to be oblained (Goodwin. 1990). 
The ultrasonic power available for reaction enhancement is lower than the electrical power consumed by an 
ultrasonic device because energy is lost as heal or sound in the conversion of electrical to ultrasonic energy 
(Martin, 1992). The steps in the conversion of energy from electrical power 10 sonochemical benefit are 
presented in Figure 2.10. 
electrical power supply 
ultrasonic generator 
transducers 
coupling structure 
ultrasonicfiekl 
cavitation field 
contacting pattern 
sonocbenncalbenefit 
Figure 2.10: Ene~' con,,'ersion steps in industrial sonocbemistry (Martin, 1992) 
The temperature of the water in an ultrasonic bath increases during sonication (Niemczewski 1980). A 
calorimetric method can thus be used to measure the power of an ultrasonic bath: ca.lorimetl)' is based on 
the transformation of an ultrasonic wave absorbed in water into therma1 energy (Goswami et al. , 1988). 
The solution temperature in a reaction vessel in an ultrasonic bath is slightly higher than that of the 
ultrasonic bath water; temperature measurements for experiments monitoring the effect of temperature on a 
reaction must thus be taken within the reaction vessel (Goodwin. 1990). 
Cbapter 1 ULTRASOUND 2·56 
2.5.3 Ultrasonic horns 
Disadvantages of an ultrasonic bath, such as the low operating intensity and lack of control of energy input 
into a chemicaJ reaction, are overcome with the use of an ultrasonic horn (Mason, 1991). The efficiency of 
energy transfer is increased with an ultrasonic horn because the horn is inserted directly into the reaction 
solution. Higher intensities (above 100 W cm-I) are produced by ultrasonic horns. Reproducible operating 
conditions are obtained since the power input of an ultrosonic born is controllable. Ultrasonic horns are 
more expensive than ultrasonic baths (Goodwin, 1990; Mason, 1991). 
An ultrasonic horn system (shown in Figure 2.11) consists of a generator, transducer, upper fixed horn 
element. detachable horn and rep1aceabl.e tip (Mason, 1991). The generator supplies the transducer (a 
piezoelectric transducer) with alternating electrical frequency (approximately 20 kHz). The low vibratory 
motion of the transducer (between 1 and to ~) is magnified by the upper fixed born element and the 
detachable horn such that the hom tip vibrates with a peak·to-peak displacement of between 10 and SO ~ 
(Gooctwin, 1990). The upper fixed horn element is attached. to the transducer and is of such a length that 
the end face vibrates with maximum. amplitude. The detachable born, connected to the upper fixed bom, 
transfers the ultrasonic energy directly into the reaction vessel. Some ultrasonic horn systems include a 
replaceable horn tip so that the detachable horn 00es not have to be replaced when the tip of the horn 
becomes eroded (Goodmn, 1990). 
upper fixed 
horn element 
replaceable 
born tip 
transd""", 
housing 
--
 uhrasonic 
born 
reaction 
mixtun: 
Figure 2.11: Schematic diagram of an ultrasonic born system (Goodwin , 1990) 
The magnification of the vibrator)' motion of a transducer to a reaction solution is dependent on the length 
and shape of a horn. For maximum magnification, the length of a horn should be a half wavelength (or a 
multiple thereof) of the sound in the material from which the horn is manufactured (Mason. 1991). The 
Cbapter 2 ULTRASOUND 2-51 
following shapes (shown in Figure 2.12) are the most common designs of ultrasonic homs (Mason,. 1991 ; 
Perlrins. 1990). 
• 
• 
linear exponential stepped 
Figure 2.12: Ultrasonic born sbapes (PeOODS. 1990) 
The uniform cylinder acts as an extension of the working end of the transducer and does not change 
the amplitude of vibration. 
The linear taper or cone is simple to manufacture but potential magnification (the ratio of the upper 
to lower diameter) is limited to approximately 4-fold 
• The exponential taper produces a higher magnification than the linear taper because of the smaller 
tip diameter (magnification is also calculated from the ratio of the upper to lower diameter). This 
design is more difficult to manufacture. The narrow length and small tip area makes this type of 
horn suitable for the sonication of small samples. 
The amplitude magnification of a stepped horn is calculated from the square of the ratio of the 
upper to lower diameter and can be as high as 16-fold 
U1trasonic hams are manufactured from materials that have a low acoustic loss, a high dynamic fatigue 
strength. are resistant to cavitation erosion and are chemically inert (Perkins, 1990). Titanium alloy is the 
most suitable material; other materials such as aluminium. aluminium bronze and stainless steel can also be 
used but the amplitude of vibration is dampened (Perkins. 1990). 
Disadvantages of an ultrasonic horn system are that the erosion of the horn tip may contaminate the 
reaction solution and that a small zone of intense ultrasonic activity is created at the tip of the horn 
(Goodwin. 1990). Pugjn used the chemical reaction of l-bromopentane with lithium wire in 
tetrahydrofuran solution to show the reduction of ultrasonic intensity with increasing distance from the tip 
of an ultrasonic horn (Pugin. 1987). It was found that the rate of decrease in l-bromopentane concentration 
Chapter 2 ULTRASOUJIro'D 2-58 
at 50 % conversion for a distance of 5 mm between the tip of the horn and the lithium wire was 2,70 % per 
min and for a distance of 15 mm was 0,85 % per min. It was concluded that the zone of high ultrasonic 
activity was relatively small and that only marginal ultrasonic activity occurred in the surrounding liquid 
(Pugin, 1987). The fonnation of a small zone of intense ultrasonic activity at the tip of a horn indicates that 
an ultrasonic horn is most suited to a flow system,. as shown in Figure 2.13, where the reaction solution is 
pumped through a flow ceU fined to the horn (Goodwin. 1990). 
I 
solution 
inlet 
• 
I ultr 
/"" 
asonic 
om h 
solution 
outlet 
• 
flow cell 
,/' 
Figure 2.13: Schematic diagram of a flow cell fitted to an ultrasonic hOrD (Good ... ·in. 1990) 
For an ultrasonic horn system, parameters such as frequency and ultrasonic power must be known to ensure 
that daily working conditions are consistent, results are reproducible and are based on some absolute values 
to allow for comparison and scale-up for industrial systems (Perkins, 1990). The frequency of an ultrasonic 
horn is provided by the manufacturer, this value can be used for comparison since a 5 to 10 % change in 
frequency does not significantly affect experimental results (Perkins, 1990). The rated electrical power of 
an ultrasonic horn system does not give an indication of the ultrasonic power because of the 
power-b)'«mand characteristic of many ultrasonic systems. The ultrasonic power drawn in a particular 
experiment wiU depend on the load placed on the horn (which is determined by the magnification of the 
horn) and the depth of the horn immersed in the reaction solution (Perkins, 1990). Ultrasonic power can be 
detennined by the following three methods (Perkins. 1990). 
• 
• 
Ultrasonic power can be calculated from calorimetry by recording the initiaJ increase rn 
temperature during sonication. 
The measurement of the vibrational amplitude at the tip of the horn (which is proportional to the 
ultrasonic power) will indicate any changes in the transmission of power. 
Chapter 2 ULTRASOtTh'D 2-59 
• Ultrasonic power can be calculated if the overall ultrasonic transfer efficiency of the transducer is 
known and the real electrical power to the transducer is measured. 
The temperature of a solution sonicated with an ultrasonic born increases during sonication and must be 
controlled by circulating cooling water around the reaction vessel (Pugin, 1987). In experiments 
monitoring the effect of temperature on a reaction, the temperature within the reaction vessel should be 
recorded since it may be more than 10 Co higher than the surrounding cooling water (Pugin, 1987). 
The perfonnance of ultrasonic baths and horns have been compared by various researchers. The results are 
specific to the ultrasonic equipment and. the model reactions used, however, an indication is obtained of the 
greater effectiveness with which ultrasonic horns operate (Mason and Berlan, 1992). Mason and Berlan 
measured the amount of iodine liberated from a 4 % potassium iodide solution (containing 10 % carbon 
tetrachloride) within a 5 min sonication period for an ultrasonic bath and horn (Mason and Berlan, 1992). 
A Kerry Pulsatron 55 cleaning bath (frequency, 35 kHz) filled with water containing 5 % detergent was 
used to sonicate a 55 mL reaction solution in a 250 mL conical flask.; a Sonies and Materials VC600 
ultrasonic horn system (frequency, 20 kHz) was used to sonicate a 27,5 mL reaction solution (horn tip 
diameter, 12 mm). The sonochemical yield was lOO ng of iodine per W for the ultrasonic bath and 350 ng 
of iodine per W for the ultrasonic horn (Mason and Serlan, 1992). Pugin compared the performance of an 
ultrasonic bath (Laborette 17; frequency, 35 kHz) and ultrasonic horn (Branson 830 Sonifier; 
frequency, 20 kHz) using the reaction of I-bromopentane with lithium wire in tetrahydrofuran solution as a 
model reaction (Pugin. 1987). The rate ofdecTease in l-bromopentane concentration, at 50 % conversion, 
was 1,90 % per min for the ultrasonic bath and 2,70 % per min for the ultrasonic horn (Pugin, 1987). 
Ratoarinoro and co-workers compared the perfonnance of an ultrasonic bath (Branson 2200E: frequency. 
47 kHz) and ultrasonic horn (Sonies and Materials; frequency, 20 kHz) at constant power per unit volume 
(1,3 W cm·3) using the Michael addition reaction of ethyl malonate to chalcone in toluene under solid-liquid 
phase transfer conditions (Ratoarinoro et a1., 1992). The reaction yield was 77 % for the ultrasonic bath 
and 98 % for the ultrasonic horn (Ratoarinoro et al.. 1992). It was concluded that an ultrasonic horn was a 
more effective emitter of ultrasonic energy than an ultrasonic bath (Ratoarinoro et al ., 1992). 
2.5.4 Ultrasonic reactors 
Industrial applications of ultrasound, because of the chemical and physical effects of ultrasonic cavitation. 
have the potential to be operated with cheaper reagants, shoner reaction cycles and less ex1rcme ph~'sical 
conditions. possibly leading to less expensive and smaller plants (Martin and Ward. 1992). 
An ultrasonic reactor is possibly nOl required if reaction enhancement is due to the physical effects of 
ultrasound, rather, a slurry could be pretreated with ultrasound before the reaction take place within a 
conventional reactor. Reaction enhancement due to the chemical effects of ultrasound will require the 
Chapter 2 ULTRASOUND 2-60 
design and development of an ultrasonic reactor since the presence of ultrasound affects the reaction 
chemisuy (Mason. 1992). 
Difficulties associated with the sca1e-up of sonochemical processes are the estimation of ultrasonic energy 
required 10 perform a particular chemical transformation and the method by which this amount of energy is 
to be transferred to a large-scale process (Mason and Berian. 1992). Laboratory equipment such as an 
ultrasonic bath or horn are not suitable for direct scale-up of a sonoc:hemical process (Martin and 
Ward, 1992). The uansmitted intensity within an ultrasonic bath is too low and the ultrasonic field is 
non-uniform and would be localised in the vicinity of the reactor waUs. Significant beating of the reaction 
solution would take place because of the large number of transduoers required for such a reactor. The 
mounting of the transducers 10 the reactor walls may also be a problem since conventional reactors are often 
constructed with an outer layer of insulating material or a water jacket, these would have to be modified or 
removed to allow for the transducers 10 be mounted 10 the inner walls (Goodwin., 1990). An immersible 
bank of transducers offers greater flexibility than permanentlY-DlOWlted transducers but a non-unifonn 
ultrasonic field is produced and problems could be experienced with erosion of the metallic transducer cover 
and sealing of transducers and electrical cables from chemical reagants that may be corrosive or highly 
inflammable (Goodwin. 1990). An ultrasonic horn generates a small zone of high intensity near the tip of 
the born, a multiple anay of ultrasonic horns would be required for the sonication of a large volume of 
solution. Other problems associated with an ultrasonic born would be the erosion of the horn tip and the 
shielding of transducers and associated electrical supply from the reaction solution (Goodwin. 1990). 
The range of ultrasound penetration in liquids is Limited because intensity decreases with distance from the 
radiating surface: the presence of solid panicles and cavitation bubbles cause further attenuation of the 
sound waves. Cavitation thus only occurs within a distance of a few tens of centimetres from the radiating 
source (Martin and Ward, 1992). The limited range of cavitation indicates that the design of large-scale 
ultrasonic equipment should rather be based on a flow system in which a small volume, of limited diameter. 
is sonicated and the reaction solution be pumped through this sonicated zone (Martin and Ward, 1992: 
Mason. 1992). An ultrasonic reactor. based on a flow system, generally consists of a flow loop outside a 
conventional batch reactor that acts as a reseryoir or bolding tank. as shown in Figure 2.14. or it can be 
configured as a cylindrical tube reactor that is included in an existing pipe network of a process (Martin and 
Ward. 1992; Mason, 1992). 
Cbapter 2 ULTRASOlJ1II"D 2';;1 
reservoir or 
bolding tan.k 
inlet 
sonicated 
ultrasonic 
transducen 
pump 
Figure 2.14: Ultrasonic reactor for use iD batch, semi-batcb or cODtiDuoos--now operation 
(Martin and Ward, 1992) 
The flow system shown in Figure 2.14 can be used for batch, semi-batch or continuous-flow operation 
(Martin and Ward 1992). M'aDtage5 of such a system are the easy control of residence time in the 
sonicated zone (by controlling the solution flow rate), simplicity of retrofit to conventional reactors and less 
uncertainty about scale-up since the same type of ultrasonic module is used in a pilot-plant and large-scale 
reactor (Martin and Ward, 1992). Temperature is controlled by heat exchange in the circulating reaction 
solution (Mason, 1992). Scale-up may be made more favourable if the limiting reagent of a reaction is 
retained in the sonicated zone or if the reagants exhibit a memory effect and remain activated after passing 
through the sonicated zone (Martin and Ward. 1992). The following issues have to be considered in the 
design of an ultrasonic module for a flow system as shown in Figure 2. 14 (Martin. 1992), 
• 
• 
• 
method of mounting transducers to a circular duct, 
coupling oftransduccrs to the liquid load, 
controlling the interaction through the hardware and in the reaction solution between multiple 
transducers, 
• ensuring reasonable transducer life. and 
• providing for the selective removal of transducers for maintenance. 
Various designs of ultrasonic reactors are commercially available. The Hanl'ell sonochemical reactor was 
designed at the Harwell laboratory of the Atomic Energy Authorit)' in the United Kingdom (Grinthal and 
Ondrcy. 1992). This reactor consists of a pumped loop as shown in Figure 2. 14; the volume of the 
sonicated zone is 4 L. The ultrasonic module, shown in Figure 2.15, consists of three transducers set 
around a cylindrical duct, 13 cm in diameter and 30 cm long (Martin. 1992; Martin and Ward, 1992). 
Chapter 2 ULTRASOUND 2-62 
F igure 2.15 : Ultrasonic module of the Hanvell sooocbemical reactor (Martin, 1992) 
The transducers operate with a frequency of 20 kHz and a combined power rating of 1 500 W. Each 
transducer can be removed separately for maintenance; transducers are coupled to the reaction medium 
through a buffer fluid with a high cavitation threshold (Martin and Ward, 1992). The ultrasonic moduJe 
can be incorporated in a flow system to handle now rates up to 230 L h'] (Pandit and Mohoikar, 1996). The 
Harwell sonochemical reactor has been used to achieve a 50-fold increase in the rate of fonnat ion of 
I-phenyl-l-butanol from the reaction of n-bromopropane and benzaldehyde (cata1ysed by lithium in 
tetrahydrofuran solution) and a 5 to lO-fold increase in the rate of fommtioD of methanol and sodiwn 
benzoate from the hydrolysis of methyl benzoate (Grinthal and Ondrey, 1992; Martin and Ward, 1992). 
The Harwelt sonochemical reactor has also been used in crystallisation to improve crystal purity 
(Martin, 1993). 
tJ]trasonic tube reactors are not omy circu1ar in dimension; hexagonal and pentagonal shaped pipes allow 
for easier attachment of transducers to a planar wall rather than a curved surface (Mason, 1992). Different 
cross sections of uJtrasonic tube reactors are shown in Figure 2. 16 (Mason, 1992). 
pentagonal hexagonal circular 
Figure 2.16: Cross sections of ultrasonic tube reactors (Mason, 1992) 
Chapter 2 ULTRASOUND 
A tube reactor can be retrofitted in an existing pipe network if the length of the reactor is designed so that a 
null point exists at each end (Mason, 1992). Tube reactors are able to handle high flow rates and can be 
used for the treatment of viscous materials. The pentagonal pipe designed by Branson Ultrasonics 
(Danbury, Connecticut, USA) provides a fairly uniform low-power ultrasonic field; the energy irradiated 
from each tube wall is reflected at an angle by the two opposite tube walls (Mason, 1992). The hexagonal 
pipe designed by Sonic Services (Paris. France) provides a high power focus in the centre of the tube, this is 
where the direct energy inadiated from each tube wall meets the reflected energy from the opposite wall 
(Mason, 1992). 
A cylindrical tube reactor also provides a focus of energy in the centre of the tube (Mason, 1992). 
Cylindrical tube reactors have been designed by a number of research group> and manufacturers. The 
resonating pipe desigood hi' BatteDe (Columbus, Ohio, USA) is 15,3 cm in diameter, oper.ltes at a frequency 
of 25 kHz and has been used for the degassing of oils (Mason, 1992). The cylindrical reactor designed at 
the University of Milan (Italy) has a central pipe that carries cooling liquid, as shown in Figure 2.17. The 
ultrasonic energy field is made more uniform by the coaxial reflection of the sound waves by the cooling 
pipe. The reactor volume is the annular space between the resonating pipe walls and the cooling core 
(Mason, 1992; Pandit and Moholkar, 1996). 
reagent 
re«! 
-reagent 
re«! 
/ 
product 
-
 reaction zone 
may contain 
solid reagent/catalyst 
Figure 2.17: Schematic diagram of tbe cylindrical reactor (with core cooling) designed at the 
University of Milan (Mason, 1992) 
The Sonilube reactor designed by Sodeva (Bonnesur·Menoge, France) consists of a transducer system 
coupled directly to an annular collar that acts as a cylindrical resonator, as shown in Figure 2.18. The 
radial collar is screw·fitted to a stainless steel pipe of specific length such that a null JXlint exists at each 
end This allows for the inclusion of the reactor in an existing pipe network without causing vibration of 
the pipe work.. A Sonilube with an operating length of 1,2 m and internal tube diameter of 42 mm was 
driven at 2 kW with an 80 % efficiency (Mason, 1992). 
Chapter 2 ULTRASOUND 
n l 
HW 
u= 
t 
t 
solution 
flow 
J 
"""'" 
......- collar 
stainless steel 
-- tu'" 
Figure 2.18: Schematic diagram of a Sodeva SonilMbe (Muoa, 1992) 
Uluasonic flow systems are not only designed as tube reactors. lbe Nearfield tJCOJIStic processor designed 
by the Lewis Corporation (Oxford, Connecticut, USA) consists of two sonicated metal plates that operate at 
different frequencies; the reaction solution flows between the plates (Mason and Berlan, 1992). The 
ultrasonic intensity in the solution between the plates (operating at frequencies of 16 and 20 kHz) is greater 
than the sum of single plate intensities. The Nearfield acoustic processor, shown in Figure: 2. 19, is driven 
by magnetostrictive transducers and can be used to sonicate high flow rates of material (Serlan and 
Mason, 1992; Mason and Berlan, 1992) , 
stainless steel 
plates 
/ 20 kHz transducers 
~--==--___________ ,'-:;;;::<~_ solution 
r flow 
'-- 16 kHz transducers 
Figure 2.19: Schematic diagram of a Nearf~ acou.stic processor (Mason, 1992) 
The Nearfield acoustic processor has been used to degrade low concentrations of water poUutants such as 
p~nitrophenol, benzene and toluene (Hua et al., 1995a; Thoma et al., 1998). The degradation of 
p-nitrophenol increased proportionally with increasing power-to-volume ratio over the range of 0,98 to 
7,27 W cm-2. Degradation was greatest when a 4:1 (vlv) mixture of argon and oxygen gas was bubbled 
through the solution. A comparison of the energy efficiency of an Nearfi eld acoustic processor and an 
ultrasonic horn system (Sonics and Materials VCX400; power, 130 W; sample volume, 25 mL) indicated 
that the G~value, defined as the number of molecules degraded per unit of energy input into the system, was 
64 x 101' and 2,3 x IOu molecules krl , respectively (Hua et al., 1995a). The investigation of effiuent 
Chapter 2 ULTRASOUND 2~5 
containing benzene and toluene concluded that for scale-up, ultrasound would be most effective as a 
pretreatment for relatively high target concentrations rather than as a polishing stage to meet effluent 
guidelines (fhoma et al., 1991!). 
The Lewis Corporation have also produced an ultrasonic vibrating tray (Mason. 1992). The vibrating tray 
was originally designed for the processing of coal and metal ores at flow rates of up to 20 ton h-l . The 
vibrating tray could be used for the ultrasonic pretreatment of reaction slurries (Mason. 1992). 
The ultrasonic push-pull system designed by Martin-WaIter (Straubenhardt, Germany) consists of a 
cylindrical bar of titanium, cut to a length equal to a multiple of half-wave1engtbs of sound, with opposing 
piezoelectric transducers bonded to each end (Mason. 1992). The combined oscillations of the transducers 
(connected electrically through a centra] bole in the cylinder) sets up a concertina effect down the length of 
the cylinder. The push-pull system can be easily retrofitted co-axiaIly in the centre of an existing pipe 
network. However, this system is susceptible to erosion and there is a lack of information relating to 
scale-up and industrial implementation (Mason. 1992; Pandit and Moholkar, 1996). 
The 250 mL ortho reactor designed by Undatim Ultrasonics S.A (Louvain la Neuve, Belgium) has two 
transducers (with frequencies of 20 kHz and 1,7 MHz) operating at right angles to eacb other (Grinthal and 
Ondrey, 1992). ]be interaction between the two ultrasonic fields is not completely understood, however, 
the different frequencies ensure efficient mass transfer and sonochemical reactions occur during sonication 
(Grinthal and Ondrey, 1992). 
The Branson sonochemica1 reactor consists of a 200 mL module with a pair of 20 kHz horn-type sonicators 
mounted on either side of a pipe containing the reaction solution (Pandit and Moholkar, 19%). Each 
sonicator operates with a power rating of I 800 W; the ultrasonic energy from the transducers is coupled to 
the pipe by means of a coupling fluid The modules may be connected together for continuous-flow 
operation. This system is susceptible to erosion of the horn tips (Pandit and Moholkar, 1996). 
The Hquid whistle reactor designed by Dawe Ultrasonics (Hayes, UK) is one of the oldest commercial 
ultrasonic Qevices (Mason. 1992). The liquid whistle, shown in Figure 2. 20, operates in a similar way as a 
gas whistle; cavitation occurs in a liquid as it passes across a vibrating blade. The vibrations are caused by 
the high Dow rate of the liquid. The liquid whistle provides efficient mixing, homogenisation and 
dispersion in a Dow system. Scale-up is easily achieved since a pump and blade are the only moving parts 
in the system. Erosion of the vibrating blade may be caused by paniculate matter in the reaction solution 
(Mason, 1992; Pandit and Moholkar, 1996). 
Cbapter 2 ULTRASOUJII"D 
cavitation 
flow guides 
sonicated 
liquid 
vibrating blade 
stainless steel 
blook 
Figu~ 2.20: Schematic diagram of aD ultruoaic liquid whistle (MUOD, 1992) 
2.5.5 Equipment design and reactor modelling 
Cavitation chemistry and thus the design of ultrasonic equipment is influenced by a variety of parameters 
that should be optimised for the sonication of a chemical process. Some examples of such parameters are 
listed in Table 2.5 (Berlan and Mason, 1992; Mason. 1992; Shah et al., 1999c; Tbompson and Doraiswamy, 
1999). 
Tabk 2.5 : Parameten that affect ca.vitatioD (Muoa, 1992) 
Parameter 
reaction medium 
reaction conditions 
type of ultrasonic system 
viscosity 
vapour pressure 
presence of solid particles 
pressure 
temperature 
presence of dissolved gas 
power 
frequency 
size and geometry of chemical reactor 
U1trasonic effects arc determined primarily by the amount of aroustic power deli'\rered to a system. Power. 
for comparison between different experimental setups, is reported in literature as intensity. either, in terms 
of the surface radiating area of the ultrasonic source or the volume of liquid sonicated. Methods of 
reporting ultrasonic intensity (also referred to as power density) are summarised in Table 2,6. Perkins 
defined intensity as the electrical power supplied to the transducer divided by the radiating surface area of 
the ultrasonic born (Perkins, 1990). 
Chapter 2 ULTRASOUl'I'D 2-<i7 
Table 2.6: Metbods of reporting ultrasonic inteosity (Perkins. 1990) 
Units 
area~ndant quantity Wcm-2 
volume-dependant quantity Wcm-3 
Calculation 
acoustic power divided by the radiating surface area 
of the ultrasonic source 
acoustic power divided by the volume of the liquid 
being sonicated 
Ultrasonic effects increase with increasing intensity. the rate of degradation of chloral hydrate increased 
linearly with increasing intensity from 0 to 1,5 W cm'2 (Sakai et al., 1977). The rate coostant for the Fricke 
dosimeter (conversion of Fe2" to Fe~ increased linearly with increasing intensity from 5 to 30 W cm-2 
(Price and Lenz, 1993). Acoustic power cannot, however, be increased indefinitely, The cavities formed by 
the sound waves become to big to implode within a single rarefaction stage and the formation of a large 
number of bubbles at the swface of the ultrasonic source act as a shield that diffuses and disperses the 
acoustic energy (Mason. 1990; Pestman et al., 1994), SehgaJ and Wang found that the sonochemical 
reaction of thymine (15 mL reaction sample; 12,6 mg L-1 solution) reached a maximum at an intensity of 
3,5 W cm'2 and decreased at higher intensities (SehgaI and Wang, 1981), Henglein and Gutierrez found 
that the yield of iodide oxidation also reached a maximum and decreased at higher acoustic powers, the 
power at which the yield was a maximum increased as the reaction volume increased from 5 to 40 mL 
(Henglein and Gutierrez, 1993). 
Acoustic power can be measured calorimetrically or from the difference in electrical power consumed by a 
transducer when the ultrasonic horn is loaded and without load. The calorimetric method to calculate 
acoustic power is based on the measurement of the rate of temperature increase in a system. taking into 
account its thermal capacity (Delchar and Melvin. 1994; Perkins. 1990). Calorimetry was used to calculate 
acoustic power in an investigation of the degradation of benzene and styrene (de Visscher el at. 1996); in 
the degradation of thymine (Mead el aI. , 1975): in a sonochemical investigation of aqueous methanol 
(Rassokhln et al.. 1995) and in an investigation of the formation of hydrogen peroxide under an argon 
aunosphere (Rassokhin et al ., 1994). Calorimetry has been used to measure a range of intensities. from 
1,06 W cm-2 in the investigation of pentacblorophenate degradation (Petrier et al. , 1992) to 37,6 W cm-2 in 
the polymerisation of methyl methacrylate (Price et aI., 1991). Acoustic power was measured using the 
difference in electrical power method during an investigation of the chemical and physical effects of 
ultrasound (Contamine et aI., 1994; Ratoarinoro et at.. 1995). 
The effect of frequency on sonochemical reactions is detennined by two opposing phenomena. The 
compression and rarefaction cycles of ultrasonic waves are shortened as frequency increases. this leads 10 a 
reduction in size of cavitation bubbles and hence a lower cavitation intensity and radical formation. The 
Chapter 2 ULTRASOUND 2-68 
smaller bubl:Nes, however, lead to a greater number of radicals escaping from the bubbles and reacting with 
solutes in the water (Mason, 1993; Petrier et al, 1992~ Pdrier and Francony, 1997). Petrier and co-v.'Orkc:rs 
investigated. the effect of frequency (20; 200; 500 and 800 kHz) on the initial rate of sonochemica1 
degradation of a 94 mg L·1 phenol solution, the rate of degradation was greatest at 200 kHz and decreased 
in the order for 500, 800 and 20 kHz (PCtrier and Franrony, 1997). The rate of soooc:hemica1 degradation 
of chloral hydrate in aqueous solution was 1,8 times greater at a frequency of 400 kHz than at 29 kHz 
(Sakai et al., 1977). Intensity and emission spectra of sonoluminescence are also dependant upon frequency 
(Mason, 1993; Sebgal et ~ 19803). Didenko and co-workers found that sonolllminesreoce intensity was 
greatest at a frequency of 337kHz and decn:ased at frequencies of I lOO and 22kHz (Didenko et al., 1994). 
The design of ultrasonic equipment should allow for the option of s:pa.rging with a gas during sonication. 
The investigation of the degradation of hydantoin compounds in an ultrasonic bath demonstrated that 
oxygen concentration of an oxygen-saturated solution was reduced, within 3 b, to that of an unsaturated 
solution due to the deg;lssing nature of ultrasound (Scbwikkard, 1995). Dissolved. oxygen concentration is 
an important parameter to quantify and control in an ultrasonic system to ensure 0JXima1 performance and 
reproducible results. 
Mason and Cordemans de Meulenaer made the following 10 recommendations in the optimisation of an 
ultrasonic process (Mason and Cordemans de Meulenaer, 1998): 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
Make cavitation easier by the addition of solids or gas bubbles to act as nuclei. 
Try entraining different gases or mixtures of gases. 
Tr)' different solvents for different temperature ranges and cavitation energies. 
Oplimise the power required for the reaction. 
When using a solid-liquid system do not charge all the components in the reactor at once. 
L[ possible, tJy to homogenise two-phase systems as much as possible. 
Try different shapes (diameters and volumes) for the reactor. 
It can be better (but not always) to avoid standing wave conditions by performing sonochemica1 
reactions under high-power conditions with mechanica.1 stirring. 
Where possible, try to transform a batch system into a continuous one. 
Choose conditions which allow comparisons betv.'een different sonochemica1 reactions. 
The choice of a continuous or batch system is dependent on the ultrasonic system. Shah and co-workeJ5 
recommended the use of continuous processes when a high throughPUt is required and the use of batch 
processes when (Shah et al., 1999d): 
• residence time is relatively long 
• there is a need for more precise control of ingredients 
Chapter 2 ULTRASOUND 2-69 
• backmixing is not desirable 
• several different products are made in essentially the same setup 
• production volume is relatively small and a continuous process is DOl eoonomic. 
A model for gas-liquid cavitation reactors must take into account the different mechanistic sub-processes 
(Shah et al., 199ge). The sub-processes ofultrasooic cavitation are the formation of reactive intermediates 
and pyrolysis products due to the high temperatures and pressures in the collapsing cavities; the rate at 
which the reaction products and other intermediates are released into the tulk solution; and the rate of heat 
and mass transfer from the collapsed cavity into the bulk: solution and the reactions between the transferred 
intermediates and other species present in solution (Shah et al., 199ge). 
Proper characterisation of the nature and distribution of gas cavities in a reaction vessel (size, number and 
location) is important in a cavitation reactor model since the cavities are the initiation sites of chemical 
reactions (Shah et al. , 199ge). Formulation of mathematical models of the location and distribution of the 
collapsing cavities is difficult due to the random nature of the process and the lack: of knowledge about the 
relationship between cavitation energy imputed 10 a reaction solution and the size, number and. distribution 
of cavities generated from the energy. Process scale-up is thus usually done on a qualitative or empirical 
basis (Shah et al. , 199ge). 
Cavitation reactor models have been reported in literature (Gogate and Pandit, 2000; Prasad Naidu et al., 
1994; Shah et al., 1999(1). Models start with the Rayleigh-Plesset equation for bubble dynamics 
(equation 2.3) which is altered to take into account different assum~ons. Prasad Naidu and co-workers 
derived a model to predict the number of free radicals generated during cavity collapse for use in a kinetic 
model of the sonication of aqueous potassium iodide solutions (Prasad Naidu et al., 1994). The growth 
phase and initial cavity collapse were assumed to be isothermal, cavity collapse become adiabatic once the 
pressure of the gas within the cavity equalled the saturation vapour pressure of the liqu.id The model 
predicted the rate of iOOine Liberation in potassium iodide solutions under an oxygen atmosphere but less so 
under a nitrogen atmosphere (Prasad Naidu et al., 1994). 
Gog;:ate and Pandit solved the Rayleigh-Plesset equation nwnerically (assuming complete adiabatic cavity 
collapse, with and without the assumption of liquid-phase incompressibility) to develop an empirical 
correlation of the pressure generated during cavity collapse as a function of intensity, frequency and initial 
nuclei size (Gogate and Pandit, 2000). The correlation is 
[2 . 131 
where initial cavity size Ho is measured in mm. ultrasonic intensity I in W cm,2, frequency f in kHz and the 
generated pressure P collllp" in ann. TIlls correlation can be used in the design of sonochemical reactors to 
Chapter 2 ULTRASOUND 2-70 
indicate the magnitude of ~ generated during cavity ooLlapse, irrespective, of the process for which 
the reactor is being used (Gogate and Pandit, 2(00). The sonochemical yield of a process is related to 
pressure generated during cavity oollapse according to equation 2.14 
Sonochemicol yield = K(P ~y 12.141 
where the oonstant K and. exponent c are determined by reactor geometry, operating parameters and type of 
reaction being performed.. The5e sort of oorrelatioDS can be used by engineers in the design procedures for 
sonochemical processes where rigorous numerical solutions may not be feasible (Gogale and Pandit, 2(00). 
2_6 CONCLUDING REMARKS 
Ultrasonic cavitation results in both chemical and physical effects. The chemical effocts of ultrasound are 
due to the free radicals formed because of the high temperatures and pressures in the coUapsing gas cavities. 
Radical reactions occur within the gas cavity, radical species and products diffuse into the bJlk solution and 
react with solutes in the water. The presence of a gas influences the radical reactions occurring during 
sonication. Thermal dissociation processes (pyrolysis reactions) dominate within the gas cavities whereas 
reaction kinetics analogous 10 radiation chemistry occur in the bJlk solution. The physical effects of 
ultrasound are due to shodcwave formation and. microject formation dnring asymmetric cavity ooUapse. 
The potential application of ultrasound in water treatment is indicated in literature. UlI.raSOWK1 enhances 
the biocidal activity of chlorine due to break-up and dispersion of bacteria clumps. Anaerobic digestion of 
sewage sludge is accelerated and sludge dewatering is enbanoed. Low intensity ultrasoWld also increases 
the activity of activated sludge microorganisms. Various organic poUutants are degraded during sonication. 
UltrasoWld, however. is best used. in conjWlction \\ith other technologies as a pretreatment step. 
Parameters such as acoustic power, ultrasonic frequency and the dissolved concentration of a gas must be 
quantified to characterise reaction conditions for comparative and scale-up purposes. The direct scale-up 
from ultrasonic baths and horns is not suitable for industrial applications. Larger-sca1e equipment such as 
the Harwell sonochemical reactor and the NearjieJd acoustic processor are based on a flow system in which 
the reaction solution is pumped through a small sonicated zone. either being recircu1ated or on a 
once-through basis. 
3 
OZONE AND HYDROGEN PEROXIDE CHEMISTRY 
Advanced oxidation processes are defined as processes (such as ozone) that generate sufficient quantities of 
hydroxyl radicals to affect water treatment (Glaze et al., 1987). The formation of hydroxyl radicals is 
enhanced by the simultaneous addition of hydrogen peroxide. Advanced oxidation processes may totally 
mineralise organic compounds, such as herbicides and pesticides, that are difficult to degrade during 
conventional water treatment. Partial oxidation, where conditions required for total mineralisation are not 
practical, may produce less recalcitrant compounds for microbial processes; under such conditions advanced 
oxidation processes are beneficia1 as pretreatment steps to other processes (Vue, 1992). 
Oz.one chemistry and the kinetics of ozone decomposition are reviewed in Section 3.1, the application of 
ozone in water treatment in Section 3.2 and hydrogen peroxide chemistry and application in water 
treatment in Section 3.3. 
3.1 OZONE FUNDAMENTALS 
Ozone is a triatomic allotrope of oxygen and is an unstable, blue, diamagnetic gas with a characteristic 
pungent odour (Greenwood and Eamshaw. 1984b). The bond angle of the triangular-shaped molecule was 
delcnnined in microwave studies 10 be 116,8 ± 0,50 and the bond lengths 0, 1278 run (Greenwood and 
Eamshaw, 1984b). The structure of ozone is a hybrid ofthc resonance structures shown in Figure 3. 1. 
,0, +0 .. + '0, ,() / 
" 
.. ~ 
,of '00_ .. ~ ,0(°"';0, .. ~ / " .. ~ / " ,0, ,0, ,0, ,0, ,0'----0-+ + 
(a) (b) (c) (d) (e) 
Figure 3.1 : Resonance structures of tbe ozone molecule (Barcourt et al., 1986; Nebel, 1981) 
The bonding of structures (a) to (d) in Figure 3.1 consist of two cr bonds and one delocalised 1t bond in 
which thc orbital is stretched over the three o>..'ygen atoms (Greenwood and Eamshaw, 1984b; Nebel, 1981). 
The strong electrophllic or electron-deficient nature of ozone is indicated by structures (a) and (d) that are 
characterised by a terminal atom with only six electrons (Nebel, 1981). Structure (e) in Figure 3.1 has also 
Cbapter 3 OZOl"l"E AND HYDROGEN" PEROXIDE CB£M1.STRY 3-2 
been proposed for the ground-state resonance of the ozone molecule. This structure consists of m'o a bonds 
and a third long bond or 7t bond between the two tenninal oxygen atoms (Harcourt and Rosa, 1978; 
Harcourt et al. , 1986). 
Ozone was initially detected by means of its smell. Van Mauren, in 1785, first recorded the distinct odour 
accompanying an electrical spark discharge in oX)'gen gas (Greenwood and Eamshaw, 1984b; Nebel, 1981). 
Ozone was named (from the Greek word ozein meaning to smell) by SchOnbein, in 1840, who recognised it 
as a new compound (Greenwood and Earnshaw, 1 984b~ Rocke, 1994). The chemical structure of ozone was 
shown by Soret, in 1866, to be that of triatomic oxygen (Streng, 1961). 
The chemical and physical properties of ozone are reviewed in Section 3.1.1, the kinetics of decomposition 
in Section 3. 1.2 and rate constants ofradica1 reactions in Section 3. 1.3. 
3.1.1 Chemical and physical properties 
Some physical properties of ozone are listed in Table 3.1. 
Table 3.1 : Physical properties of ozone (Greenwood and Ea rnsbaw. 1984b; Streng, 1961) 
Parameter 
molecular weight 
molecular formula 
[Ollll 
melting point 
boiling point 
density (liquid) 
density (solid) 
viscosity (liquid) 
viscosity (gaseous) 
Value 
47,998 
0, 
blue gas at room temperature 
·192,5 °C 
-111 ,9°C 
1 354 kg m ') at 0119,4 °C 
1728kg m·3 at-195,8°C 
1,57 cp at _183°C 
1.33 x 10.2 cp at 25 °C 
Ozone condenses to a deep blue liquid and to a violet-black solid (Greenwood and Eamshaw, 1984b). 
Liquid and solid forms of ozone are explosive due to rapid decomposition to gaseous oX)'gen. Gaseous 
ozone is also thennodynamically unstable with respect to decomposition to oxygen gas. though the 
decomposition is slow in the absence of catalysts or ultraviolet radiation (Greenwood and Eamshaw. 
1984b). 
f o ,(g) - O, (g) M/t' ~ 142,7 kJ mor ' ; "Gt' ~ 163,2 kJ mor' 
Scbeme 3.1 
Chapter 3 OZONE AND HYDROGEN PEROXIDE CHEMISTRY 3-3 
Although ozone is a powerful oxidant, it is selective in its action (Bablon et al., 1991b; Nebel, 1981). The 
ozone oxidation of potassium iodide, shown in Scheme 3.2, indicates the strongly oxidising nature of ozone 
and the tendency of ozone to transfer an oxygen atom during reaction to produce oxygen gas (Greenwood 
and Earnshaw, 1984b). 
0 3 + 21- + H20 - O2 + h + 20H 
Scbeme 3.2 
3.1.2 Classification of ozone reactions 
Oxidation of organic solutes in water by ozone is due either to direct reaction with ozone molecules or 
reaction with free radicals fonned from ozone decomposition (Beltrfm, 1995; Hoigne, 1975; Hoigne and 
Bader, 1976). Ozonation is a gas-liquid reaction and the influence of mass transfer (gas absorption) has to 
be considered in determining whether both direct and decomposition reactions occur (Beltrin, 1995). 
Beltcin and co-workers investigated the competition between direct and decomposition reactions using film 
and surface-renewal models for gas absorption (Beltnin, 1995). The film model was based on the concept of 
a stagnant film at the surface of a liquid next to a gas; convection does not occur within the film and the 
dissolved gas moves through the film by molecular diffusion alone (Ch.arpentier, 1981 ; Danckwerts, 1970). 
A schematic diagram of ozone concentration in the gas bubble, film layer and bulk solution is presented in 
Figure 3.2. 
g" 
bubble 
gas-liquid 
interface 
. stagnant t stagnant ; 
gas film : liquid film: 
, , 
, 
, 
Dislance 
bulk 
solution 
moderate 
ozone demand 
high ozone 
demand 
Figure 3.2: Schematic diagram of OZOne concentration in the gas bubble, rUm layer and bulk 
solution as described by the two-film theory (Adams and Randtke, 1992a) 
Ozone. as shown in Figure 3.2, can be depleted in the film layer by a fast chemical reaction (high ozone 
demand) resulting in a negligible amount of ozone in the bulk solution whereas a slow chemical reaction 
results in the bulk solution being saturated with ozone. The kinetic regime of a gas absorbing into a liquid 
Chapter 3 OZONE AND HYDROGEN PEROXIDE CBEMlSTRY 3-4 
and then Wldergoing chemical reaction is determined in film theot)' by the value of the dimensionIess Hatta 
number. The Hatta number is defined in equation 3.1 for an irreversible first-order reaction and in 
equation 3.2 for an irreversible second-order reaction (Charpentier, 1981 ; Danckwerts, 1970). 
Ha [3.1) 
Ha [3.2) 
where k is the first-order rate constant, Do. the ozone diffusivity in water, kL the liquid-phase mass transfer 
coefficient. k.J the rate constant for the direct reaction between ozone and compoWld B in solution and CB 
the concentration of compound B. The classification of the kinetic regime of gas absorption reaction 
according 10 the Hatta number is presented in Table 3.2. 
Table 3.2 : Classification of kinetic regime of gas absorption reactions according to the Hana 
number (Charpentier, 1981) 
Ba 
< 0.02 
0.02 to 0.3 
0,3 to 3 
>3 
Kinetic regime Description 
vel)' slow reaction in bulk Reaction does not occur in the film; mass transfer keeps 
solution the concentration of the absorbed gas (ozone) in the 
bulk solution close to saturation leveL 
slow reaction in bulk 
solution 
moderately fast reaction 
fast reaction in film layer 
The process is that of physical absorption followed by 
reaction in the bulk solution; a negligible proportion of 
ozone reacts in the diffusion film. 
A substantial amount of ozone reacts in the film: ozone 
concentration in the bulk solution is very low. 
Reaction is fast and occurs completely in the film layer; 
ozone concentration in the bulk solution is negligible. 
The Hatta number indicates whether the overall reaction rate is controlled. by reaction kinetics or by the 
masNransfer coefficient. and thus whether a large liquid hold-up or large specific interfacial area is 
required in the design of a reactor for a specific reaction (Charpentier, 1981). As shown in Table 3.2. a 
Hatta number below 0,3 indicates that the reaction occurs completely in the bulk solution and that the 
reaction rate is controlled by the bulk volume; necessitating a large volume of liquid for the reaction. 
Conversely. a Hatta number greater than 3 indicates that the reaction occurs completely within the boundary 
layer and that the reaction rate is controlled by the interfacial surface area; net:essitating a contacting device 
that creates a large interfacial area. A large liquid volume and interfacial area are required for a reaction 
that occurs within both the film layer and the bulk solution (Charpentier. 1981). 
Chapter 3 02.0/10"1: AND HYDROGEN PEROXIDE CHEftUSTRY 3-5 
The surface·renewal model is 00sed on the concept of periodic replacement of elements of liquid at the 
interface by liquid from the bulk solution~ the rate of gas absorption is a function of the exposure time of the 
liquid element., initially being rapid and decreasing with time (Charpentier, 1981 ; Danckwens, 1970). The 
diffusjon time (ID) is defined as the time between two consecutive renovations of liquid elements. it is the 
available time for ozone to diffuse into the liquid elements; reaction time (tR) is defined as that required for 
the reaction to proceed at an appreciable rate (Astarita., 1967 in BeltJin, 1995). The expressions for the 
diffusion and reaction times are presented in equations (3.3], (3.4] and [3.5]. 
= Do, 
to kl. 
(for a second-{)roer reaction) 
1. (for a first~rder reaction) k 
[3 .3J 
(3 .4J 
[3.5J 
where k is the first-{)rder rate constant, Do, the ozone diffusivity in water, kL the Jiquid·phase mass transfer 
coefficient, ko the rate constant for the direct reaction between ozone and compound B in solution and CB 
the concentration of compound B in solution. The ratio of the di1fusion time to the reaction time as defined 
according to the surface·renewal model is equal to the square of the Hatta number as defined according to 
the film layer model (Beltnin, 1995). The classification of kinetic regime of gas absorption reactions 
according to diffusion and reaction times is presented in Table 3.3. 
Table 3.3 : Classification of kinetic regime of gas absorption reactions according to diffusion and 
reaction times (Beltnin, 1995) 
Kinetic regime 
slow reaction 
ID» l R fast reaction 
Description 
Reaction is slow since the surface liquid elements are 
renewed before the reaction can develop appreciably. 
Reaction is fast since the dissolved gas is completely 
depleted between renovations of surface liquid. 
where ID is the diffusion time, the time between two consecutive renovations of liquid elements 
I~ is the reaction time, the time required for a reaction to proceed at an appreciable rate 
A kinetic regime of moderate reactions is defined to occur when tR is between 0,1 and 10 times that of tD 
(Beltnin, 1997). The limit bem'een slow and moderate reactions (tR ~ lOrD) is called the diffusional regime, 
ozone diffusion controls absorption rate in this region (BeltJin, 1997). 
Cbapter 3 OZONE AND HYDllOC!.N PEROXIDE CHEMISTRY 
The presence of dissolved ozone in the b.J.lk solution, according to Table 3.2. indicates that reactions are 
slow and take place in the bulk solution whereas the absence of dissolved ozone in the solution indicates 
that reactions are fast and take place in the proximity of the interface (Beltn\n, 1995). Beltn\n investigated 
the effect of pH on the ozone decomposition reaction and found that at a pH of 2 and 7 the bulk solution was 
saturated ·with ozone, thus, decomposition was slow and took place in the bulk: solution whereas at a pH of 
12 the dissolved ozone in the bulk solution was negligible and ozone disappeared due to a fast reaction 
while diffusing through the film layer (Beluin, 1995). 
OLone decomposition and the direct ozone reaction with a generic compound B takes place in the same 
reaction zone during ozonation when kd. < 106 M"I S· I and CB < 10-' M or when pH ~ 10 and kd. ~ 106 M·I 5.1 
with CB < 10~ M (Be11:l"!ln, 1995). Beltdn concluded that the ozone decomposition reaction would not be 
able to compete with the direct ozone reaction when the rate constant for the direct reaction between ozone 
and a surface water pollutant (concentration typically below IO~ M) was significantly greater than 
106 M"I 5.1 (Beltr.in, 1995). 
3.1.3 Kinetics of ozone decomposition 
OLone decomposition occurs through a series of radical reactions that in pure water are initiated by the 
reaction of ozone with hydroxide ions, HO' (Hoigne, 1975; Staehelin and Hoigne, 1982). Decomposition 
increases with increasing alkalinity (Hoigne and Bader, 1976; Sotelo et al., 1987; Staehelin and Hoigne. 
1982). Staehelin and Hoigne found that the decomposition of ozone was slow but measurable at a pH 
below 8, easily measurable at a pH between 8 and 10 and too rapid to be measured at a pH above 10. 
decomposition was pH dependant because it was determined by HO' concentration (Staehelin and Hoigne, 
1982). Ozone decomposition is also temperature dependant., at a pH of 7 decomposition increased with 
increasing temperature from 10 to 40 cC: decomposition was extremely fast in experiments pcrfonned at a 
temperature above 30°C and a pH above 8 (Sotelo et al.. 1987). 
Various reaction mechanisms have been proposed for the decomposition of ozone. The reaction mechanism 
proposed by Hoigne and co-workers for ozone decomposition in pure water at a neutral pH is presented in 
Scheme 3.3 (Sablon et ai. , 1991b; Stachclin et aI. , 1984). 
initiation step 
0 3 + HO- ~ 0;- + HO; 
Propagation steps 
HOi ~ Oi- + H+ 
0 3 + O2- ~ 0;- + O2 
pK. ' 4,8 
lal 
(bJ 
Ic] 
Chapter 3 OZONE AND HYDROGEN PEROXIDE CHEMISTRY 
0 ;- + H~ !i HO; 
HO; ~ HO· + O2 
0 3 + HO· ~ HO. 
HO. ~ HOi + O2 
Tennination steps 
HO. + HO. ~ H20 2 + 203 
HO~ + HO; - H20 2 + 0 3 + O2 
Scheme 3.3 
k2 = 5,2 X 1010 MI s·' 
k3 = 1,1 x 10' M' S·I 
k. = 2 X 109 M"' S·' 
kj = 2,8 X 104 M"' S·, 
Idl 
lel 
In 
IgI 
3-7 
[hI 
lil 
The mechanism for ozone decomposition shown in Scheme 3.3 includes, as intennediate compounds, the 
species HDJ· and H04· that bave not been measured directly (Bablon et al., 1991b; Westerhoff et al., 1997). 
Computer simulation was used by Staehelin and co-workers to prove that HO.· must be formed, most likely 
as a charge-transfer complex (Ho·en) that decays into HOz· (Staehelin et al. , 1984). Bahnemann and Hart 
also proposed the fortrultion of H~· radicals from the reaction of ozone and hydroxyl radicals, as in 
reaction If] of Scheme 3.3, with subsequent decay into HOz· and ox-ygen, as in reaction [g) of Scheme 3.3 
(Bahnemann and Han, 1982). The chain reactions occurring during ozone decomposition (Scheme 3.3) are 
presented in Figure 3.3. The reaction numbers of Scheme 3.3 are included in Figure 3.3. 
0 , 
HO- (a) 
,t-- _ 
ehain 
breakdo'Ml 
Figure 3.3; Reaction diagram for the ozone decomposition process (Staehelin et al •• 1984) 
A mechanism for ozone decomposition has also been proposed by Toniiyasu and co-workers: the 
mechanism does not include the formation of Hen· and H0.4• compounds but has only been verified with 
Chapter 3 OZOI'l"E.ur-'1J HYDROGEN PEROXIDE CHEMISTRY 
pulse radiolysis data at pH values above 10 (Bablon et al., 1991b; Westerhoff et al., 1991). Ozone 
decomposition according to the mechanisms proposed by Hoigne and co-workers, and Tomiyasu and 
co-workers was modelled by Westerhoff and e<rworurs using ACUCHEM, a kinetic-modelling computer 
package (Westerhoff et al., 1991). The objective was to demonstrate the use of computer models to generate 
profiles of radical species during ozonation. Both simulation models predicted similar ratios of ozone to 
hydroxyl radicals in the pH range 6,5 to 8.5 ~ ozone to hydroxyl radical ratio dc:creased by a factor of ten for 
each increase in pH of one unit. Predicted ozone oonsumption according to the Hoigne model was faster 
and matched the measured rate of ozone depletion more closely for oonditions of 5,76 mg L-1 initial ozone 
ooncentration., pH of 7,5 and 30 min reaction time (Westerhoff et al. , 1997). The Hoigne model was used to 
show the formation profiles of hydrogen peroxide, hydroxyl radicals, superoxide radicals (Oi - ) and ozonide 
radicals (0; - ) during ozone decomposition; formation of hydrogen peroxide and the subsequent reaction 
between ozone and perhydroxyl radicals become dominant with increasing pH. Ozone oonsumption in the 
presence of natural organic matter was also simulated with the model (Westerhoff et al., 1991). 
The rate-detennining step in the reaction mechanism of ozone decolDJXlSition in water is the formation of 
superoxide radical anions (0;- ). shown in reaction (a] of Scheme 3.3 and Figure 3.3 (Bablon et al., 199Jb; 
Sotelo et al ., 1987; Staehelin and Hoigne, 1982). Ozone decomposition occurs in a chain reaction (shown 
in Figure 3.3) since the fonnation of perhydroxyI radicals in reaction [a) and (g) of Scheme 3.3 lead to the 
formation of superoxide radical anions via reaction (hI and further reaction with ozone according to 
reaction (cl (Staehelin and Hoigne, 1982). 
In non pure water, ozone decomposition is initiated, promoted or inhibited by solutes present in the water 
(Staebelin and Hoigne, 1985). Decomposition is accelerated by organic solutes that convert bydroxyl (HO·) 
radicals into superoxide radical anions (Oi - ) since superoxide radical anions react fast and selectively with 
ozone while hydroxyl radicals react with a variety of compounds (Bablon et al., 199 1b: Staehelin and 
Hoigne, 1985). The reaction rate for reaction between superoxide radical anions and ozone (reaction (cl of 
Scheme 3.3) is significantly grealer than that for the reaction of superoxide radical anions with other solutes 
present in water. The chain reaction of ozone decomposition in non pure water (shown in Figure 3.4) 
consists of reactions (a] to le1 of Scheme 3.3 followed by reactions [a] to [cl of Scheme 3.4 (Staehelin and 
Hoigne, 1985). Perhydroxyl radicals, that continue the cycle of ozone decomposition via reaction [b] of 
Scheme 3.3. are formed as shown in Scheme 3.4 from the reaction of hydroxyl radicals and an organic 
solute B (Staehelin and Hoigne, 1985). 
B + HO· .... B-
 0 , + B· - BOO· 
BOO· - HO; + B"" 
Scbeme 3.4 
lal 
[bl 
Icl 
Cbapter 3 OZONE AND HYDROGEN PEROXIDE CBEMlSTRV 
0, 
HO r-___. 
0, 
Figure 3.4: Reaction diagram for tbe ozone decomposition process in the pretenee of a solute B 
(Staebelin and Boignl, 1985) 
Oxidation of solules in water during ozonation takes place, under an acidic pH, predominantly through 
direct reaction with ozone and, under an alkaline pH predominantly through reaction as shown in 
Scheme 3.4 with hydroxyl radicals fonned from ozone decomposition according to reactions la] 10 re] of 
Scheme 3.3 (Hoigne and Bader, 1976; $otelo et al., 1987). 
Initiators of ozone decomposition are compounds that generate superoxide radical anions from ozone 
molecules (Bablon et aI., 199Ib). Examples of initiators are inorganic compounds, such as perhydrnx)'1 
ions. hydroxide ions and some cations (such as Fe2+); and organic compounds, such as glyoxylic acid, 
fomUc acid and humic substances (Bablon et al ., 1991b; $olelo et al.. 1987; Staehelin and Hoigne. 1982: 
von Sonntag, 1996). Humic substances are complex macromolecu1ar structures (containing a mixture of 
predominantly carboxylic and phenolic aromatic or aliphalic carbon functional groups) that are able to act 
as initiators. promoters and inhibitors of radical reactions (Bablon et al ., 1991b; Westerhoff et al.. 1997). 
Ozone decomposition is also initiated by ultraviolet radiation at 253.7 nm (Bablon et al.. 1991b: 
Vue, 1992). Promoters of ozone decomposition are compounds that generate superoxide radical anions 
from hydrox)'1 radicals. Examples of promoters are inorganic compounds such as phosphates; and organic 
compounds such as aryl groups, formic acid, glyoxylic acid, primary alcohols and humic acids 
(Bablon et al.. 199Ib). 
Radical scavengers are compounds that react with hydroxyl radicals but do not produce species (such as 
superoxide radical anions) that propagate the chain reaction of ozone decomposition. Common radical 
scavengers (and hence inhibitQf'S of ozone decomposition) are carbonate (C0:l2") and bicarbonate (HCDJ") 
ions. alkyl groups, tertiary a1cohols (such as tert-butyl alcohol) and humic substances (Bablon el al .. 1991b: 
Glaze et al , 1992; Hoigne and Bader, 1976). Carbonate and bicarbonate ions are typically the principle 
Chapter 3 07..01'0: Al'II"D HYDROGEN PEROXIDE CHEMISTRY 3-10 
radical scavengers in natural waters but do not, however, always lead to the full inhibition of ozone 
decomposition (Glaze et al., 1992; Staehelin and Hoigne, 1985). 
Different rate equations, as shown in Table 3.4. have been proposed for the: r;IccompositioD of ozone. 
Typically, the reaction order varies, with respect to ozone concentration, from 1,5 at acidic pH to I at 
alkaline pH; and from 0 to 1 with respect to hydroxide ion concentration (Nadezhdin, 1988). Second order 
reaction kinetics, with respect to ozone concentration, may be observed in the presence of hydroxyl radical 
scavengers (Nadezhdin, 1988). 
Table 3.4: Some kinetic studies of OZODe decomposition in water (SoteIo et aL. 1987) 
Reference pH Temperarure Reaction order witb respect to 
("C) ozone bydroxide ion 
(Kilpatrick et al. , 1956) o to 7 25 1,5 
(Morooka et al., 1979) 2 t09 10 to 30 1,5 to I 0,28 to I 
(Staehelin and Hoigne, 1982) 8 to 10 20 
The reaction kinetics of ozone decomposition proposed by Staebelin and co-workers, as shown in Table 3.4. 
are first order with respect to both ozone and hydroxide ion concentration (Staehelin and Hoigne, 1982; 
Staehelin and Hoigne, 1985). The expression derived by Staehelin and co-workers for the steady stale 
concentration of hydroxyl radicals during ozonation of water containing an organic solute B is presented in 
equation 3.6 (Staehelin and Hoigne, 1985). 
[HO'] " 13.61 
where kHO.O, is the rate constant for the reaction between ozone and hydroxide ions, kl the rate constant for 
the initiation reaction between solute B and hydroxide ions. ks the rate constant for the scavenging reaction 
between solute Band hydrox')'l radicals, and (B] the solute concentration (Staehelin and Hoigne, 1985). 
Staehelin and co-workers estimated the steady-state concentration of hydrox-yl radicals to be < 10.9 M based 
on an ozone decomposition rate of 10" M s·\ producing hydroxyl radicals that are consumed by solutes of 
concentration 10-1 M with typical rate constants .. 101 M-I S·I (Staehelin and Hoigne. 1985). Superoxide 
radical anions also exist as short-lived intennediates, the steady-state concentration was estimated to be 
... 10-9 M assuming a formation rate of " 10·' M S· I and an ozone concentration of 10·' M controlling its 
consumption (Staehelin and Hoignc, 1985). 
Chapter 3 OZONE Al'I'D HYDROGEN PEROXIDE CR£MlSTRY 3-11 
The number of hydroxyl radicals formed per mol of ozone decomposed has been estimated to be between 0.5 
and I mol (Hoigne, 1975~ Hoigne and Bader, 1976; Westerboff et al., 1997). The theoretical number of 
hydroxyl radicals formed, according to reactions la] to le] of Scheme 3.3 and the dissociation reaction of 
water, is 2 mol per 3 mol of ozone decomlX)Sed. The overall reaction of hydroxyl radical formation is 
shown in Scheme 3.5 (Westerhoff et al., 1997). 
30 3 + H1 0 - 2HO· + 40 2 
Scbeme 3.S 
The presence of hydrogen peroxide or the application of other advanced oxidation processes during 
ozonation initiates additional radical reactions in solution (BabIon et al., 1991b). Perbydroxyl ions, as 
shown in reaction (c) of Scheme 3.6, are formed from the dissociation of hydrogen peroxide in water. The 
direct reaction between ozone and bydrogen peroxide is slow whereas the reaction between ozone and 
perhydroxyl ions is rapid (Bablon et al. , 1991b; Taube and Bray, 1940). Low concentrations of perhydroxyl 
ions are kinetically effective in initiating ozone decomposition (Bablon et al., 1991b). The radical reactions 
initiated by hydrogen peroxide during ozonation are listed in Scheme 3.6 (Beltnin et al ., 1998; Glaze and 
Kang, 1989a; Glaze and Kang, 1989b). 
Initiation steps 
0 ] + HO- ~ Oi- + HO; 
HOi • Oi- + H+ 
H20 2 """ H02" + H.o-
 0 3 + H02" ~ 0 ; - + HOi 
Propagation steps 
la] 
(b] 
pK - II ,8 Ic] 
k'j "" 2,8 X 106 M""I 5.1 Id] 
0 ) + Oi- ~ 0;- + O 2 kl "" 1,6 X 10' MI 5.1 [e) 
O J- + H+ !; HOj k2"" 5,2 X 1010 M"1 5.1 {f} 
HO; !t HO· + O2 k3"" 1,1 x 1O~ M"1 5.1 (g] 
0 3 + HO· ~ O 2 + HOi k.. "" 2,0 X 10' M"I 5'\ (h) 
HO· + H20 2 ~ 0;- + H20 + H"' k$ = 2,7 X 107 M"l 5'\ [i) 
HO· + HOi ~ 0;- + H1 0 k, "" 7,5 X 10' M""I 5.1 lil 
Termination steps 
HO· + B t~ B~ 
0 3 + B ~ B0"Jti4 
Scheme 3.6 
fkl 
fll 
Cbapter 3 O ZONE AA" HYDROGEN PERoXIDE CIlEMISTRY 3-12 
The termination stepS of Scheme 3.6 are the reaction of a solute B with hydroxyl radicals, reaction (k], and 
the direct reaction of B with ozone, reaction [I]. The reactions listed in Scheme 3.6 are presented in 
Figure 3.5. The reaction numbers of Scheme 3.6 are included in Figure 3.5. 
(c) 
Hp, :oo=f': HO; 
0 , 
(d) 
B 
-0) 
Figure 3.5: Reaction diagram for the ozone decomposition process in tbe presence of hydrogen 
peronde and a solute B (Glaze and Kang. 1989.; Glaze and Kang. 1989b) 
The overall stoichiometry for the reaction between ozone and hydrogen peroxide, as summarised from the 
reactions listed in Scheme 3.6, is shown in Scheme 3.7; a total of 2 mol of ozone are consumed per mol of 
hydrogen peroxide (Beltr.in et 31 ., 1998; von Sonntag, 1996). 
Scheme 3.7 
Hydrogen peroxide is formed during ozonation from reactions such as rh] and (i] of Scheme 3.3, however. it 
exists as a short-lived. intermediate since upon dissociation the perhydroxyl ions formed initiate funher 
ozone decomposition (Hoigne and Bader, 1976; Staehelin and Hoigne, 1982; Staehelin and Hoigne, 1985: 
Westerhoff et al. , 1997). Perhydroxyl ions consumed by ozone are replaced from the dissolved hydrogen 
peroxide by an equilibrium shift in reaction (e] of Scheme 3.6 (Bablon et al., 1991b). 
Hoigne and Bader measured the concentration of hydrogen peroxide during ozonation (with and without 
prior addition of hydrogen peroxide) at different pH levels (Hoigne and Bader, 1976). Hydrogen peroxide 
concentration (for an approximate ozone addition of 2,88 mg L·1) was between 0,34 and 0.51 mg L-t at pH 2 
and below the detection limit (0, 10 mg L- l ) at pH values 6 and 10.5. Hydrogen peroxide concentration at a 
Chapter 3 OZONE AND HYDROGEN PEROXIDE CBEMISTRY 3-13 
pH of 2 was equal to the sum of that added (6,80 mg L-I ) and that formed during ozonation whereas at a pH 
of 10,5 less hydrogen peroxide was present (0,14 mg L-I ) than what was initially added (Hoigne and 
Bader, 1976). Hydrogen peroxide is not observed as a significant intermediate in ozonated water, unless the 
pH is low (below 6), bxause of the high reactivity of perhydroX1'1 ions (Staehelin and Hoigne, 1982). 
Hydrogen peroxide initiated decomposition of ozone, as shown in Scheme 3.6. increases with increasin& pH 
(Bablon et al ., 1991b; Hoigne and Bader, 1976; Staehelin and Hoigne, 1982). The reaction between ozone 
and perhydroxyl radicals, reaction {b] of Scheme 3.6, is classified (according to calculation of the reaction 
time, tR , and diffusion time, tD, for an ozone mass transfer coefficient of 2 x 104 m S-I) as a slow reaction at 
pH 4 for an hydrogen peroxide concentration " 340 mg L- I and as a moderate to fast reaction at pH > 10 for 
a hydrogen peroxide concentration .... 3,4 mg L·1 (Beltnin, 1997). Computer simulation also indicated that 
the reaction between ozone and perbydroX1'1 ions becomes more dominant with increasing pH 
(Westerhoff et al., 1997). 
Staehelin and Hoigne determined the relative contribution of perhydroxyl and hydro"'-'YI ions to the 
decomposition of ozone in aqueous solution (Staehelin and Hoigne, 1982). Perhydroxyl ions had a greater 
effect than hydroxyl ions on ozone decomposition in solutions with a hydrogen peroxide concentration 
greater than 3,4 x 10.3 mg L· I and a pH below 12 (Staehelin and Hoigne, 1982). Beltnio calculated that the 
rate of ozone decomposition due to reaction with perhydroxyl ions in a 0,34 mg L·I hydrogen peroxide 
solution at pH 7 and 12 was approximately 63 and 24 times greater, respectively, than that due to reaction 
with hydroxyl ions (Beltr.ln, 1997). 
3.2 OZONE IN WATER TREATMENT 
Ozone is used in potable water treatment for disinfection, degradation of organic compounds and control of 
taste and odour (Glaze. 1987; Masten and Davies, 1994: Nebel. 1981). Pre~zonation has been found to 
enhance flocculation of suspended particles in surface waters (Glaze, 1987). By-product formation during 
chlorination (trihalomethane compounds are potentially harmful to humans) has increased the application 
of ozone in water treatment (Bowers et aI., 1973; Brink et al ., 1991 : Rachwal et aI. , 1992). The limitations 
associated with ozone are the lack of residual disinfectant capacity and the potential biological grO\\1h that 
can occur in a distribution network (Glaze, 1987). The lack of residual disinfectant is due to the instability 
of ozone and subsequent decomposition to o"'-"ygen. the half-life of ozone at pH 8 is less than 1 h. OLofl:'ltion 
of natural organic matter pnxJ.uces more readily biodegradable compounds that if not removed through 
filtration can lead to biological growth in a distribution network. Improved performance is obtained when 
ozone is combined with other disinfectants that maintain an active residual for longer periods and a method 
of filtration for the removal of biodegradable material (Glaze. 1987). 
Chapter 3 OZONE AND HYDROGEN PEROXIDE CHEMISTRY 3-14 
Ozone is used in the treatment of industrial wastewaters to improve water quality for recycling or discharge 
(Masten and Davies, 1994; Nebel, 1981 ; Rice, 1991). Ozonation can improve the biodegradability of 
non-biodegradable compounds since the oxygenated functional groups introduced during ozonation are 
potential starting points for metalx>lic processes (Nebel, 1981). Compounds associated with odour problems 
contain functional groups possessing a high electron density such as sulfides, antines and olefins. These 
functionaJ groups are oxidised by ozone to produce oxygenated groups that are not odoriferous, for example, 
dimethyl sulfoxide is fonned from the oxidation of dimethyl sulfide (Nebel, 1981). 
The history of ozone use in water treatment is detailed in Section 3.2.1, the equipment for ozone generation 
and transfer in Section 3.2.2 and the applications of ozone in water treatment in Section 3.2.3. 
3.2.1 History of ozone use 
The first full -scale ozonation plants for water treatment were constructed at the turn of the century in 
Westem Europe, the majority were located. in France, a plant was built in Paris in 1898 and Nice in 1906 
(Brink et al., 1991 ; Gomella, 1972). Other early plants were located at Oudsboom, Netherlands (1893), 
Niagara Falls, New York (1903), Wiesiloden, Gennany (1901), and Madrid, Spain (1910), The initial 
expansion in ozone usage \\-'35 halted by World War 1, research conducted during the war into poisonous 
gases led to the development of inexpensive chlorine (Brink et al., 1991). Water treatment was performed 
by chlorination, chloramination and the application of chlorine dioxide (Gomella, 1972). 
The initial application of ozone in water treatment was for disinfection. Other early applications of ozone 
were for taste and odour control (Brink et al. , 1991). Ozonation was generally implemented during the final 
stage of water treatment; during the 19605 applications such as colour removal and oxidation of iron and 
manganese that required ozone addition early during treatment (hence the term pre--ozonation) were 
developed The coagulating effects of ozone and the oxidation of specific micropollutants (such as phenolic 
compounds and various pesticides) were also first exploited during the 1960s (Brink et al., 1991). The 
ability of ozone to control algal growth was utilised in France during the 19705. Further applications of 
ozone include the control of disinfection by-products and the minimisation of microbiological growth 
potential of water by ozonation prior to granular activated carbon filtration (Brink et al.. 1991). 
Ozone application, post World War 11, increased more rapidly in Europe than in America. By 1976. 
approximately I 000 installations world·wide used ozone in water treatment: the treatment plants were 
situated predominantly in Europe, however, 20 were in operation in Canada. The largest (that treated water 
flow rates of up to 910 ML d· l ) supplied drinking water to the City of Quebec (Peleg, 1976). By 1984, the 
number of ozone facilities grew to 20 in the United States and 50 in Canada: the largest at Los Angeles 
treated approximately 2360 ML <fl of water (Glaze, 1987). By 1994. 40 effiuent treatment plants using 
ozone were situated in the United States (Masten and Davies, 1994). Three installations using ozone in 
water treatment were operating in South Africa during 1991 (Brink et al .. 1991). 
Cbapter 3 OZONE AND HYDROGEN PEROXIDE CHEMISTR\:' 3-15 
3.2.2 Ozone generation and transfer 
The instability of ozone requires that it is generated on-site (Brink et al., 1991; Glaze, 1987; RoseD, 1973). 
Ozone is formed according to the reactions listed in Scheme 3.8. Cleavage of the double bond of an oxygen 
molecule produces two short-lived oxygen atoms, these atoms react rapidly with oxygen molecules to 
produce ozone (Brink et al. , 1991 ; Glaze, 1987; Nebel, 1981). 
O2 - o· + o· 
O· + O2 - 0 3 
Scbeme3.8 
lal 
lb] 
Ozone generation in water treatment is perfonned by the corona discharge method (Brink et al., 1991 ; 
Glaze, 1987; Nebel, 1981 ; Rasen. 1973). This method consists of an electrical discharge maintained 
ootween two charged electrodes of a double-tube ozone generator (shown in Figure 3.6). The electrical 
discharge (lasting up to 10 ns) accelerates electrons so that they possess sufficient kinetic energy to, upon 
impact, split the double bond ofan oxygen molecule (Brink et al. , 1991 ; Glaze, 1987). 
electrode cooling water jacket 
:;:~ 
or oxygen 
_'LI __ Jl'lTlLI"LlIZ 
! ,t~1 . < .: • 
Figure 3.6: Double-tube ozone generator (Glaze, 1987) 
air or oxygen 
and ozone 
• 
eleclrOde 
glass tube 
The outer (grounded) electrode. as shown in Figure 3.6. consists of a stainless steel tube surrounded by 
cooling water (Rosen, 1973). The high potential electrode is centred inside the stain1ess steel tube within a 
tubular glass or ceramic dielectric material ; the electrode is usually a conductor coated on the inner surface 
of the dielectric material. The feed gas (dry oxygen or air) passes through the discharge area around the 
sealed glass tube (Rosen, 1973). An alternating current induces electrons entering the discharge area to 
vibrate between the electrodes in accordance with the frequency of the current. The higher the frequency 
the longer the period electrons remain within the discharge area to interact with oxygen molecules 
(NebeJ, 1981). Ozone generators using oX'ygen as a feed gas are designed with' a discharge gap thickness of 
I to 1.5 mm (Brink et aL 1991). Oz.one fonnation is approximately 1,7 to 2,5 times greater in oxygen than 
Chapter 3 OZONE AND HYDROGEN PEROXIDE CHEMISTRY 3-16 
in air; typical concentrations of ozone are 1 to 3 % in air and 3 to 7 % in oxygen (Brink et al. , 1991 ; 
Glaze. 1987). Generators can contain as many as 400 of the double tubes shown in Figure 3.6 and can 
generate 600 kg of ozone per day with dry oxygen as feed gas (Glaze, 1987). The maximum concentration 
of ozone produced in air and oxygen is approximately 4 and 8 %, respectively; reaction of ozone with 
oxygen atoms within a generator prevents concentrations fonning (greater than 20 wt%) at which 
spontaneous detonation may take place (Nebel, 1981). 
Ozone production is an inefficient process, only 5 to 10 % of the applied electrical energy is used in the 
generation of ozone; an insignificant amount is liberated as light enefgy and the bulk is converted to heat 
(Glaze, 1987; Rosen, 1973). Ozone generators have to be cooled to prevent heat build-up and the therma1 
decomposition of ozone. Ozone fonnation is also decreased with increasing moisture and oil content of the 
feed gas. Oxygen and air feed gas must be dry with a low dew point of approximately ·52 to · 58 °c 
(Glaze, 1987; Rosen, 1973). The moisture content in air leads to the formation of nitrous and nitric acids 
that are deposited in the ozone generator and decrease the life of the glass or ceramic dielectric material 
(Brink et al., 1991; Nebel, 1981). Oxygen atoms are not always produced during the impact of electrons 
with oxygen molecules; the addition Of removal of an electron from a molecule produces a negatively (Oi) 
or positively (ot) charged ion (Nebel, 1981). Approximately SO % of electrical power (or electrons) in a 
double-tube ozone generator can be lost (because of conduction) as the electrons move through the field of 
charged ions (Nebel, 1981). 
Ozone production is a direct function of electrical frequency and is doubled when the frequency is doubled 
(Nebel, 1981 ). The frequency at which generators are designed to operate varies between SO and 2000Hz; 
practical limitations such as heat removal and current saturation vary for different ozone generators. The 
rate of ozone production is usually controlled with the voltage applied to the high voltage electrode since 
ozone generation varies exponentially with the voltage (Nebel, 1981). Factors such as the width of the 
corona discharge gap and the thickness of the dielectric material influence the power efficiency of an ozone 
generator. The smaller the corona discharge gap, the greater the power efficiency. The thickness of the 
glass (or ceramic) dielectric material is inversely proportional to the ozone production. The glass electrode, 
however. has to be sufficiently thick to withstand some mechanical impact and the voltage stress under 
which it is placed.. The higher the applied voltage. the thicker the electrode has to be to prevent failure by 
electrical arcing (Nebel, 1981). Although ozone production increases with increasing voltage and 
frequency. the power efficiency of ozone generators does not necessarily increase (Nebel, 198 1). Both ozone 
production and power efficiency vary indirectly with gas pressure. Ozone production decreases whereas 
operating efficiency increases with an increase in flow of the ozone-generating gas (Nebel, 1981 ). 
The solubility of ozone in water is a function of the pressure at which the gas is applied and the temperature 
of the water (Nebel. 1981). Ozone solubility increases with increasing pressure and decreasing water 
temperature. The residual ozone in water decreases with increasing temperature due to the faster rate of 
Cbapter J OWNE AND HYDROGEN PEROXIDE CHEMlSI'RY 3-17 
decomposition and faster reaction rates in the water. The partial pressure of ozone in the gas phase is 
related. to the solubility of ozone in water by Henry's Law. Henry' s Law is defined in equation 3.7 
[3 .7[ 
where C* is the liquid-phase ozone concentration in equilibrium with the gas-phase concentration, H the 
Henry' s law constant for ozone (0,082 aun m3 gmor1 at 25 °C) and Po, the partial pressure of ozone in the 
gas phase (Glaze, 1987). Ozone solubility is expressed in terms of an absorption coefficient (P> or as the 
ratio of the ozone concentration in water to that in the gas (Brink ct al., 1991). 
A number of devices are used to transfer ozone from gas to water. Ozone transfer is increased by 
maximising the surface area between the water and the ozonated gas; smaller b.itX>les are characterised by a 
larger surface area for a given volume of dispersed gas (Nebel. 1981). Bubble size is controlled by the pore 
size of an ozone diffuser. Transfer efficiency also increases with increasing contact time between gas and 
water, the slower the rise time of the bubbles, the greater the contact time (Nebel. 1981). Contact time 
increases \\i.th increased water depth in an ozone contactor (optimal depths vary between 3,7 and 5,5 m). 
Mixing increases the transfer of ozone but also increases the rate of destruction of residual ozone. The 
volume of an ozone contact chamber should also minimise contact between ozone bubbles so as to prevent 
bubble coalescence and the reduction of contact surface area (Nebel, 1981). The speed of ozone transfer is 
usually limited by thc ratio of gas volume to water volume in the contactor (Nebel, 1981). 
Devices used for ozone transfcr include packed and plate columns, static mixers, jet reactors and agitated 
vessels (Brink et aI., 1991; Martin and Galey. 1994; Munter et al .. 1993). One of the most commonly used 
devices (shown in Figure 3.7) is the counter-current bubble column (Glaze. 1987). 
water inlet 
water outlet 
spent gas 
pressurized 
ozone inlet 
Figure 3.7: Counter~urrent bubble column for ozone transfer (Glaze, 1987) 
Challter 3 OZONE AND HYDROGEN PEROXIDE CUEMlsrRY 3-18 
The efficiency of an ozone contactor is important because of the low panial pressure of ozone in the process 
gas. An ozone contactor should maximise mass transfer efficiency, gas/liquid mixing and contact time 
(Schulz and Prendiville, 1993). Both contact time and bubble surface area are determined by pore size in 
dispersion systems using porous disks or rods (Bowers et aI. , 1973). A smaUer pore size leads to an 
increase in bubble surface area but also an increase in fouling and gas pressure drop across the diffuser 
(Bowers et al. , 1973). O.lOne contacting systems using porous disks usually consist of alternating diffusion 
and reaction chambers (Schulz and Prendiville, 1993). Water flows downward in tbe ozone diffusion 
chamber, counter-current to the rising gas bubbles, and upward in the reaction chamber. Multi-chamber 
systems satisfy the ozone demand for both fast- and slow-reacting chemicals while maintaining a required 
ozone residual at the outlet of the reaction chamber for disinfection. Mass transfer efficiency is determined 
by gas-phase ozone concentration, the ratio of gas flow rate to liquid flow rate, water depth and bubble size 
(Schulz and Prendiville, 1993). Transfer efficiency may be reduced at higher gas flow rates due to bubble 
coalescence and at very low gas flow rates due to channeWng of gas bubbles. The actual contact time in 
counter-current ozone diffusion contactors is approximately 50 % of the theoretical hydrauJic retention time 
(Schuiz and Prendiville, 1993). 
Static mixers (shown in Figure 3.8) are systems using fixed internal nrixing elements to rapidly transfer 
ozone into water (Martin and Galey, 1994). Mixing occurs due to the division of the main water flow by the 
mixing elements into partial flows that are continually recombined and again divided; O'L.one loss occurs 
because of the intense mixing. Static mixers require low maintenance and are robust due to the lack of 
moving parts (Martin and Galey, 1994). 
Figure 3.8: A static miIer (Martin and Galey, 1994) 
Cbapter 3 OZONE AND HYDROGEN hRoXIDE CIIEMlSTRY 3-19 
A pilot-scale study (8 to 12 m3 b-I ) at the Mery-sur-Oise water treatmenl plant showed that ozone could be 
dissolved in water within IS s with the use of a static mixer. Ozone transfer increased with decreasing 
water flow rate and increasing gas flow rate (Martin and Galey, 1994). 
lnline injection contactors either introduce ozone directly into the mainstream of water flow or into a side 
stream that has been diverted and pressurised by a _ pump (Bowers et al., 1973; Schulz and 
Prendiville, 1993). Static mixing elements are used in a side stream 10 generate fine gas OObbles that 
enhance mass transfer and mixing of the ozone in the side stream back into the main stream of water. 
Advantages of side stream injection contactors over diffusion oontactors include operation at a higher ratio 
of gas flow rate to liquid flow rate and thorough mixing without bubble cbannelling problems (Schulz and 
PrendiviUe, 1993). Injection contactors are inexpensive but are subject to bumping and corrosion 
(Brink et al., 1991). Contact time is minima l unless a secondary reactor is installed. downstream for the 
oxidation of slow-reacting chemicals and for settling and removal of oxidised precipitates. lnline injectors 
are not designed for the disinfection of water because of the difficulty in maintaining a minimwn ozone 
residual for a specific contact time in such a secondary reactor (Schulz and Prendiville, 1993). A twcrstage 
ozone water treatment plant in Henrico County, Virginia (USA) uses inline injection for pre-ozonation and 
fine-bubble diffusers for intermediate ozonation. Inline injection was used in the ~onation stage in 
order to avoid maintenance problems associated with plugging of fine-bubble diffusers by suspended solids 
and oxidised precipitates in the raw water (Schulz and PrendiviUe, 1993). 
3.2.3 Application 
Ozone is used in water treatment for disinfection., oolour removal, taste and odour removal, oxidation of 
organic compounds and for the enhancement offlocculation (Koga et al. , 1992). Trihalomethane formation 
is also lowered by the use of ozone instead of chlorine. Ozone has traditionally been used for water 
treatment in only North America and Europe; facilities are being installed. in other countries, such as that in 
Nizhny Novgorod, Russia, that treats 280 ML crI of water (Brink et al. , 1991 ; Haglund, 1997). 
The application of ozone in water potable treatment is reviewed in Section 3.2.3. 1 and in industrial effiuent 
treatment in Section 3.2.3.2. 
3.2.3.1 Potable ll'ater treatment 
O.lone is used in potable water treatment as a biocide, an oxidant and a pretreatment to improve the 
performance of processes such as coagulation (Martin and Galey, 1994; Nebel, 1981 ; Schulz and 
PrendiviUe. 1993). Ozone disinfection is limited. by the lack of residual disinfectant capacity and the 
potential for biological regrowth in a distribution network. Ozone disinfection is thus usually combined 
with another disinfectant to maintain a longer active residual and a method of filtration for removal of 
biodegradable material (Glaze, 1987). The two-stage ozone water treatment plant in Henrico County, 
Virginia (USA) included ozone disinfection ahead of granular activated carbon filters so as to maintain 
Cb.pt~r 3 OZONE AND HYDROCl.N PDtoXlD£ CIRMlSTRY 3-20 
biologica1 growth on the filters; filter bed. life was extended by a factor of between 4 and 10 before 
reactivation was required (Schulz and Prendiville, 1993). 
Ozone is a strong biocide with disinfection occurring due to cell wall rupture (Nebel, 1981). Inactivation of 
bacteria by ozone varies from that of Escherichia coli. the most sensitive to i.nactivaIion, to bacteria such as 
the Gram-positive cocci (Staphylococcus and StreptOCOCClJS), the Gram-positive bacillae (Bacillus) and 
Mycobacteria that are more resistant (BabIon et al., 1991a). Viruses are typically more resistant to ozone 
inactivation than vegetative bacteria but not more so than the sporular forms of Mycobacleria. The 
resistance of parasites (protozoan cysts such as Giardia lamb/ia) to ozone inactivation is greater than that of 
viruses (BabIOD et al., 1991a). Ozone is more effective than chlorine for the removal of Escherichia CQli , 
Giardia, viruses and certain forms of algae (Glaze, 1987; Nebel, 1981). Disinfection with a static mixer to 
enhance ozone transfer achieved the required I-log reduction in colony forming units for Giardia lamblia 
and 3-log reduction for viruses (required by the US Environmental Protection AfJ:.Dcy) with a 90 s contact 
time and 2 mg L·! ozone transfer (Martin and Gaie:y, 1994). Disinfection under idea1laboratory conditions 
occurs when an ozone residual of 0,4 mg L'] is maintained for a 4 min contact time (Martin and 
Galey, 1994). Ozone dosage required for disinfection is dependant on pH and water temperature; the decay 
rate of ozone increases with increasing pH and water temperature (Meijers et al., 1995). The rate of ozone 
disinfection was improved with the simultaneous application of ultrasound rut unaffected by the addition of 
hydrogen peroxide (Dahi, 1976; Martin and GaJe:y, 1994). Dahi found that the rate of ozone decomposition 
was significantly greater during sonication than without sonication, k = 0,207 min'] versus k "" 0,032 min· 1 
(Darn. 1976). Ozone disinfection has also been used in the recycling of marine aquaria wastewaters such as 
that at Sea World in Orlando, Florida (Rice, 1997). 
Taste and odour are associated with various types of organic com~ds including polysulfides, aldehydes 
and alicyclic alcohols such as geosmin ( 1 , 10-".ans~ethyl-".ans-<9)~ol) and 2-methylisobomeol 
(Bablon et al .. 1991a; Glaze et al ., 1990). Taste- and odour-producing compounds are present in raw water 
or are produced during water treatment itself. Such compounds are formed from the decomposition of plant 
matter and by the activity of organisms such as actinomycetes and blue-green algae and occasionally by 
organisms such as zooplanklon and nematodes (Bablon et al. , 199Ia). Taste and odour problems occur 
more frequently when raw water is distributed from impoundments, especially in summer when water 
demand is high and water levels are low. Taste and odour problems may be of industrial origin such as 
from chlorinated solvents. hydrocarbons and pesticides (BabIon et al .. 1991a). Formation of taste and odour 
compounds may take place in a distribution net\\'ork if conditions lead to the growth of nticroorganisrns or if 
products used for the lining of pipes and reservoirs contain chemicals that are released into the water 
(BabloD et al ., 1991a). 
Chapter 3 OWNE AND HVDllOGf.N hRoXIDE CIII'.MISTRV 3-21 
Glaze and co--workers evaluated various oxidising agents for the removal of six model taste and odour 
compounds from a municipal water supply (Glaze et al., 1990). The model compounds were geosmin, 
2-metbylisobomeol, 2,4-decadienal, dimethyltrisulfide, I-beptanal and l-hexanaL Effective oxidation of 
geosmin and 2-methylisoborneol (100 ng L· I concentration) by potassium permanganate, chlorine dioxide, 
chlorine, chloramine or hydrogen peroxide did not take place; percentage removals were less than 31 % for 
a 3 mg L·I oxidant c:k>se and a 120 min contact time. Ozone oxidation of geosmin and 2-methylisoborneol 
were 35 and 40 %, respectively. for a 2 mg L·I ozone dose and a 20 min contact time; oxidation increased to 
86 and 73 0/0, respectively, for a 4 mg L·1 ozone dose. The simultaneous 3A)lication of hydrogen peroxide 
(2 mg L·I) and ozone (4 mg Lol) over 20 min lead to 95 and 89 % oxidation, respectively. of geosmin and 
2-methylisoborneol (Glaze et al., 1990). The five-year pilot-plant investigation by Weng and co-workers at 
Hackensack Water Company, New Jersey (USA) found that ozone transferred the musty or grcwy odour of 
the raw water into a sweet odour in the pilOl-plant finished water (Weng et al., 1986). 
Ozone is used as an oxidant in water treatment to oxidise synthetic organic compounds and compounds 
produced by natural processes (BabIon et al .• 1991a). Naturally occurring organic matter such as humic 
acids, fulvic acids and proteins are characterised by non-specific measurements such as chemical oxygen 
demand. UV absorbance {at 254 run} and tota1 organic carbon. Synthetic organic chemicals in raw water 
(such as biocides, hydrocarbons, phenols, plasticizers, dyes, amines and solvents) are often toxic; these 
poUUtaDts are present in concentrations in the I ng L-1 to I mg L-1 range (BabIon et al. , 199Ia). 
Pesticides are regularly detected in the raw water of potable water treatment plants (Kang et al. , 1997; 
Lambert et al.. 1996; Xiong and Grahant, 1992). Koga and co-workers detected 18 pesticides in river water 
used as a water source for the City of Kitakyushu in China; 11 of the pesticides were still present after 
treatment with pre-chlorination, Oocculation, sedimentation, rapid sand filtration and post<hlorination 
(Koga et al. . 1992). Triazines (such as atrazine and simazine) and phenoxyalkyl acid derivatives (such as 
Mecoprop) are two groups of herbicides detected in water supplies in England and Wales at concentrations 
above 0,1 ~g L·1, the maximum admissible concentration stipulated by the EC Drinking Water Directive 
(Xiong and Graham, 1992). Ozone is used as an oxidant for the degradation of pesticides that arc not 
readily degrnded in conventional water treatment processes. Kang and co-workers found that the 
concentration of pesticides in most raw waters was reduced by 90 % during ozonation (Kang et al., 1997). 
Triazine pesticides are more resistant to oxidation by ozone than phenolic pesticides and Olner 
organonitrogen pesticides (Masten and Davies. 1994). The complete degradation of chlorophenoxy acid 
herbicides required 4 to 5 moles of ozone per mole of solutc while atrazine required 25 moles of ozone 
(Xiong and Graham, 1992). Pesticide degradation by ozone is dependant on ozone dose and contact time: 
atrazine degradation increased from 45 to 65 % with an increase in ozonc dose from 2 to 5 mg LoL for a 
10 min contact time, and increased from 60 to 90 % for an increase in contact time from 10 to 60 min for a 
Cbapter 3 OZOl'll'"'£ AND HYDROGEN PEROXIDE CHEMISTRY 3-22 
3 mg LO\ ozone dose (Roche, 1994; Roche and Prados, 1995). Pesticide degradation is also influenced by 
pH and temperature; the degradation of diuron by ozone increased with increasing pH, in the range 7,2 to 
8,3, and increasing temperature, 5 to 20 QC (Meijers et al ., 1995). 
Ozone is a selective oxidant and the degradation of pesticides is hindered by oxidation of compounds that 
have higher oxidation rates or are present in greater quantities, for example, natural organic matter and 
bicartxmate ions (Orlandini et al., 1996; Roche and Prados, 1995). Pesticide degradation is enhanced when 
ozone is used in combination with hydrogen peroxide or ultraviolet light (Kang et al., 1997; Koga et al ., 
1992; Prados et al. , 1995). Atrazine degradation (1 ,42 f.Lg L-\ initial concentration) for a 10 miD con.tact 
time was 40, 65 and 70 0/0. respectively. for treatment with ozone, ozone with ultraviolet light and ozone 
with hydrogen peroxide where the ozone dose was 3 mg Lot and the hydrogen peroxide: to ozone mass ratio 
was 0,4 (Prados et al. , 1995). Roche and co-workers found that atrazine degradation (835 ng LO\ initial 
concentration) for a 10 min contact time was 40 and 90 0/0, respectively. for ozone and ozone with hydrogen 
peroxide for a 3 mg LO\ ozone oose and a hydrogen peroxide 10 ozone mass ratio of 0,4 (Roche. 1994). 
Lambert and co-workers investigated the degradation oftive pesticides for ozone doses of 1; 2 and 3 mg L'\ 
and mass ratios of hydrogen peroxide to ozone of between 0 and I for each of the ozone doses 
(Lambert et al ., 1996). Pesticide degradation increased with increasing ozone consumption and applied 
hydrogen peroxide concentration. Ozone transfer was independent of hydrogen peroxide concentration. 
Hydrogen peroxide was not totally consumed during ozonation and the residual increased with the 
increasing applied concentration (Lambert et al., 1996). 
Aromatic compounds (such as the chlorinated derivatives of phenol and benzene) have been detected in 
potable water supplies (Boncz et al .. 1997: Masten et al .. 1997: Trapioo et al., 1997). CWorinated phenols 
are oxidised effectively by ozone at a high pH~ the toxicity of the degradation products, as tested by Daphnia 
magna 24 h tests. was less that that of the parent compounds (Boncz et al ., 1997; Trapido et ai. , 1997). 
Fewer disinfection by-products that may induce the formation of trihalomethane compounds were produced 
during the ozonation of phenolic and benzene compounds than during cWorination (Kjm et al .. 1997). The 
degradation of p-nitrophenol increased with increasing pH; the kinetic regime of the reaction between ozone 
and p-nitrophenol changed. from slow to fast to instantaneous as the pH increased from 2 to 6.5 to greater 
than 8,4 (BeltrAn et a1.. 1992). Degradation of 1,3,5-trichlorobenzene by ozone and ozone with hydrogen 
peroxide was greatest in the pH range 7 to 8,5 (Masten et al .. 1997). Hydrogen peroxide and ultraviolet 
light enhanced the ozone oxidation of volatile organic compounds such as tricWoroethylenc and carbon 
tetrachloride (Bellamy et al., 1991: Peyton et al., 1982). 
Oxidation of organic compounds by ozone in potable water treatment is also enhanced with the addition of a 
catalyst or with the simultaneous application of ultrasound (Olson and Baibiero 1994: Yolk et al .. 1997). 
Ozone. ozone with hycirogeb peroxide and ozone with titanium dioxide catalyst mineralised 15, 18 and 
Cbapter 3 OZONE AND HYDlt<XiEN hRoXID£ CBDUSTIlY 
24 %, respectively, of the initial dissolved organic carbon (2,84 mg L- I ) of a synthetic solution of fulvic 
acids (yolk et al., 1997). The rate of removal of toeal organic carbon (TOC) in a 10 mg L- I fuJvic acid 
solution was increased from 40 to 91 % with the simultaneous application of ultrasound with ozone 
(60 min; power, 55 W; frequency, 20 kHz); the amount of original carbon removed as carbon dioxide 
increased from 28 to 87 %, respectively (Olson and Barbier, 1994). Ozonation may increase the 
biodegradability of non-biodegradable material (Nebel, 1981). The partia1 oxidation of naturally occurring 
organic matter (humic and fulvic acids) produce com.pounds such as aldehydes and ketones that are more 
biodegradable than tile parent compounds (Foster et al., 1992). Van Leeuwen demonstrated tile improved 
biodegradability of South African water sources and industrial effluents sucb as Rietvlei Dam water, a Sasol 
effluent and industrially-poUuted Ngagane River water, ·the water samples were inoculated with 
Pseudomonas fluorescens and the growth of the microorganisms was studied by measuring adenosine 
triphosphate concentration and UV absorption at 400 nm (van Leeuwen, 1987). The formation of 
trihaJomethane compounds during chlorination of naturally occurring organic matter is reduced with 
pre-ozonation (Racbwal et al., 1992). Simultaneous application of hydrogen peroxide or ultraviolet light 
with ozone enhances the removal oftribalomethane precursors (Duguet et al., 1985; Glaze et al. , 1982). 
3.2.3.2 Industrial effluent treatment 
Ozone is a strong oxidising agent and is thus used in the treatment of various industrial effiuents. Effluents 
from the textile industIy are highly coloured, contain dyes and ancillary chemicals such as sizing agents and 
surfactants (Perk:owsld et al. , 1996; Rice, 1997). Textile effluents are toxic and tend to be non 
biodegradable due to the complex chemical structure of organic compounds in the effluent. The source of 
colour, biological oxygen demand (BOO) and chemical oxygen demand (COO) changes since textile plants 
use different organic dyes depending upon the season (Rice, 1997). Ozone decolourises textile effluent, 
however. the large organic load (represented by high COD and total organic carbon (TOC) residues) are not 
completely converted to carbon dioxide and water. Treatment of textile effluent is performed more 
effectively by a combined method of ozone pretreatment and subsequent biodegradation; the removal of 
COD and non-ionic surfactants by biodegradation increased, respectively, from 10 and 9 % without 
pre~zonation to 65 and 87 % with pre-ozonation (Perkowsld et a1., 19%). Textile efi1uent treatment with 
ozone is practised in Japan, Germany, Italy, Taiwan, the USA, the UK and Spain (Rice, 1997). 
Ruppert and Bauer found that a red-coloured wastewater with an initial TOe of 275 mg L'] and pH of 10,06 
was completely decolourised by ozone within 20 min (Ruppert et al., 1994). The TOC reduction was 20 % 
and pH decreased from 10,06 to 8,08. The first step in colour removal, the cleavage of double bonds in 
aromatic rings or azo groups, foUowed by hydroxylation, did not lead to the elimination of carbon resulting 
in a low reduction in TOC (Ruppen et al. , 1994). Liakou and co-workers used Orange U (Cl 15510), 
4-{2-hydroxy-l-naphthylazobenzenesulfonic acid), as a model dye to demonstrate ozone decolourisation 
(Uakou et al., 1997). Oxalate, formate and benzenesulfonate were identified as intermediate compounds of 
Orange U oxidation. The chemical structure of Orange U is shown in Figure 3.9. 
Cbapter 3 OZONE AJII'l> HYDROGEN Pr.aoXIDI: Cu:uosrRv 
HO 
Na03~N=N 
Figure 3.9: C ...... icaJ 5Inoct1ore of 0 .... n dye (Liakoo et aL, 1997) 
3-24 
Ozonation can be used to convert cyanide ions into the less toxic cyanate ions in electroplating efDuents 
(Rice, 1997). Chlorination is less expensive than ozonation but cyanide ion conversion is also lower. An 
ozone treatment facility was installed at the Cactillac Motor Car Division of General Motors Corporation in 
Oeuoit. Michigan, in 1978; the ozone to cyanide ratio ranged. from 2 to 3, ozone utilisation was greater than 
99 % and total cyanide levels bave been consistently reduced to below 1 mg L-t (Ric::e, 1997). 
Emuents from petroleum refineries are saturated with hydrocarbon compounds. Ozonation followed by 
granular activated carbon (GAC) adsorption bas been used to reduce concentrations of petroleum«rived 
organic material in efiluent at five USA petroleum product storage and transpOrtation facilities to be10w 
0, I mg L· t (Rice, 1997). Pre-ozonation increasc:d. the lifetime of the GAC adsorbent by between 10 and 
50 %. Ozone is used in the treatment of effiuent {predominantly phenolic compounds} from the oil shale 
industry in Estonia; approximately 2 000 tons of efI1uent is produced annually (Rice, 1997). 
O.lOne has also been used to degrade protocatechuic acid (a phenolic compound) in the effiuent from an 
olive oil manufacturing plant (Benitez et al., 1996) and for colour removaJ in the effluent from a molasses 
processing plant (Gebringer et al ., 1997). 
3.3 HYDROGEN PEROXIDE 
Hydrogen peroxide was first made by J.L. Tbenard in 181 8; relatively pure hydrogen peroxide was produced 
from the acidification of barium peroxide and the removal of the barium precipitate that formed 
(Greenwood and Earnshaw, 1984<:: Manly, 1962). A hydrogen peroxide molecule consists of two hydrogen 
and two oxygen atoms covalently bonded in a non-polar H-O-O-H structure (Kirchener, 1981 : Pakiari and 
Linnctt, 1975). The molecular structure of hydrogen peroxide is shown in Figure 3. 10. 
H , 
0 - 0 
'H 
Figure 3.10: Molecular structure of bydrogen peroxide (Greenwood and Earnsbaw, 1984c) 
Cbapter 3 OZONE AND HYDROGEN Pr.itoXIDt: Cllt:MlSTRy 3-25 
The bond lengths and bond angles of liquid and crystalline hydrogen peroxide are different to the 
dimensions of gaseous hydrogen peroxide because of hydrogen bonding between neighbouring molecules. 
The OD bond length of crystalline hydrogen peroxide (at -163°C) is 145,8 pm (145,8 x 10·" m); the ().H 
bond length is 98,8 pm~ the OOH angle is 101,90 and the dihedral angle between the planesofO-H bonds is 
90,20 (Busing and Levy, 1963 ~ Greenwood and Eamshaw, 1984c). The 0.0 bond length of gaseous 
hydrogen peroxide is 147,5 pm~ the O-H bond length is 95,0 pm; the ()()H angle is 94,80 and the dihedral 
angle between the planes of O-H bonds is 1I1,5° (Busing and Levy, 1963~ Greenwood and Eamshaw, 
1984<:). Hydrogen peroxide is the smallest molecule that shows hindered rotation about a single bond, the 
rotational barriers for the trans and cis oonformations are 4,62 and 29,45 kJ mor" respectively (Giguere 
and Srinivasan, 1977; Greenwood and Eamshaw, 1984c). 
Hydrogen peroxide is manufactured commercially from the auto-oxidation of an anthraquinone comp:>und 
(Greenwood and Eamshaw, 1984c~ Kircbener, 1981 ; Manly, 1962). The process developed by 
LG. Farbenindustrie iD Gennany during World War n (shown in Figure 3.11) consists of dissolving 
2~ylanthraquinone in a mixed est.erlhydrocarbon or alcohoVhydrocarbon solvent and reducing it to the 
corresponding quinol by a Raney nickcl or supported palladium catalyst. The quinal is oxidised 
non-catalytically in a stream of air to produce hydrogen peroxide and 2-ethylanthraquinone. The catalyst is 
separnted and the hydrogen peroxide e>."tracted by water and ooncentrated to 30 ~/o by distillation under 
reduced pressure. Higher concentrations of hydrogen peroxide are produced. by further distillation 
(Greenwood. and Eamshaw, 1984<:; Kirchener, 1981). 
0 OH 
CH2CH, CH2CH, 
+ H2 catalyst .. 
o ~ y OH separate 0 
and redlssolve 
CH2CH, 
+ H20 2 
extract and 
• 
(-1 % cone) concentrate 
2-ethylanthraquinone 0 
Figure 3.11: Commercial production of hydrogen perol.ide via tbe auto-oxidation of an 
anthraquinone compound (Greenwood and Earnshaw, 1984c) 
Hydrogen peroxide is miscible in water in all prop:>rtions and is sold in concentrations from 3 to 98 % 
(Kirchener, 1981). Hydrogen peroxide is used industrially as a bleach for textiles and paper, in the 
Chapter 3 OZONE AND HYDROGEN hRoXIDE CIIEMlSTRY 3-26 
manufacture of chemicals, for pollution control, in hydrometa1lurgy and for minor purposes such as in 
per.;onaJ care products, food processing and furniture manufacture (Greenwood and Eamshaw, 19840). 
Chemical reagents are added. to stabilize commercial hydrogen peroxide solutions. Stabilizers deactivate 
catalytic impurities by complex fonnation or by adsorption (Kirchener, 1981). Sod:ium pyropbospba1e acts 
in acidic solutions as a complexing agent whereas sodium stannate forms a protective colloid Alkaline 
solutions can be stabilized with alkali metal silicates. Various stabilizing systems have been developed, 
combinations of tin salts and phosphates are frequently used. Some stabilizers contain organic compounds 
that are effective in dilute solutions but cannot be used in concentrated solutions. Tbe required quantity of 
stabilizer decreases with increasing peroxide concentration. Container corrosion is also a source of catalytic 
impurities, nitrate salts are frequently added to inhibit corrosion (Kirchener, 1981). 
The chemical and physical properties of hydrogen peroxide are reviewed in Section 3.3.1 and the 
application of hydrogen peroxide in water treatment in Section 3.3.2. 
3.3.1 Chemical and physical properties 
Some physical propenies of hydrogen peroxide are listed in Table 3.5. 
Table 3.5: Pbysical properties of hydrogen peroxide (Greenwood and Earnshaw, 1984c; 
Kircbener, 1981) 
Parameter 
molecular weight 
molecular formula 
form 
melting point 
boiling point 
density (liquid) 
density (solid) 
viscosity 
heat of dissociation 
Value 
34,016 
Hp, 
colourless liquid at room temperature 
-O,43 °C 
150,2 °c 
1 442 kg m') at 25 °c 
I 643 kg m-3 at -4,5 °c 
1,249 cpa120 °C 
34,3 kJ mor l 
Hydrogen peroxide is a weak acid and may react directly or after dissociation into free radicals. Free 
radicals, shown in Scheme 3.9. are formed from the homolytic cleavage of either an D-H or 0-0 bond 
(Bablon et al., 199 1b; Kirchener, 198 1; Manly, 1962). 
H20 2 ... H" + HO; M/= 380 kJ mor l la] 
H101 ... 2HO· [hI 
Scheme 3.9 
Cbapter 3 OZONE AND HYDROGEN PuoXID~ CuDusntv >-27 
The formation ofbydroxyl radicals, reaction [b) of Scheme 3.9, is the dominant reaction during uncatalysed 
decomposition and pbotocbemicalJy initiated reactions (Kirchener, 1981). Hydrogen peroxide reactions are 
listed in Scbeme 3.10. 
decomposition 
molecular addition 
substitution 
oridation 
reduction 
2H 20 2 ..... 2H 20 + O2 
H20 2 + Y ..... Y.H 20 2 
H, O, + RX - ROOH + HX 
H,O, + 2RX - ROOR + 2HX 
H20 2 + W ..... WO + H20 
H20 2 + Z ..... ZH2 + O2 
Scheme 3.10 
la) 
[b) 
Ic) 
Id) 
le) 
In 
Hydrogen peroxide is stable in the absence of catalysts, however, decomposition is strongly catalysed by 
metal surfaces (metal ions or metal oxides), MnOz or traces of alkali dissolved from glass (Kirchener, 1981 ; 
Koubek et aL, 1963). Peroxyhyd:rates are formed from molecular addition reactions with hydrogen 
peroxide; peroxyhydrates such as that formed from sodium carbonate or urea are oom.mercially available. 
Peroxygen comJXlUllds prepared by substitution reactions of hydrogen peroxide with organic reagents are 
used commercially as polymerisation catalysts or as oxidising agents for a number of special reactions. 
Hydrogen peroxide is a strong oxidising agent and is capabie of oxidising a wide variety of organic and 
inorganic compounds. Reduction reactions occur when hydrogen peroxide reacts with a stronger oxidising 
agent than itself, for example chlorine, sodium hypochlorite, potassium permanganate or eeric suifate 
(Kircbener, 1981 ). 
3.3.2 Application in water treatment 
Hydrogen peroxide is applied in water treaunent because it is a strong oxidant. Hydroxyl radicals, fonned 
from the decomposition of hydrogen peroxide, react with organic matter by either adding to a C=C double 
oond, as shown in reaction fa] of Scheme 3. 11, or by abstracting a carbon-bound H atom, as shown in 
reaction [bl of Scheme 3. 11 (Clarke and Knowles, 1982: von Sonmag et al., 1997). Radical cations are 
formed from molecules containing heteroatoms such as nitrogen or sulfur. 
HO· + R,C=CR, - R,(HO)C-CRi 
HO· + RH - H,O + R" 
Scheme 3.11 
la) 
[bl 
The addition to a double bond, reaction la] of Scheme 3. 11 , is generally faster than hydrogen abstraction, 
reaction [b) of Scheme 3.11. Hydroxyl radicals react non-selectiveiy with most organic and inorganic 
Chapter 3 OZONE AND HYDROGEN Pr.RoXIDI: ClD.MISTlty 3-28 
solutes with reaction rates (typically of the order of 10' sot) that approach diffusion-<antrolled limits 
(Hoigne, 1997; von Sonntag et al., 1997). Hydroxyl radicals are scavenged in water by bicarbonate and 
carbonate ions (von Sonntag et al., 1997). 
Hydrogen peroxide is used in both municipal and industrial effluent treatment. Hydrogen peroxide is used 
in municipal effluent treatment for the abatement of odours (such as hydrogen sulfide) and as an oxygen 
source (Cole et al., 1974~ Eul et al., 1991). Hydrogen sulfide is formed from bacterial action and leads to 
the corrosion of sewer systems as well as odour problems. A supplemental oxygen source is required when 
overloaded activated sludge aeration systems, denitrification in secondary clarifiers and stale or septic 
wastewaters result in low dissolved oxygen concentrations (Ayting and Castran1as, 1981 ; Cole et aL, 1974). 
Hydrogen peroxide is used in industrial effluent treatment for the detoxification of cyanide, nitrite and 
hypochlorite; for the destruction of phenol aromatics and formaldehyde and for the removal of sulfite, 
t1Uosu1fate and sulfide compounds (Bull and Zef!; 1991 ; Eul et al., 1991). Hydrogen peroxide is also used 
as a pI"C-oxidation technology to enhance biodegradation of persistent and roc:alcitrant organic compounds 
such as trichloroethylene and pentachlorophenol (CaJberry and Benzing. 1991 ; Maroo et al., 1997). 
E nvironmental disasters such as accidental chemical spills and offensive odours can be treated with 
hydrogen peroxide. A broken tanker spilt 900 ton of fennented molasses into a marine dock along the River 
Mersey in England. contaminating approximately 3 million m ) of water. 1be water turned black and began 
to evolve hydrogen sulfide as the dissolved oxygen was depleted and a sulfate-reducing bacteria began to 
grow. Hydrogen peroxide was dosed in a 1,5 to I ratio of hydrogen peroxide to hydrogen sulfide. The 
water was mixed with tug boats and nonnal boating and recreational activities were allowed to resume after 
48 h (Robinson and. Monsen, 199 1). An irKlustriallanctfill operation in England experienced problems with 
offensive odours emanating from four leac:hate lagoons. Drums of hydrogen peroxide (35 %) were placed 
every 20 m along the upwind bank of the lagoons and. allowed to drip at a rate of 100 mL min·' into the 
water. Hydrogen peroxick: was dosed in a 3 to I ratio of hydrogen peroxide to hydrogen sulfide and was 
followed by jetting for aeration and. mixing. The concentration of hydrogen sulfide in the lagoons (3 mg L· t 
in two lagoons and greater than to mg L" in the other lagoons) was reduced to below the limits of detection 
after treatment with hydrogen peroxide (Robinson and. Monsen. 1991 ). 
Hydrogen peroxide is used in conjunction with advanced oxidation technologies such as in UVIH~; 
o,tH,o, and UVIO,lH,o, to enhance the degradation of stable organic compounds (Bull and Zeff, 1991 ; 
Clarke and Knowlcs, 1982; Glaze et al., 1987). Compounds typically destroyed are highly substituted 
aromatics. chlorinated aromatics, chlorinated alipbatics, furans and dioxim (Eul et al., 1991 ). 
Chapter 3 OZONE AND HYDROGEN hRoXlDE CIRMlSTRY 
3.4 CONCLUDING REMARKS 
Ozone is used in water treatment for disinfection, colour removal,. taste and odour removal, oxidation of 
organic comJXlUllds and for the enhancement of Oocculation. Ozone is a power 1:ut selective oxidant. 
Oxidation of compounds takes place through the direct reaction with ozone molecules but also via reaction 
with hydroxyl radicals formed from the decomposition of ozone. Hydroxyl radicals are not only a stronger 
oxidant than ozone but are also less selective. hydroxyl radicals react with most organic and many inorganic 
solutes with high rate constants that approach diffusion~ntroUed limits. 
Ozonation is a gas-liquid reaction and is dependant upon the diffusion of ozone into solution. The kinetic 
regime of ozone absorption is calculated from the Hatta number. Dissolved ozone concentration in the bulk 
solution is negligible when all the ozone is consumed in fast reactions (Ha > 3) in the film layer. 
Converrely, slow reactions (Ha < O,02) take place in the bulk solution and ozone mass transfer maintains 
the dissolved ozone concentration in solution close to saturation. 
The decomposition of ozone in solution is initiated by the reaction with hydroxide ions, decomposition is 
thus pH dependant and increases with increasing pH. Oxidation of organic compounds is dominated by the 
direct reaction with ozone under acidic conditions and b)' the reaction with hydroxyl radicals under alkaline 
conditions. Hydrogen peroxide is formed from the free radical reactions initiated by the decomposition of 
ozone, however, it is not produced as a significant intermediate because of the fast reaction of perhydroxyl 
ions with ozone. Perhydroxyl ions are formed from the dissociation of hydrogen peroxide. 
The oxidation of organic colIlJXlUllds during ozonation is enhanced during the combination with 
technologies such as hydrogen peroxide, UV radiation and ultrasound that enhance the decomposition of 
ozone. Ozone decomposition is enhanced during the combination with hydrogen peroxide or ultrasound by 
the reaction with perhydroxyl ions. The rate constant for the reaction of ozone with perhydroxyl ions is a 
factor of more than I Q4_fold greater than the rate constant for the reaction of ozone with hydro,,"),1 ions. 
Ozone disinfection is improved when ozone is combined with other disinfectants that maintain an active 
residual for longer periods and with a method of filtration for the removal of biodegradable material. 
Ozonation of natural organic matter produces more readily biodegradable compounds that if not removed 
through filtration can lead to biological growth in a distribution network. The biodegradability of 
non-biodegradable compounds is improved during ozonation since the oxygenated functional groups 
in~ced during ozonation are potential starting points for metabolic processes. 
4 
ATRAZINE CHEMISTRY 
Atrazine is a herbicide that is used world-wide, it is one of the most commonly used herbicides in the 
United States. Runoff into surface waters and seepage into groundwater has resulted in the contamjnation 
of potable water resources. Atrazine is the most frequently detected herbicide in groundwater in countries 
of the European Community. Concern for the environmental impact and potential effects it may have on 
humans has resulted in the use of atrazine being banned or regulated in many countries. Water treatment 
processes to degrade atrazine are being investigated by various organisations and university research 
groups. 
The general chemistry, physical properties and toxicity data for atrazine are presented in Section 4.1. The 
role of atrazine in soil and in aquatic systems in the environment are reviewed in Section 4.2. Atrazine 
degradation kinetics and different methods used to degrade atrazine in water treaUnent are reviewed in 
Section 4. 3. 
4.1 CHARACTERISTICS OF ATRAZINE 
The Chemical Abstracts name of atrazine is 6-chloro-N-ethyl-N' -(1-methylethyl)-I ,3,5-triazine-2,4-diamine 
and the IUPAC name is 6-chloro-N2-ethyl-,v-isopropyl.l ,3,5.triazine.2,4-diamine (fomlin, 1997). It is 
also referred to in earlier literature as 2-<:h10r0-4.ethylantinQ-6.isopropylamino--s·triazine. Atrazine is a 
six·membered heterocyclic ring, the chemical structure of atra2.ine is shown in Figure 4.1. 
Cl 
NJ--N 
H3CH2CHN~N)lNHCH(CH3), 
Figure 4.1 : Chemical structure of atrazine (Tomlio, 1997) 
The discovery of atrazine is reviewed in Section 4.1.1, chemical and physical properties in Section 4.1 .2 and 
toxicity data in Section 4.1.3. 
Chapter 4 A TRAZIN!: CHEMISTRY 4-2 
4.1.1 Discovery 
Atrazine was developed by the Geigy Chemical Company in Basel. Switzerland, from research begun in 
1952 into derivatives of symmetrical triazines. It was patented in Switzerland in 1958 and in the United 
States in 1959 (Solomon et al., 1996). Atrazine has been produced by other companies since the early 
1970s when the patent protection of Ciba-Geigy expired. Ciba·Geigy in 1989/1990 supplied approximately 
60 % of the market (Sei1er et al., 1992). In 1992, atrazine was produced by over 10 different companies 
marketing it aJone or in mixtures with their own active ingredients (Seiler et al. , 1992). 
Atrazine is prepared, in the presence of an acid-binding agent, by the reaction of cyanuric chloride with one 
equivalent of ethylamine followed by one equivalent of isopropylamine (Sittig, 1980). The reaction pathway 
for the preparation of atrazine is shown in Figure 4.2. 
cyanuric chloride atrazine 
Figure 4.2: Reactioo patbway of the iDdustriaJ prep.ratioo of .tmine (Sittig, 1980) 
4.1.2 Chemical and pbysical properties 
Chemical and physical properties of atrazine are listed in Table 4.1. 
Table 4.1 : Chemical and physical properties of atrazine (Tomlin, 1997) 
Parameter 
molecular weight 
molecular formuJa 
form 
melting point 
boiling point 
vapour pressure 
Henry's law constant 
standard gravity (density) 
water solubility 
pKa 
Value 
215.687 
colourless powder 
175,8 °C 
205.0 QC at 10 I kPa 
0,0385 mPa at 25 QC 
Log P = 2.5 at 25 ' C 
1,5 X 10-4 Pa m1 mol" 
1,23 at 22 QC 
33 mg L-' at pH 7 and 22 QC 
1,7 (very weak base) 
Chapter 4 ATRAZlNE Cm:M!sTRy 4-3 
Alrazine is relatively stable in neutral, weakly acidic or weakly alkaline media; it is rapidly hydrolysed to 
the hydroxy derivative in strong acids or alkalis, and at 70 DC in neutral media (Tomlin. 1997). Pure 
atrazine may be stored for several years, decomposition begins immediately after application in soil 
(Huber, 1993). The stability of the s-triazine ring is due to the electronic: configuration of the heteroc:yclic 
ring with the 1t electrons being delocalized over the six ring atoms. The 1[ electrons are not evenly 
distributed over the whole ring and are localized in the vicinity of the niuogen atoms. The strong 
electrooegativity of the nitrogen atoms causes the s-triazine ring to display few of the characteristics of a 
compound containing an aromatic nucleus (Esser et al. , 1975; Weber, 1970). 
Properties of s-triazine derivatives are determined by the substituents OD the ring (primarily by the 
substitueDt in position 6), the ring itself does not contribute except for its effect 00 the charge distribution 
(Leclercq and Pacikcva, 1979; Weber, 1970). Substituents in position 6 are usually chlorine, methoxy 
(OCH3) or thiomethyl (SCIh) groups, various alkylamino groups are substituted in positions 4 and 6 
(Leclercq and Pacikova, 1979). The dissociation constant is strongly affected by the substituent in position 
6, basicity increases in the oroc:r Cl < SCIh < OCH3. The effect of the groups in positions 2 and 4 are 
smaller but still pronounced. the greater the number of ethyl groups replacing hydrogen atoms, the more 
basic the compound (Leclercq and Pacakova, 1979). 
Atrazine acts as a herbicide because of its inhibition of photosynthesis, it is thus less toxic to animals than 
to plants (Solomon et al., 1996). The toxicity of atrazine to different animal species is listed in Table 4.2. 
The terms LD~ and LC~ refer to the dosage and concentration of a chemical that kills 50 % of the group of 
animals to which it is administered.. Acute LD~ values indicate that the chemical administration was by a 
single dose; dosage is expressed in mg active ingredient per kilogram of anima1 bodyweight (Jones et al., 
1968). 
Table 4.2: Acute toxicity of atrazine (Seiler et al., 1992) 
Parameter Species 
acute oral LD~ rnl 
acute oral LD~ mouse 
acute dermal LD~ rn. 
acute inhalation LC~ rn. 
skin irritation rabbit 
eye irritation rabbit 
sensitisation guinea pig 
Concentration 
(mg Iqfl) 
I 869 to 3080 
1 750 to 3 992 
> 3 100 
> 710 mg m·l 
slight irritant 
no irritant 
sensitising 
Chapter 4 ATRA:mII'E CIlEMISrRY 4-4 
The evaluation performed by Ciba-Geigy indicated that atrazine had no mutagenic effects on manunals 
(Seller et al. , 1992) . .A!razine was also found not to be mutagenic in the Ames Salmonellaimicrowme assay 
(Simmon et al. , 1979). Long term experiments indicated a slight impairment in weight development 
following high 00ses of atrazine in rats, mice and dogs; dogs fed for a year with I 000 mg kg-I of atrazine in 
their food showed changes in heart and circulation. No carcinogenic effects, indepencient of the dose 
administered, were found in mouse trials (Seiler et al., 1992). 
Exposure to atrazine uncier factory conditions did not adversely affect the health of workers or increase the 
incicience of benign or malignant disease atnibutable to atrazine (Loosli, 1995; Seller et al., 1992). 
Weighted evicience of studies of farming JX)pulations in the United States support no causal association of 
malignant changes with atrazine (Loosli, 1995). Two studies of a rural Italian population suggested an 
increase in ovarian tumours in exposed women; statistical review suggested the increase could be by chance, 
exposure data was unsatisfactory and. no supp:>rting evidence has been presented in other studies 
(Loosli, 1995). High doses of atrazine (500 and 1 000 g) were ingested. in two medically documented cases 
of intended suicide. No signs of poiwning due to atrazine were reported, therapeutic measures were 
required. to control the adverse effects due to the c:o-formulants (Loosli. 1995). 
4.2 ATRAZINE IN TIlE ENVIRONMENT 
Atrazine js a herbicide and is introduced into the environment for weed control during crop cultivation or 
on railway tracks (Huber, 1993). Point-source contamination arises from accidents, inadequately cleaned 
tools, spillage during pesticide mixing and from diSJX)sal of waste pesticide and empty containers 
(Gan et al.. 19%; Huber, 1993). Atrazine contaminates water wurces by seeping into groundwater or from 
runoff processes into surface water (Huber, 1993). 
Atrazine persists longer in soils than other herbicides and has been the most frequently detected herbicide in 
groundwater in Europe (Cbopra et al. , 1973; Galassi et al. , 1996). The half-life of atrazine in wil (the time 
for 50 % dissipation) varies between 14 and I 09 ~ degradation is usually biphasic in that initial rapid 
degradation occurs during the first 2 months after application and slower degradation occurs during the 
drier swnrner and cold winter months (Koskinen and Clay, 1997). Auazine concentrations of between 0 
and 10 jJ.g L'\ 0.4 and 10 Ilg L-1 and 0 and 26 jJ.g L· 1 have been detected in rivers in France, Austria and 
Germany. respectively; concentrations in rivers in the United States are usually between 0 and 20 jJ.g L-1 
(Huber, 1993; Solomon et al. , 1996). At.ra2.ine residues in rivers in agricultural regions are usually episodic 
with major peaks in spring and early summer following application (Solomon et ai" 1996). The presence 
and persistence of atrazine in potable water WUTCeS has resulted in the regulation or banning of atrazine use 
in many European countries. Health standards have been set in the United States by the Environmental 
Protection Agency for atrazine at 3,0 ~ L"] and in the European Community the health advisory standard 
Chapter 4 ATRAnIIo"J: CHEMISTRY 4-5 
for each triazine herbicide and. metabolite is 0,1 ~g L-\ the total triazine concenuation is not to exceed. 0,5 
~g L-\ in groundwater (Thurman et al. , 1991 ; Thurman et al. , 1994). 
The role of atrazine in the environment has been reviewed in the following sections, application and. usage 
in Section 4.2. 1, persistence and. clegradation in soil in Section 4.2.2, persistence and. clegradation in aquatic 
systems in Section 4.2.3 and. metabolism in plants in Section 4.2.4. 
4.2.1 Application and usage 
Atrazine is used for the control of broadleaf and. some grassy weeds in maize, sorghum and sugarcane 
cultivation (Koskinen and Clay, 1997; Solomon et al. , 1996). It can be applied. before or after the 
emergence of weeds and. generally provides season-long weed control, typically, the herbicide is applied 
pre~mergence as a water-disrx:rsed spray or in liquid fertiliser (Koskinen and Clay, 1997; Solomon et al., 
1996). It is sold as the active ingredient in various wettable powders or in granular formulations and is 
marketed under trade names such as Gesaprim~ and Aatrex~ (Buber, 1993). Atrazine is also used as a 
non-selective herbicide for weed control on railway tracks (Huber, 1993). 
Atrazine is one of the most heavily used herbicides in the United States, approximately 29 000 t active 
ingredient was applied annually between 1987 and 1989 to aOOut 84 % of the maize crop (Koskinen and 
Cia)'. 1997). The estimated use in 1993 was between 32 000 and. 34 000 t active ingredient and in 1995 
between 31 000 and 33 000 t active ingredient (Koskinen and Clay, 1997; Solomon et al., 1996). 
The detection of atrazine in ground and swface water resulted in the reduction of the maximum alJowabJe 
application rate of atrazine. The Good farmi ng practice programme for atrazine, presented in Table 4.3, 
was proposed by Ciba-Geigy in 1991, it was implemented in 1992 in European Conununity countries that 
did not already have more stringent atrazine restrictions in place (Seiler et al .. 1992). 
Table 4.3: Good farming practice programme for atrazine (Seiler et al.. 1992) 
Parameter 
target crops : 
maximum use rate : 
timing of application: 
product offer: 
maize and sorghum only 
1,5 kg active ingredient per hectare per year 
pre-emergence and early post-emergence. no autumn applications 
prepack mixtures favoured compared to straight atrazine formulations 
The registered application rate (as recommended on the label) of atrazine in the United States was reduced 
from 5.6 to 3,4 kg active ingredient per ha in 1990 and further. in 1993. to a maximum total pre- and 
post~mergent use rate of 2,2 to 2,8 kg active ingredient per hectare (Solomon et aI. , 1996). 
Chapter 4 A TRAZINE CHEMISTRY 
4.2.2 Atrazioe in soil 
The applied atrazine tends to remain in the surface soil as it degrades, however, off-site movement has been 
documented (Koskinen and Clay, 1997). Movement of atrazine through soil into groundwater is affected by 
soil processes that vary throughout the soil profile. The two major processes that affect the amount of 
auazine present in soil and the movement thereof are transformation and retention (Koskinen and 
Clay, 1997). Transformation processes (by partial or complete degradation) reduce or eliminate the amount 
of atrazine present in soil and retention processes (by adsorption) decrease or eliminate the amount of 
atrazine available for ~rt to groundwater (Koskinen and Clay, 1997). 
4.2.2.1 Transformation 
Atrazine is transformed in soil by chemical and microbial processes such as hydrolysis, N-<iealkylation. 
deamination and ring cleavage (Kaufman and Kearney, 1970; Koskinen and Clay, 1997). The primary 
metaOOlites in the transformation of atrazine, as shown in Figure 4.3, are hydroxyatrazine 
(6-hydroxy-N .. ethyl-N' -( I -methylethyl)-I ,3 ,S-triazine-2, 4--di.amine), deethylatrazine (6-<:hlor<rN-{ I-methyl 
ethyl)-1 ,3,S-triazine-2,4-<1iamine) and deisopropylatrazine (6-<:hlo~N-ethyl·I .3 ,S-triazine-2 .4-<1iamine) . 
These are further degraded to deethyldeisopropyiatrazine (6<hlo~I ,3 ,S-triazine-2 ,4-<1iamine) .. deethyl 
hydroxyatrazine, deisopropylhydroxyatrazine and ammeline (Koskinen and Clay. 1997; Winkelmann and 
Klaine, 1991a). 
hydro",.-yatrazine deethylatrazine deisopropylatrazinc 
deethyldeisopropylatrazine deethylhydro", .. yatrazine deisopropylhydroxyatrazine 
ammeline 
Figure 4.3: Primary and secondary metabolites of atrazine (Koskinen and Clay, 1997) 
Cbapter 4 A TRAZlNE CHEMISTRY 4-7 
Attazine is transformed. in soil by adsorption-catalyzed. chemic:al hydrolysis to hydroxyatrazine (Burkhard 
and Guth, 1981 ; Skipper et al., 1967; Skipper and Yolk, 1972). The reaction follows first order reaction 
kinetics (Burkhard and. Guth, 1981) and. is dependant on factors such as pH, temperature and. concentration 
of dissolved organic carbon (Koskinen and. Clay, 1997). Burkhard and. Guth found. that the formation of 
hydroxyatrazine was greater at a lower pH, the half·life of anazine (at 22°C) in two German stand.ard soils 
of pH 4,8 and 6,5 was 53 and 11 3 d., respectively (Burkhard and Gulh, 1981). The hydrolysis half·life of 
atrazine (at 25 °C) in aqueous buffered solutions of pH 5, 7 and 9 was 64, > 200 and > 200 d., respectively 
(Burkhard and Guth, 1981). The degradation of atrazine is dominated by chemical transformation to 
hydroxyatrazine in acidic soil and by microbial transformation to N--dealk:ylation metabolites in alkaline soil 
(Blumhorst and Weber, 1994; Qiao et al., 1996). 
The metabolites, deethylatrazine and deisopropylatrazine, are fonned by microbial degradation through 
N~lation of atrazine (Blumhorst and Weber, 1994; Kaufman and Keamey, 1970). The presence of 
both alkyl groups on the triazine ring inhibits the microbial dechlorination ofatrazine, however, elimination 
of one or both of the alkyl groups allows for the microbial dechlorination of the metabolites producing the 
hydroxylated derivatives, as shown in Figure 4.3 (Belli and Khan, 1986). 
Greater amounts of deethylatrazine than deisopropylatrazine are produced. from the N..QeaJ..kylation of 
atrazine (Gan et al.. 1996: Sbapir and. Mandelbaum, 1997; Winkelmann and Klaine, 199Ia). Deethyl· 
atrazine. deisopropyiatrazine and hydroxyatrazine are all formed in surface soil, however, with increasing 
soil depth atrazine is transformed mainly through N~lation rather than through chemical hydrolysis 
(Rodriguez and Harkin, 1997; Sorenson et al., 1995). The major metabolite produced in the top 10 cm of 
soil is hydroxyatrazine whereas deethylatrazine is the most COmmon metabolite produced. at greater depths 
(Sorenson et al ., 1994; Sorenson et al., 1995). Constant levels of hydroxyalraZine and deisopropylatrazine 
throughout the soil profile indicate that the rates of fonnation and degradation are relatively equal. the 
increasing concentration of deethylatrazine with depth indicates that it is more resistant to degradation and 
can accumulate over time (Sorenson et al., 1994: Sorenson et al., 1995). 
The ratc of atrazinc degradation depends on soil and environmental factors that influence the presence and 
aClivi~' of atrazine-<Segrading microorganisms (Koskinen and Clay, 1997). These factors include atrazine. 
water and 0"'1'gen concentration, soil temperature, soil type, pH and organic matter content and previous 
soil management and crop practices (Koskinen and Cia)', 1997; Skipper and Yolk, 1972; Stojanovic et al., 
1972). Atrazine is degraded faster at a soil temperature of 2S °C than at 10 °C, atrazine half·life is 28,4 d 
at 25°C and 102.9 d ID °C (Qiao et al., 1996). Attazine concentration (5 to 5 ()()() mg kg" of soil) did not 
influence the rate of biodegradation in a clay loam soil (haif·life < 56 d), however. degradation decreased 
with increasing concentration in a sandy loam soiL the half·life was 56 d at 5 mg kg" and 91 d at 5 000 mg 
kg'\ of soil (Gan et al., 1996). Airazine is degraded more rapidly in silt loam than in clay loam or sandy 
Cbapter 4 ATRAllNI: CHEMISTRY 4-8 
loam soil, atrazine concentration after 4 months decreased to 64, 75 and 80 0/0, respectively, of the initial 
concentration in sill loam, clay loam and sandy loam soils (Sorenson el al., 1993; Sorenson et al. , 1994; 
Sorenson et al. , 1995). Atrazine degradation is reduced during unsaturated moisture conditions, 22 % of 
initial atrazine remained after 60 d under sarurated conditions whereas 34 % remained under unsaturated 
conditions (Kruger et al., 1997). 
Atrazine degradation is slower in subsurface soil (Lavy et al. , 1973; Miller et al., 1997; Shapir and 
Mandelbaum, 1997). Lavy and co-workers found that the phytotoxicity of atrazine in a silty clay loam soil 
dissipated within 5 months at a depth of 15 cm, within 17 months at a depth of 40 cm and no dissipation 
occurred after 41 months at a depth of 90 cm (Lavy et al., 1973). Shapir and Mandelbaum found that after 
30 d of incubation more than 50 % of atrazine in surface soil had degraded whereas less than 33 % had 
degraded in soil from a depth below 50 cm (Sbapir and Mandelbaum, 1997). Atrazine degradation is 
greater in soil that has a history of atrazine treatment. Vanderheyden and co-workers found thal the 
half-life of atrazine in soil, typically between 35 and 50 d, was reduced to between 5 and 10 d in maize 
fields that had been treated annually with atrazine for more than five years (Vanderheyden et al. , 1997). 
The factors that affect the degradation of atrazine also affect the degradation of the primary metabolites of 
atrazine. Deethylatrazine degrades slower in subsoil than surface soil and is also lower under saturated than 
unsaturated moisture conditions (Kruger et al. , 1997). The degradation of deethylatrazine, deisopropyl  
atrazine, deethyldeisopropylatrazine and hydroxyatrazine follow first order reaction kinetics, 2 % of the 
deethylatrazine and 2 % of the deisopropylatrazine, 33 % of the hydroxyatrazine and no deethyldeisopropyl 
atrazine was found to remain in soil 180 d after application (Winkelmann and Klaine, 199Ib). The soil 
half-lives of atrazine. deethyIatrazine, deisopropylatrazine. deethyldeisopropylatrazine and h}'droxyatrazine 
were 21. 26, 17, 19 and 121 cl, respectively, under similar conditions (Winkelmann and Kiaine, 1991b). 
Mineralisation of atrazine to carbon dioxide involves cleavage of the triazine ring. Early investigations 
found that the triazine ring was not readily broken by soil microorganisms and the fonnation of 14C02 from 
ring·labeled atrazine in soil was generally low. Skipper and co-workers found that approximately 0.4 to 
0.5 % 14C02 evolved from ring-Iabeled atrazine in a sandy loam and silry clay loam soil after a 14 d 
incubation period (Skipper et al. , 1967). Kruger and co-workers found that less than I % of the applied 14C 
had mineralised after a 120 d period (Kruger et al., 1997). Skipper and Yolk measured a 0.1 % formation 
of 14COz from ring-Iabeled atrazine after a 14 d incubation period and Dao and co-workers found that the 
formation of 14~ was less than 0,5 % after an 8 month incubation period (Dao et al., 1979; Skipper and 
Yolk, 1972). Amounts of carbon dioxide evolved in the mineralisation of atrazine range from 0,05 % 
within 3 months to I % per month (Geller, 1980). Higher rates of atrazine mineralisation have been 
reported for soils that have a history of atrazine application (Vanderheyden et ·a!.. 1997: Winkelmann and 
Klaine, 199Ia). Gan and co-workers found that an average or 64 % mineralisation occurred after 140 d for 
Chapter 4 A TRADl"OE CHEMISTRY 4-9 
atrazine concentrations of between 5 and 500 mg kg- I of soil (Gan et al. , 1996). The degradation products 
of atrazine are more easily biodegradable than the parent compound (GeHer, 1980). Mineralisation of 
ring-labeled 14C deethylatrazine, deisopropylatrazine and hydroxyatrazine yielded 2S, 16 and 21 %, 
respectively. of 14C(h after 180 d (Wiokelmann and Klaine, 199Ib). 
The atrazine-ring carbon atoms are in the same oxidation state as in carbon dioxide and cannot be used by 
microorganisms as a source of carbon or energy for growth, however, energy can be obtained from the 
oxidation of the carbon atoms in the side chains substituted on the s-triazine ring (Cook, 1987). Atrazine is 
usually utilised as a nitrogen source for microbial growth when carbon and energy are supplied by other 
substrates (Cook, 1987). 
Microbial strains that can mineralise atrazine, especially in soils with a prolonged history of annual atrazine 
application, have been identified (Bebki and Khan, 1986). Mirgain and co-workers identified three 
bacterial species (Sphingomonas paucimobi/is. Sphingobacterium mu/rivorum and Agrobacterium 
radiobacler) that together, but not individually, could degrade atrazine (Mirgain et al., 1995). Behki and 
Khan identified three strains of Pseudomonas (Pseudomonas putida, Pseudomonas JIuorescens and 
Pseudomonas slutzeri ) that degraded atrazine via N-dealkylation to deethylauazine and deisopropylatrazine 
(Behki and Khan, 1986). Two of the species could dechlorinate deethylatrazine and deisopropylatrazine 
(but not atrazine) following incubation in glucose-supplemented mineral salts me<lium (Bebki and 
Khan, 1986). Shapir and Mandelbaum demonstrated that a Pseudomonas strain completely degraded 
atrazine. between 90 and 100 % mineralisation occurred in soil within IS d after inoculation with the 
Pseudomonas strain (Shapir and Mandelbaum, 1997). The first step in the mineralisation was 
dechJorination to hydroxyatrazine (Shapir and Mandelbaum., 1997). 
Mirgain and co-workers, in an investigation in north-eastem France, isolated several aerobic gram-negative 
bacteria that were capable of metabolising atrazine in association with other species or in pure culture 
(Mirgain et al .. 1993). The degradation curves consisted of three phases. an adaptive period in which 
degradation was slow or absent. a rapid degradation phase in which 80 % of the pesticide was consumed 
over a few days and the final stage in which the degradation of the remaining atrazine was again slow 
(Mirgain et ai., 1993). Isolates from agricultural soil samples indicated that atrazine was degraded by 
Acinetobacter calcoaceticus in pure culture and by a combination of Pseudomonas alcaligenes and an 
Agrobaclerium species. Atrazine was also degraded by a consortium of microorganisms from a garden soil 
sample (Pseudomonas jluorescens, Pseudomonas purida, FltlYobaclerium mllltivorum. Enlerobacler 
c/oacae and an Acinetobacter species). No microbial species in a groundwaler sample were found that 
degraded atrazme whereas two species (Pseudomonas purida andXanlhomonas maltophilia) were identified. 
in a surface water sample (Mirgain et al.. 1993). Giardi and co-workers demonstrated that a Nocardia 
Chapter 4 ATRAZlJIo'"E CHEI\USTRY 4-10 
lOstrain was capable of degrading atrazine, degradation (as shown in Figure 4.4) was through 
N-dealkylation and deamination yielding dicyanodiamidine (Giardi et al., 1985). 
atrazine 
o 
H,N)lNH 
HN..,lNH, 
dicyanodiamidine 
Cl 
NJ.-N 
• A)l 
H,N ~N NHCH(CH,n 
deethylatrazine 
deethyldeisopropylatrazine 
• 
4-amin0-6~hloro-s-tria:zine 
Figure 4.4: Microbial degradatioD of atrazine by a Nocardia strain (Giardi el al., 1985) 
The biodegradation pathway of atrazine is presented in Figure 4.5 . Biodegradation is predominantly 
through N-dealkylation to deeth)'latrazine and deisopropylatrazine even though Shapir and Mandelbawn 
showed thal atrazine can be dechlorinated 10 hyclroxyatrazine by a Pseudomonas strain (Erickson and 
Lee, 1989: Shapir and Mandelbaum, 1997). Several pathways e:cist for subsequent degradation of 
deethylatrazine and deisopropylatrazine, as shown in Figure 4.5. these depend on the order in which 
dechlorination. deamination and further dealkylation take place. Behki and Khan found that dechlorination 
occurred first in the degradation by two Pseudomonas bacterial strains whereas Giardi and co-workers 
found., as shown in Figure 4.4, that dealkylation and deamination occurred before dechlorination in the 
degradation by a Nocardia bacterial strain (Behki and Khan, 1986: Giardi et al., 1985). Degradation 
through ammeline. ammelide and cyanuric acid was caused by a Pseudomonas species strain A (Jutzi et a1.. 
1982). The different degradation pathways. as shown in Figure 4.5, converge with the fonnation of 
cyanuric acid (Cook et al. , 1985; Cook., 1987). Two soil fungi (Slachybolrys char/arum and Hendersonula 
toroloidea). a Pseudomonas species strain D and a facultative anaerobic bacterium have been shown to 
degrade cyanuric acid (Cook et a1.. 1985; Jessee et al. , 1983: Wolf and Martin, 1975). rung cleavage of 
cyanuric acid., as shown in Figure 4.5. produces biuret that is further degraded to urea and then 10 carbon 
dioxide and ammonia (Cook et aL 1985). Three moles of ccn and three moles of NIL T are produced per 
mol of cyanuric acid (Cook et al .. 1985). 
Chapter 4 ATRAZIN'E CRE.'dISTR\' 
atrazine 
hydtoxyatra2ioc: 
dccthylbydroxyatrazine 
2-~h1oJ"0-4.hydrOXY­
 s-triarine 1 
cyanuric acid 
biurel 
4-amino-2,6-dihydroxy 
s-tr:iazine 
• 
4-11 
dc:isopropyJatrazine 
OH 
N~N 
H~H~HNAN~NH2 
deisopropylbydroxyatrazine 
4 ,6-di.hydroxy-2-cthylamino 
,s-.triazine 
Figure 4.5: Mineralisation pathway of atrazine (Erickson and Lee, 1989; Koskinen and 
Cia),. 1997) 
4.2.2.2 Retention 
Retention is defined as the restriction or retardation of mobility of a chemical in soil because of the 
interaction of the chemical with the surface of a soil panicle (Cben&. 1990). Retention of atrazine in soil 
Chapter 4 ATRAZlI'i"E CHEMISTRY 4-12 
decreases the amount available for weed control. microbial transformation and the amount that can move to 
greater soil depths leading to groundwater contamination (Koskinen and Clay, 1997). Typically between SO 
and 80 % of atrazine under sluny conditions is retained in silt loam, loam or clay loam surface soils 
(Koskinen and Clay, 1997). The energy required and available for desorption determines whether the 
retention process is reversible or irreversible (Cheng, 1990) . 
Different binding mechanisms are involved in the retention of auazine and its metabolites in soil, the 
binding mechanisms or binding strengths change with time, more auazine was extracted from soil after 
35 d than after 138 d using methanol or supercritical fluid extraction (Koskinen et al, 1995). Mechanisms 
such as ion exchange, metal complexation, covalent bonding, hydrogen bonding, van der Waals forces and 
physical trapping have been postulated for the retention of auazine in soil (Bailey and White. 1970; 
Cheng, 1990; Chopra et al., 1973). Cheng proposed that hydrogen bonding (the partial transfer of charge 
between an electron donor and acceptor) could be a major force in the binding of atrazine to soil 
(Cheng, 1990). Welhouse and Sleam evaluated the hydrogen-bonding potential of atrazine by measuring 
fonnation constants for hydrogen-bond complexation with hydrogen-bond oonors and acceptors typically 
found in organic matter in soil (Welhouse and SIeam, 1993a; Welhouse and Bteam, 1993b). Atrazine was 
found to fonn strong C(H)perative hydrogen bonds with compounds such as pyrrolidinone and acetic acid 
because of the favourable geometry for orbital overlap and the ability of both molecules 10 redistribute 
electronic charge and reinforce complexation (Welhouse and SIeam, 1993b). The most stable bound 
residues of alrazine in soil are associated with humified organic matter (Koskinen and Clay, 1997). 
Atrazine adsorption is dependant upon factors such as organic carbon content, soil clay content, soil type, 
pH. atrazine concentration and atrazine-soil contact time (Bailey and White, 1970; Koskinen and 
Clay. 1997). The formation of soil..oound residues of alTaZine decreases in subsoil. Kruger and co-workers 
found that after 120 d soil-bound residues in the surface soil accounted for between 62 and 73 % of the 
applied atrazine and in the subsoil for between 13 and 22 % of the applied atrazine (Kruger et al. , 1997). 
Miller and co-workers found that more than 50 % of the applied atrazine after 6 months at a depth of 45 cm 
was still unbound and untransfonned and at a depth of 75 cm almost 80 % was unbound and wmansfonned 
(Miller et al ., 1997). Atrazine adsorption increases with contact time with soil (Dao et al.. 1979: 
Miller et al .. 1997) and with increasing soil organic matter and clay content (Burkhard and Guth. 1981 ; 
Chopra et al . 1973: Koskinen and Rochene, 1996). Atrazine adsorption is greater at lower pH values. more 
atrazine was adsorbed to soil in the pH range 4 to 6 than to soil of pH 7 or greater (Clay and Kosldnen. 
1990; Koskinen and Clay, 1997). Atrazine adsorption increased in sand as moisture content increased from 
4 to 16 % and in silt loam as moisture increased from 9,6 to 27 % (Koskinen and. Rochette, 1996). 
The adsorption characteristics of atrazine metabolites are different to those of atrazine. Hydro,,:yatrazine 
was more strongly adsorbed to soil than atrazine (Clay and Koskinen, 1990; Dao et al. , 1979). Seybold. and 
Mersie found that the adsorption of atrazine and its major metabolites decreased in the order 
Chapter 4 AJ'RA27NE Cm:MIsrRv 4-13 
bydroxyatrazine > atrazine > deisopropylatrazine > deethylatrazine (Seybold and Mersie, 1996). The 
desorption process decreased in the order deethylatrazine > atrazine and deisopropylatrazine > 
hydroxyatrazine (Seybold and Mersie, 1996). Both adsorption and desorption influence the mobility of a 
chemical in soil, high adsorption and low desorption restrict the mobility and the potential of a chemical to 
leach through the soil column. Seybold and Mersie ranked the overall order of decreasing mobility as 
deethylatrazine > deisopropylatrazine > atrazine > hydroxyatrazine (Seybold and Mersie, 1996). Kruger 
and co-workers, in an investigation of five Iowa soils, identified three mobility groups for atrazine and 
selected metabolites: most mobile - deethylatrazine; intermediate mobility - atrazine, deisopropylatrazine 
and deethyldeisopropylatrazine; nearly immobile - hydroxyatrazine and ammeline (Kruger et al. , 1996). 
4.2.3 Atrazine in aquatic systems 
Atrazine has been detected in both groundwater and surface water systems (Huber, 1993; Koskinen and 
Clay, 1997; Solomon et al., 1996). Atrazine contamination of aquatic systems is particularly widespread 
where maize is the dominant crop such as in the Midwestern United States (Solomon et al., 1996). The 
persistence and mobility of atrazine and metabolites such as deethylatrazine in soil allow for leaching into 
groundwater (Sorenson et al., 1994). Atrazine contamination of surface water systems is caused by runoff 
following precipitation or irrigation of agricultural fields (Huber, 1993). 
4.2.3.1 Groundwater 
Atrazine has the potential to contaminate groundwater because of its persistence and mobility in soil. 
Measurements of atrazine half life in soil range between 14 and 109 d (Koskinen and Clay. 1997). 
Sorenson and co-\\-'Orkers found that the amount of atrazine persisting 16 months after treatment was 32, 24 
and 22 %, respectively, in clay loam, silt loam and a sandy loam soil (Sorenson et al., 1993; Sorenson et al ., 
1994: Sorenson et al. , 1995). Gan and co-workers and Sorenson and co-workers found that atrazine moved 
to a depth of 70 to 80 cm within the first month after application (Gan et al., 1996; Sorenson et aI., 1994). 
Atrazine movement to this depth in clay loam was due to macroporous water flow (macropores arc caused 
by roots and earthworm burrows); insufficient water from precipitation (2,3 cm in 4 increments) and 
irrigation (5 cm in 2 increments) was available to move the atrazine to such a depth un1ess macroporous 
flow occurred (Sorenson et al., 1994). The rate of atrazine leaching is soil dependent. the greater water 
holding capacity of silt loam and clay loam allows for greater atrazine movement (Sorenson et al .. 1995). 
Leachate collected from an 80 cm clay loam column 16 months after treatment contained 0,04 % of the 
applied atrazine (Sorenson et ai. , 1994). The metabolites most likely to leach into groundwater because of 
their lower retention in soil are decthylatrazine, the dominant metabolite in subsoil, and deisopropylatrazine 
(Koskinen and Clay, 1997; Krugcr et al ., 1996). 
A representative survey of the groundwater of the United States by the US Environmental Protection 
Agency recorded atrazine concentrations of between 0.18 and 1,04 IJ.g L' \ half of the tested wells showed 
concentrations below 0,28 IJ.g L· j (Huber. 1993). The maximum concentration recorded in wells in the 
Cbapter 4 ATR.A.ZIl"I'E CRDfiSTRY 4-14 
Netherlands. Switzerland and Germany was 0,58; 1,34 and 3, 1 ~g L' I, respectively; the higher values were 
found in shallow wells in the vicinity of intensive agricultural production (Huber, 1993). The 
concentrations of atrazine. deelhylatrazine and deisopropylatrazine measured in grounm..'ater at Dalmine, 
Italy, in June 1995 were 0,232; 0,045 and 0,058 f.Lg L·I , respectively (Galassi et al., 1996). 
Significant degradation of atrazme and its metabolites does not occur in groundwater because of the low 
microbial populations, low amount of dissolved organic carbon (below I mg of carbon L· l ) and temperatures 
below 10 °C (l-Airgain et al ., 1993). Rodriguez and Harkin found that more than 85 % of alrazine in 
groundwaler was still present after 270 d (Rodriguez and Harkin, 1997). The atrazine concentration in a 
well in the Lombardy area in Italy (where atrazine use has been banned since 1986) was in the range of 0 ,25 
to 0,46 }.lg L'\ bet\\'een February and December 1987 (monitored monthly) and between 0,147 and 
0,232 f.Lg L' \ between May and July 1995 (Galassi et al" 1996), The resistance of deethylatrazine to 
degradation in subsoil indicates that it is just as persistent as atrazine in groundwater (Rodriguez and 
Harkin, 1997). 
4.2.3.2 Surface water 
Atrazine contamination of surface water is caused by runoff following precipitation and irrigation of 
agricultural fields (Huber, 1993). Losses due to runoff range between 0,1 and 7,7 % of the applied atrazine, 
the loss in a year with average rainfall is typically 1,2 % (Gaynor et al., 1992; Gaynor et al., 1995). 
Atrazine transport in runoff occurs throughout the year but losses are higher during the growing season 
when runoff occurs soon after herbicide application (Gaynor et al., 1992). 
Atrazine concentrations of between 0 and 87 Ilg L'\ have been detected in rivers in the Urtited States, 
concentrations. however. rarely exceeded 20 f.Lg L'l (Huber, 1993; Solomon et al .. 1996). Atrazine 
concentration in ri\'ers in France ranges between 0 and 10 f.Lg L'\, in Austria between 0.4 and 10 Ilg L'\ and 
in Germany between 0 and 26 f.Lg L'l , the mean concentration levels are significantly lower than the upper 
peak values (Huber. 1993). The atrazine concentration in standing freshwater systems (ponds. natural lakes 
and reservoirs) is lower than that in flowing water; for example, in Germany, most of the detected atrazine 
concentration in standing freshwater systems is below 0, 1 IJ.g L'l (Huber, 1993), Atrazioe concentration in 
surface water in the United States is highest in the first two months after application in early summer (late 
spring and early summer rainfall flushes atrazine from croplands) and decreases during autumn and ",'inter 
months (Thunnan et aI., 1991 ; Thwman et aI., t 992). Atrazine is also found to persist from year to year in 
both soil and water (Thurman et al. , 1991). 
Atrazine metabolites such as doethylatrazinc and dcisopropylatrazine have been detected in surface runoff 
and surface water systems (Gaynor et aI" 1995: Thunnan et aL. 1991). Metabolite concentrations in surface 
Cbapter 4 A TRAZlNE CHEMISTRY 4-15 
water is dependant on the hydrologic conditions of the basin and timing of runoff; early rains and a dry 
summer leads to an peak of metabolites in surface water, a wet summer delays but increases the maximum 
metabolite concentration (Thurman et al. , 1994). The deethylatrazine peak concentration occurs later in the 
season than the atrazine peak concentration (Gaynor et al., 1992). The investigation by Thurman and 
co-workers of the surface water systems in the Midwestern maizebelt of the US found that the ratio of 
deisopropylatrazine to deethylatrazine was approximately 0,6 and that the ratio of deethylatrazine to 
atrazine, 2 monlhs after application, varied between 0,5 and 0,7 (norman et al., 1994). The persistence of 
atrazine and its metabolites in surface water in the United States decreases in the order of atrazine, 
deethylatrazine and deisopropylatrazine (Thunnan et al. , 1991; Thunnan et al. , 1992). The concentration 
of metabolites is greater in standing water than in flowing water; hydroxyatrazine, although not detected in 
flowing water, has been detected in Midwestern reservoirs in the United states (Solomon et al. , 19%). 
Deethylatrazine and deisopropylatrazine are also fonned, as shown in Figure 4.6, from the degradation of 
other triazine herbicides such as simazine, cyanazine and propazine (Thunnan et al., 1994). The 
investigation by Thurman and co-workers of the surface water systems in the Midwestern maizebeit of the 
United States found that deethylatrazine was formed predominantly from atrazine (98 %) with trace levels 
derived from propazinc; deisopropylatrazine was formed from atrazine (75 %), cyanaz.ine (25 %) and trace 
levels from sitnazine (Thurman et al., 1994). Atrazine is the most widely used triazine herbicide, over 
2,5 times more atrazine than cyanaz.ine was found by Thurman and co-workers to be applied on a regional 
scale, propazine had already been removed from the US market (Thmman et al. , 1994). 
deisopropyiatntzine deethylatr3Zine 
Figure 4.6: Formation of deetbylatrazine and deisopropylatrazine from simazine, cyanazine 
and propazine (Thunnan et al., 1994) 
Surface water ecosystems may be affected by the long tenn presence of atrazine and its metaOOlites. The 
two mechanisms by which atrazine is taken up by aquatic organisms are adsorption from water or uptake 
Chapter 4 A TR.A2lN£ CHEMISTRY 4-16 
via the food chain (Huber, 1993). The investigation by Ellgehausen and co-workers of a model aquatic food 
chain consisting of an algae (Scenedesmlls acutus), daphnid (Daphnia magna) and catfish (lCIa/urus me/as) 
species found that direct uptake from water was significantly more important than transfer of atrazine 
residues via the food chain (Ellgehausen et al. , 1980). Atrazine is taken up by nearly all aquatic organisms, 
however, a large portion of the adsorbed atrazine is eliminated quickly by the organisms (Huber, 1993). 
Ellgehausen and co-workers found that the degradation of atrazine in the investigated algae, daphnid and 
catfish species followed second order reaction kinetics (EIlgehausen et al., 1980). 
Phytoplanktons are the most sensitive organisms to atrazine in aquatic ecosystems (freshwater and 
saltwater), followed in decreasing order of sensitivity by macrophyteS, benthic invertebrates, zooplankton 
and fish (Solomon et al. , 1996). Phytoplankton biomass and production rates are unaffected by exposure to 
atrazine concentrations below 20 ~g LO ] , ecologically important effects occur only with explS\lre to 
concentrations above 50 ~g L'\ (Solomon et al., 1996). The investigation, initiated by the Environmental 
Protection Agency, of the ecological risk of atrazine in North American surface waters concluded that 
atrazine did not pose a significant risk to the aquatic environment (Solomon et al. , 1996). It was 
recommended that site-specific assessments be performed of the inhibitory effects of atrazine on algae, 
phytoplankton or macrophyte production in small streams that are vulnerable to agricultural runoff 
(Solomon et al., 1996). The review by Huber on the ecotoxicoiogical relevance of atrazine in aquatic 
systems concluded that the potential hazard. of atrazine in the environment depended on the concentration 
and degree of exposure, atrazine concentrations of up to 20 ~g LO ] would not cause any pennanent damage 
to aquatic ecosystems (Huber, 1993). 
4.2.4 Mode of action in plants 
Atrazine is readily taken up by all plants, take up is primarily through roots and foliage (Esser et al. , 1975; 
Solomon et al., 19%). Plant uptake of atrazine increases with increasing atrazine concentration, time of 
exposure, higher temperatures and low relative humidity (Esser et al. , 1975). Atrazine take-up in maize 
plants 2 months after treatment was found to be greater in sanely loam (3,6 % of appliecl atrazine) than in 
clay loam (2,9 % of appliecl atrazine) even though organic content levels in the soil were similar and soil pH 
in the sandy loam was lower than in the clay loam (Sorenson et aI., 1993 ~ Sorenson et al., 1994). Atrazine 
take·up in maize plants is also greater in silty loam that in clay loam (Sorenson et aI. , 1995). 
Atrazine is not toxic to plants in the dark, it thus interferes with the photosynthesis process in plants 
(Esser et al. , 1975). Atrazine inhibits photosynthesis by interrupting the light-driven flow of electrons from 
water to NADP. rticotinamide adertine dinucleotide phosphate (Esser et al , 1975). Chlorophyll molecules 
are oxidized by light energy during photosynthesis. Replacement electrons for the oxidized chlorophyll 
molecules (as shown in Figure 4.7) are obtained from the cleavage of water wi.th o:\")'gen being produced as 
a by-product, the reaction is known as the Hill reaction (Eben and Dumford, 1976; Solomon et al., 1996). 
Chapter 4 ATRA2INl: CHEMISTRY 4-17 
Atrazine inhibits the Hill reaction and competes with plastoquinone at its binding site thus blocking the 
transport of electrons from photosystem 11 (Ebert and Dumford, 1976; Solomon et al., 19%). 
·05 NADPH NADP 
Cyclic '-../ 
pbotopbospboryiatioo ferredoxin 1 
ATP '-../ ADP 
~NOtne'o FRS 
o 
0" 
tttt 
UGHT 
pbotosystem I 
Figure 4.7: Photophospborylation (Solomon et al.~ 1996) 
Inhibition of the Hill reaction leads to the destruction of the chlorophyll molecules in photosystem n since, 
without replacement electrons from water, the light-induced electron flow occurs until all of the chlorophyll 
molecules are oxidized (Solomon et al., 1996). The blockage of electron flow from photosystem n causes 
photophosphorylation (the production of A TP), reduction of photosystem I chlorophyll molecules, cyclic 
photophosphor:'-'Ialion and the production of NADPH to cease (Solomon et al.. 19%). The binding of 
atrazine to the plastoquinone n binding site is reversible since photosynthetic activity increased in plants 
removed from atrazine-contaminated growing medium (Solomon et al. , 1996). 
The phytotoxicity of atrazine is reduced by degradation, dechlorination metabolites are non-toxic and 
N-dealkylation metabolites are less toxic than atrazine (Harris, 1967; Kaufman and Keamey. 1970). 
Stratton found that atrazine was more inhibitory to growth and photosynthesis in two species of green algae 
and three species of cyanobacteria than hydroxyatrazine, deethylatrazine, deisopropylatrazine and 
deethyldeisopropyJatrazine (Stranon. 1984). Deethylatrazine and deisopropylatrazine are the most toxic 
transformation products for plants. Atrazine is between 7 to 10 and 20 to 50 times more inhibitor:'-' to 
cyanobacteria, respectively, than deethylatrazine and deisopropylatrazine, similarly, atrazine is bet1A:een 4 to 
6 and 7 to 13 times more inhibitory to green algae than these metabolites (Stranon. 1984). Hydroxyatrazine 
and deethylaeisopropylatrazine are relatively non-toxic, atrazine is between 200 and 1 000 times more toxic 
than either ofthesc metabolites (Stranon. 1984). 
Chapter 4 A TRAZlN'E CIIEMlSTRY 4-18 
Plants such as maize possess efficient detoxification me:chanisms for atrazine (Ebert and Dumford, 1976). 
Three possibly pathways for atra2ine detoxification in maize are hydroxylation, N-dealkylation and 
glutathione conjugation (Eben and Dumford, 1976; Shimabukuro, 1967; Shimabulruro et al., 1970). The 
presence of benzoxazinone (2,4-dihydroxy-7-methoxy-I .4-benzoxazin-3-one) in maize leads to the 
hydroxylation of atrazine producing hydroxyatrazine (Esser et al.. 1975; Shimabukuro. 1967). Raveton and 
co-workers found that the reaction followed first order reaction kinetics and proposed the reaction pathway 
shown in Figure 4.8 (Raveton et al., 1997a). 
H,C°yY°XOH 
~N 0 
6H 
benzoxazinone 
bydroxyatrazine 
atnlZine 
Figure 4.8: Reaction pathway of atrazioe hydroxylation in maize (Ra"'eton et al., 1997a) 
Ravelon and co-workers found thal after 72 h less than 5 % of applied atraz.ine was still present as atrazine 
in extracts of roots and shoots of maize seedlings, most atrazine had been transformed 10 hyclrox-yatrazine 
and small quantities of deethylhydrox-yatraz.ine and deisopropylhydroxyatrazine (Raveton et al.. 1997b). 
The displacement of the 6<bloro group in atrazine with glutathione also reduces the loxicily of atrazine 
(Lamoureux et a1.. 1972). The glutathione ronjugate of atrazine. shown in Figure 4.9, has been identilied 
in maize and sorghum leaves (Lamoureu." et al ., 1972; Srumabukuro et al .. 1970). It is fonned b)' an 
enzymatic reaction catalysed b)' glutathione s-transferase (Shimabukuro Cl al .. 1971). Shimabukuro and 
co-workers found that within 24 h between 60 and 80 % of applied atrazine to leaf blades of six maize 
varieties was convened to the glutathione conjugate of atrazine (Shimabukuro et al., 1971). 
Chapter 4 ATRAllNE CUEMlSJ'RY 4-19 
Figure 4.9; Glutathione conjugate of atrazine in maize leaves (Lamourew.: et al., 1972) 
Atrazine hydroxylation (which occurs mainly in the roots) and glutathione conjugation (which occurs only 
in the leaves) are independent processes (Raveton et al ., 1997a; Raveton et al. , 1997b). Hydroxylation is the 
dominant atrazine detoxification process in maize plants, the relative imponance of glutathione conjugation 
is greater with post-emergence herbicide treatment (Raveton et al. , 1997a; Raveton et al., 1997b). 
4_3 ATRAZINE IN WATER TREATMENT 
A comparative study of four disinfection methods (chlorine, monochloramine, chlorine dioxide and ozone) 
and an evaJuation of the effectiveness of granular activated carbon versus sand filtration for chemical 
contaminant remov31 was performed over a year at Jefferson Parish, on the mouth of the Mississippi river, 
in me United States (Lykins and Koff'skey. 1986; Lykins et al .. 1986). The raw water, after clarification and 
filtration. was split into five process streams, ODe for each of the disinfection methods and a control without 
an)' disinfection. The disinfection process streams consisted of a contact chamber, the contact time was 
approximately 30 min, followed by parallel filtration through a sand column. a granular activated carbon 
column and a duplicate granular activated carbon column. The control non-disinfection stream did not 
have a contact chamber but was also split into three parallel filtration streams. The influent concentration 
of atrazine ranged from 23 10 249 ng L'l and was not affected by chlorine dioxide. chloramine or chlorine 
disinfection whereas ozone disinfection reduced the atrazine concentration, on average. by 83 %. Atrazine 
concentration was unchanged after sand filtration, however, filtration through the granular activated carbon 
column removed between 95 and 97 % of the atrazine throughout the 1 year operational period (Lykins and 
Koffskey. 1986: Lykins et al ., 1986). 
A South African study investigated the fate of organic micropollutants in a pilot-scale integrated effiuent 
treatment/water reclamation system consisting of a denitrification reactor, a chemical clarifier. a combined 
nitrification pond and biological clarifier, prechiorination, dual media filtration. granular activated carbon 
adsorption and final chlorination (van Rensburg et 31., 1981). The atrazine concentration through the 
system, relative to the feedwater, was 81 % after denitrification. 64 % after the chemical clarifier. 60 % 
after the biological clarifier. 53 % after prechlorination. 62 % after dual media filtration and 0 % after 
adsorption with granular activated carbon (van Rensburg et al ., 1981). An investigation of a pilot plant 
consisting of a Clarifier, rapid sand filter. ozonation tank and slow sand filter indicated that the atrazine 
Cbapter 4 ATRAZlllf£ CHEMISTRY 4-20 
concentration was not reduced by flocculation and coagulation in the clarifier but was partly reduced (by 
between 30 and 45 %) after the ozonation tank (Montiel and Welte, 1992). Atrazine concentration in the 
effiuent from an agrocbemical formulation plant has been reduced from 6 300 mg L-1 to below 0,1 mg L-1 in 
an integrated. chemicallbiological oxidation system consisting of an aerobic fluidised bed, an aerobic 
trickling filter, a chemical oxidation reactor for ozonation and catalysed hydrogen peroxide oxidation, an 
anaerobic fixed bed filter, ultrafiltration modules and reverse osmosis and nanofiltration membranes 
(Haverhoek et al., 1997). An investigation of a nanofiltration pilot plant at the Mery-sur-Oise water 
treaUDent plant in France indicated that nanofiltration, for initial au-azine concentrations of between 0 and 
900 ng L-l , reduced the concentration by between 80 and 90 % (Agbekodo et al., 1996). 
Activated. caroon and advanced. oxidation processes, su<:h as ozone, are some of the most effective processes 
for reducing atrazine concentration in water treatment (Adams and Randtke, 1992b). Atrazine is strongly 
adsorbed by activated carbon, however, naturally occurring organic matter competes with atrazine and 
lowers the adsorptive capacity (Adams et al., 1990). Breakthrough of granular activated carbon filters can 
occur rapidly (Duguet et al. , 1990). Advanced oxidation technologies involve the generation of highly 
reactive radical intermediates, panicu1arly hydroxyl radicals (Beltrin et al., 1998). The reaction pathway 
proposed by Chan and oo-workers for the acid and alkaline hydrolysis of atrazine is shown in Figure 4. 10 
(Chan et al .. 1992). 
alkaline hydrolysis 
• 
atrazine 
! HO-
 Cl \ HO Cl 
I ,.../ acid hydrolysis X hydroxyatrazine 
NA N .. N \ N 
H3CH,cHN)",NANH,cH(CH3n H3CH,cHN,-l~Jl.NHCH(CH3n 
Figure 4_10: Acid and alkaline hydrolysis of atrazine (Chan et al., 1992) 
Chan and co-workers proposed that alkaline hydrolysis was a two-step nucleophilic displacement of 
chlorine ~' hydroxyl ions and that the acid hydrolysis proceeded via protonation at one of the side-chain 
nitrogen atoms (Chan et al. , 1992). The C-CI bond is readily broken (45.2 kJ mor' ) because of the electron 
Cbapter 4 A TRA2INE CHE.!.IISTRY 4-11 
deficiency of the ring carixm atoms and the electron withdrawing nature of the chlorine atom. The carbon 
atoms in a triazine ring are slightly positively charged because the 1t-elecuons are not evenly distributed 
over the ring but are localised in the vicinity of the nitrogen atoms (Esser et al. , 1975; Paris and Lewis. 
1973). 
Atrazine reaction with hydroxyl radicals also occurs at the site of the side chains producing dealkylated 
products. The mechanism proposed by Hapeman and co-workers for the formation of dealkylaled products 
from the reaction of atrazine with hydroxyl radicals in the presence of oxygen is shown in Figure 4. 11 
(Hapeman et aI. , 1995). 
atrazine 
2-acetami<L>6-chloro-
 4-isopropylamino-.,.-triazine 
deethylatrazine 
Figure 4.11: Reaction mecbanism for tbe formation of dealkyiated degradation products from 
tbe reaction of atrazine witb bydroxyl radica15 (Bapeman et al., 1995) 
Hydroxyl radical attack, shown in Figure 4.11. occurs at the side<hain carbon Cl to the amine nitrogen 
(Hapeman et aI. , 1995). The carbon is relatively electron-rich since the amine nitrogen provides additional 
electron density. The formation of deethylatrazine is favoured above that of deispropylatrazine as the 
Cbapter 4 A TRAZL"'£ CHEMISTRY 4-22 
reaction is sterically controlled, hydroxyl radical attack thus occurs more readily on the ethyl side chain 
than on the isopropyl side chain. Oxygen is required in the degradation of atrazine to form the perhydroxy 
radical (step 2 of the reaction mechanism presented in Figure 4.11) that reacts with an identical molecule to 
yield the final products. Hydrogen peroxide, shown in Figure 4. 11 , is formed during atrazine degradation 
(Hapeman et al., 1995). 
Second order rate constants for the reaction between atrazine and hydroxyl radicals, measured at different 
temperatures and pH v31ues. are listed in Table 4.4. 
Table 4.4: Se<:ond order rate COIlStants for tbe reaction of atrazine and hydroxyl radicals 
Reference Temperature pH Rate constant 
("C) (M"I S,I) 
(Beltnin et aI., 1993) 20 7 1,8 X 1010 
(Chramosta et al .. 1993) 20 7,5 1,7 x 10' 
(De Laat et 31., 1994) 20 7,5 2,4 x 10' 
(llaag and Yao, 1992) 24 3,_ 2,6 ± 0,4 x 10' 
The second order rate constants for the reaction. at 20 °C and a pH <= 7.5 to 8,1, of hydroxyl radicals with 
various atrazine degradation products are listed in Table 4.5. The lower rate constants for the reactions of 
dcethyldeisopropylatrazine and cyanuric acid with bydroxyl radicals indicate that these compounds are not 
readily oxidised by hydroxyl radicals (De Laat et 31. , 1994). 
Table 4.5: Rate constants for tbe reaction of b~'droI)'1 radicals with atrazine degradation products 
al 20°C and pH Q:I 7,5 to 8,1 (De Lot et al., 1994) 
Compound 
deethylatrazine 
deisopropylatrazine 
hydroxyatrazine 
deethyldeisopropylatrazine 
cyanuric acid 
Rate constant 
(M, I 5,1) 
1.2 X 109 
1.9 X 10' 
2,6 X 10' 
< 5 X 10' 
-< 2 x 10' 
The following advanced oxidation tcchnologies for atrazine degradation in water treatment have been 
reviewed, ozone in Section 4.3.1, ultraviolet radiation in Section 4.3.2 and uluasound in Section 4.3.3. The 
biological degradation of attazine in water treatment is reviewed in Section 4.3.4 . 
Cbapter 4 ATRAZINE CHEMISTRY 4-23 
4_3_1 Ozonation 
Trihalomethane formation in disinfection with chlorine in water treatment has lead to an increase in the use 
of ozone for disinfection (Adams and Randtke. 1992b). Ozone also significantly reduces the concentration 
of roicropoUutants (such as atrazine) in water (Lykins and. Koffskey, 1986; Lykins et al., 1986). Atrazine 
degradation products are fonned during ozonation since complete degradation involving ring cleavage does 
not occur (Hapeman, 1994; Legube ct al ., 1987; Nelieu ct al., 1996). The ozonation products of atrazine are 
shown in Figure 4 .12. 
Cl 
Cl I atrarine 
NAN 
H]CH~HNAN~NHCH(CHlh 
khloro--2,4-diaeetamid<>-
 S-triazin~ l i 
NJ-N 0 
H_N).,)lNACH 
- I' H 
4-acelllm ido-2-amino 
khloro-.Nriazine l 
Cl 
/ 
NJ-N 
H,N).,NJlNH, 
dcethyldcispropylatnuine 
-
 2-amin<>-6-chloro-4-hydroxy 
s-triazine 
4-amin<>-2.6-dihydroxy 
s-triazine 
Figure 4.12 : Degradation patb1\' a~' of tbe ozonalion of atrazine (Bapeman. 1994) 
Chapter 4 ATR.A2l1'\£ CHEMISTRY 4-24 
The four primary ozonation products of atrazine, shown in Figure 4.12, are deethylalraZine, 2-acetamido-
 6-<:h10r0-4-isopropylamino-s-triazine, deisopropy1atraz.ine and 4-acetam.ido-6-<:h1oro-2-ethy1amino 
s-tr1azine (Hapeman. 1994; Hapeman-Somich et al ., 1992). The three secondary products fonned are 
6-<;hloro-2,4-diacetamido-s-triazine. 4-amino-2-acetamid0-6-<:hloro-s -triazine and deetbyldeisopropyl 
atrazine. Extended ozonation of deethyldeisopropylatrazine leads to the stepwise formation of 
2-amin0-6-<:hJor0-4-hydroxy-s-Uiazine and 2-amin04,6-<libydroxy-s -Uiazine (Hapema..n. 1994; Hapeman 
Somicb et al ., 1992). 
Deetbylatrazine and deisopropylatrazine are in the same oxidation state as atrazine and are not oxidized 
by-products of atrazine despite the strong oxidizing power of ozone, the leaving side-<;hain alkyl group is 
oxidised and not the portion of the molecule containing the s-triazine ring (Adams and Randlke, 1992a). 
The amino alkyl groups are the first site of attack during ozonation. The alkyl group is either removed or 
convened to the N-acetyl, according to the degradation pathway shown in Figure 4. 12, with the s-triazine 
ring remaining intact (Hapeman, 1994). 
Atrazine degradation during ozonation is due to both reaction with hydroxyl radicals and the direct reaction 
with ozone. The reaction mechanisms illustrating the formation of hydrolysed and dealkylated degradation 
products due to reaction with hydroxyl radicals are presented, respectively, in Figure 4. 10 and Figure 4.11 . 
The mechanism for the direct reaction of atrazine with ozone is shown in Figure 4. 13. 
Cl 
o N~N + H20:2 
H C)lNA N)lNHCH(CH ,~ 
3 I In 
H 
2-acctamido-6-<:h1oro-
 4-isopropylamino-s-lriazinc 
• 
deethylatrazine 
Figure 4.13: Reaction mechanism for the formation of dealkylated degradation products from 
the direct reaction of atrazine with ozone at acidic pH (Bolzattbini et al •• 1994) 
Cbapter 4 ATRA2lNE CHEMISTRY 4-25 
The reaction of atrazine and ozone at acidic pH. as sbown in Figure 4.13, is initiated with the formation of a 
complex from the coordination of ozone to a side chain nitrogen. Transfer of the hydride from carbon to 
oxygen gives the hydrotrioxide which either undergoes N-dealkylation to form deethylat.raz.ine, acetaldehyde 
and hydrogen peroxide or rearranges to fonn 2-acetamid0-6-chlor0-4-isopropylamino-s-triazine and 
hydrogen peroxide (Bolzacchini et al ., 1994). Steric hindrance causes the formation of deethylatrazine to be 
5-fold greater that that of deisopropylatrazine (Bolzaccbini et al., 1994; Hapeman, 1994). 
The mixture of degradation products obtained from the ozonation of atrazine was dependant on reaction 
conditions, particularly pH, ozone concentration and contact time (Adarns et al., 1990). All the products 
identified in the investigation by Hapeman and co-workers contained a chlorine atom (Hapeman. 1994; 
Hapeman-Somich et al. , 1992). Hydrolysed degradation products from the ozonation of atrazine have been 
identified under certain reaction conditions (Adams and Randtke, 1992b; Legube et al., 1987; Nelieu et al ., 
1996). Adams and Randtke found that under conditions that prevented the autodecomposition of ozone to 
hydroxyl radicals, amy dealkylated products were formed and that under conditions that favoured the 
formation of hydroxyl radicals (such as high pH values), minor quantities of hydrolysed products were 
formed in addition to the dealkylated products (Adams, 1990; Adams and Randtke, 1992a). It was proposed 
that the hydrolysis of the CI-C oond was due only to reaction with hydroxyl radicals and nOI due 10 direct 
reaction wilh ozone (Adams and Randtke, 19923). The formation pathway of hydrolysed products during 
the ozonation of atrazine is shown in Figure 4 .14. 
atrazine hydroxyatrazine 
deethylhydroxyatrazine 
Figure 4.14: Fonnation pathway or hydroJy!led products during atrazine ozonatiOD (Adams 
and Randtke, 19920) 
Cbapter 4 A TRAlIl'I'E CHEMISTRY 4-26 
AlIazine, during ozonation, reacts either directly with ozone or with hydroxyl radicals formed from the 
decomposition of ozone. Rate constants, measured at different temperatures and pH values, for the reaction 
of atrazine with hydroxyl radicals are listed in Table 4.4 . Rate constants for the direct reaction between 
ozone and atrazine, at pH values of2, 7 and 12 are listed in Table 4.6. 
Table 4.6: Rate constants for the direct reactioD between ozone aDd atrazine at pH 2, 7 and 12 
(Beltr.in et al, 1993b) 
pH 
2 
7 
12 
4,5 
6,3 
6,3 
Atrazine degradation increases with increasing pH (Adams, 1990; Beltran et al., 1993b; Beltran et al., 
1994a). Hydroxyl ions at high pH values catalyse the decomposition of ozone, atrazine is degraded through 
reaction with the radicals formed from the ozone decomposition (Beltran et al., 1993b). The presence of 
hydroxyl radical scavengers, l-butanol and carbonatelbicarbonate ions, decreased the degradation of atrazine 
at pH values of 2, 7 and 12 (Beltran et al., 1994a). Atrazine is thus degraded by radical reactions even at 
low pH values. The direct reaction between atrazine and ozone at 20 cC and a pH of 2 represented 24 and 
81 %, respectively, of the total degradation in the absence and presence of 0,74 g L-I I-butanol (Beltrin 
et al ., 1994a). The higher rate constant for the reaction between atrazine and hydroxyl radicals (fable 4 .4) 
than that of atrazine and ozone (Table 4.6) indicates that atrazine degradation is more likely to occur via 
free radical reactions than by direct ozonation. Atrazine degradation during ozonation at a pH below 12 
occurs due to both direct reaction with ozone and free radical reactions whereas atrazine degradation at a 
pH above 12 is due only to radical reactions (Beltran, 1995). 
The reactions for the degradation of atrazine are 
A + 0 1 ~ producls 
O 
.~, 
A + H • -+ producls 
Scheme 4.1 
where kd and kHO.A are the rate constants for the reactions of atrazine wilh ozone and atrazine with hydroxyl 
radicals (Beltran et al., 1994a). 
Hydroxyl radicals are generated from the decomfX>Sition of ozone according to the mechanism shown in 
Scheme 3.3 in Section 3.1.3. Beltnin and co-workers modelled the rate of atrazine degradation using 
equation 4 .1 (Beltnin et al. , 1993b; Beltnin et al. , 1994a). 
Chapter 4 ATRAZlJIo"E. CH£MlSTRY 
-dCA 
dt 
4-27 
(4.1 J 
where CA, Co, and CHO represent, respectively. the concentrations of atrazine. dissolved ozone and hydroxyl 
radicals. The first term on the right hand side of the equation represents the contribution of atrazine 
degradation from direct ozone reaction and the second term the contribution from reaction with hydroxyl 
radicals. Equation 4.1 is only valid if the decomposition of ozone is slow relative to the ozone mass transfer 
rate (Beltnin et al .• 1 994a). The kinetic regime of ozone absorption is determined from the dimension1ess 
Hatta number for the reaction between ozone and atrazine; the Hatta number, as detailed in Section 3.1.2, 
relates the chemical reaction rate to the physical absorption rate of ozone (Be1tran et al., 1993b; Beltnin 
et al. , 1994a). Reactions undergone by gases dissolving in solution develop in the proximity of the 
gas-liquid interface (fast reactions), in the bulk liquid (slow or very slow reactions) or in between both zones 
(moderate reactions); the kinetic regime is classified as slow when the Ha number is less than 0,3 (Beltnin 
et al. , 1993b; Beltrim et al. , 1994a). The Hatta number for an irreversible second"rder reaction is defined 
as 
Ha 
_ (kdCADo. )o.~ 
- kc (4.2J 
where kd is the rate constant of the direct reaction between ozone and atrazine, CA the concentration of 
atrazine, Do, the ozone diffusivity in water and kL the liquid-phase mass transfer coefficient (Beltran el al ., 
1993b). The Hatta numbers for the direct reaction between ozone and atrazine at pH 2, 7 and 12 are listed 
in Table 4.7. 
Table 4.7: Hatta numbers and kinetic regimes of O'L.ODe absorption for the direct reaction between 
owne and atrazine at pH 2, 7 and 12 (Beltran et al., 1993b) 
pB 
2 
7 
12 
Ba 
o 
o 
o 
Kinetic regime 
very slow 
very slow 
very slow 
The direct reaction between ozone and atrazine at any pH value is a very slow reaction. thus. the 
degradation rate of atrazine is determined on1y by chemical reaction and is not limited by ozone mass 
transfer nor the ozone decomposition rate (Beltran et al. , 1993b; Beltnin et al., I 994a). Saturation of the 
bulk solution with ozone during ozonation also indicates that the ozonation of atrazine occurs by a set of 
slow gas-liquid reactions (Beltcln et aI. , 1994a). Both the direct reaction with ozone and the free radical 
reactions develop in the bulk solution (Beltnin et al. , I 993b) . 
Chapter 4 ATRAZlNI.: CHEMISTRY 4-28 
Atrazine degradation was not significantly affected by temperature (3 , 10 and 20 QC) for ozonation at a 
partial pressure of 1,1 kPa (Beltran et al., 1994a). Atrazine degradation increased with increasing 
temperature when ozonation was performed at partial pressures of 0,67; 0,84 and 1,1 kPa for temperatures 
of 3, 10 and 20 °C, respectively; the partial pressure was adjusted since ozone solubility, and thus the 
amount of ozone available in solution, decreased with increasing temperature (Beltran et al., 1994a). 
Duguet and co-workers found that an applied ozone dose of 5,1 mg L-1 caused a 46 % reduction in atrazine 
concentration from 0,35 to 0,19 )J.g L-t (Duguet et al., 1990). The addition of 2,8 mg L- I hydrogen peroxide 
increased the degradation to 84 % (Duguet et al. , 1990). Atrazine degradation is enhanced by the addition 
of hydrogen peroxide because of the increase in hydroxyl radicals (Paillard et al., 1991). Typical conditions 
applied during ozonation in water treatment plants are a maximum dosage of 5 to 6 g of ozone per m3 of 
water (equivalent 10 a concentration of 5 to 6 mg L- t ) and a maximum contact time of 20 min. Atrazine is 
not completely degraded under such conditions and the oxidation potential of ozone towards pesticides is 
improved with the addition of hydrogen peroxide (Paillard et al. , 1991). The combination of ozone and 
hydrogen peroxide in comparison with other advanced oxidation technologies (UV/D3 and UVIH~) is 
more efficient, costs less and is simple in application (Beluan et al. , 1998; Paillard et al., 1991). The 
reactions occurring during the oxidation of atrazine by ozone and hydrogen peroxide, Scheme 3.6 in 
Section 3. 1.3, are repeated in Scheme 4.2. 
Initiation stepS 
0 3 + HD~ !;. 0 ;- + HO; 
HO; -.,t 0;- + H+ 
H2D2 -.,t HO; + H· 
0 3 + HO; ~ D3~ + HOi 
Propagation steps 
pK = 1l,8 
k'· = 2 8 X 106 M"I s·' . , 
la] 
[h] 
Ic] 
Id] 
0 3 + 0;- ~ 03 ~ + O2 kl = 1,6 X 109 M t S-I le] 
0 ; - + H+ !;. HO; k2 = 5,2 X 1010 M-I S·l If) 
HO; ~ HO· + O2 k3 "" 1,1 x 10' M I S-l Ig] 
0 3 + HO· ~ O2 + HOi ~ = 2,0 X 109 M"I S-I [hI 
HO· + H,O, ~ 0 ;- + H,O + W k, : 2,7 x 10' M' s' li] 
HO· + HOi ~ Oi ~ + H20 14. = 7,5 X 109 M"I S-I U1 
Tennination steps 
HO· + B !~ .. BOJIid 
0 3 + B ~ B<WiJ 
Scheme 4.2 
[k] 
[IJ 
Chapter 4 A 'fRA1lI\o'1: CHEMISTRY 4-29 
The main initiation step in the oxidation of atrazine by ozone and hydrogen peroxide is reaction Id] of 
Scheme 4.2, the reaction between ozone and perhydroxyl ions (H<X), the ionic form of hydrogen peroxide 
(Beltcin et al. , 1998). Hydrogen peroxide does not act as an overall scavenging agent of hydroxyl radicals 
during ozonation, despite reactions [i] and [j] of Scheme 4 .2, since the superoxide radical anions produced 
in reactions [i) and [j] promote the further decomposition of ozone, regenerating hydroxyl radicals via 
reactions le] , tfl and [g] of Scheme 4.2. Reactions between hydroxyl radicals and atrazine intermediates 
also act as termination steps in the reaction mechanism shown in Scheme 4.2 (BeltrAn et al., 1998). Beltnin 
and co-workers modelled the degradation of atrazine, based on the mechanism in Scheme 4.2, using 
equation 4.3. 
-dCA 
dl 14.31 
where CA, Co, and CHO represent, respectively, the concentrations of atrazine, dissolved ozone and hydroxyl 
radicals (Beltr.in et aJ ., 1998). Beltcln and co-workers calculated, using equation 4.3, the contributions to 
atrazine degradation from hydroxyl radical reaction (reaction [k] of Scheme 4.2) and direct ozone reaction 
(reaction [I] of Scheme 4.2). Hydroxyl radical reactions accounted for 76,6 % of the total atrazine 
degradation for an atrazine concentration of 10,8 mg L-1 and an ozone concentration of 3,3 mg L-1. The 
addition of 0,34 mg Lol hydrogen peroxide increased the degradation due to radical reactions to 86 %; 
degradation was due only to radical reactions (90 to 100 %) for hydrogen peroxide concentrations above 
34 mg L·1 (Beltnin et al ., 1998). The deviation in data for the modelled degradation of atrazine and 
intermediates. deethylatrazine and deisopropylatrazine, was within 15 % of the experimental results. The 
kinetic model also predicted that the removal of atrazine, deethrlatrazine and deisopropylatrazine (of 
similar concentrations to that found in natural waters) wouJd occur in less than 7 min for oxidation with 
ozone and hydrogen peroxide even in the presence of bicarbonates or other hydroxyl radical scavengers 
(Beltcln et al.. 1998). 
Paillard and co-workers found that the optimum pH for atrazine degradation in pure water was between 
8 and 9 for ozone and between 7 and 8 for ozone combined with hydrogen peroxide (paillard et al ., 1991). 
Ozonation contact time (usually between 5 and 20 min) is determined by the required disinfection that has 
to be achieved during water treatment (Meijers et al. , 1993). Contact time for advanced oxidation processes 
depend on the rate of formation of hydroxyl radicals since, once formed, the radicals react with organic 
compounds in the water ""ithin 1 IlS to 10 ms time period. The estimated contact times for advanced 
oxidation with ozone and hydrogen peroxide at different pH values and ratios of ozone to hydrogen peroxide 
are listed in Table 4.8 (Meijers et al. , 1993). 
Cbapter 4 ATRA7.JNE CBEl'ofiSTRY 4-30 
Table 4.8: Enimated contact times fol'" advanced oxidation with OZODe and bydrogen peroxide 
(Meijers et at, 1993) 
pH 0] dosage BlOt dosage Estimated contact time for 95 % BO' 
(mg L-1) (mg L-1) formation (min) 
bubble column contact chambel'" 
7,8 3 1,5 4 
7,8 3 2 1,5 2 
7,8 3 4 1,5 
7,8 3 1,5 IS 
8,3 3 I ,S 0,5 
Radical formation for advanced oxidation at a pH of 7,8 01'" greater (as indicated in Table 4.8) occurs within 
minutes. Contact time for oxidation at a pH below 7,8 is longer since the rate of hydroxyl radical formation 
decreases by a factor of 10 for each pH unit decrease (Meijers et al., 1993). The estimated contact time for 
95 % hydroxyl radical formation in a contact chamber (as indicated in Table 4.8) decreases with increasing 
hydrogen peroxide concentration. The overall stoichiometry of the reaction of ozone with hydrogen 
peroxide, shown in Scheme 3.7 in Section 3.1.3 , indicates that 2 mol of ozone are consumed per mol of 
hydrogen peroxide (Beltnin et al., 1998; Meijers et al., 1993). 
Paillard and co-workers found that atrazine degradation in pure water was a maximum at the molar ratio of 
ozone to hydrogen peroxide of 2:1 for ala min contact time, 1,75 mg L·1 initial atrazine concentration, 
1,58 mg L-1 ozone concentration and pH of 7,5 (Paillard et al, 1991). Meijers and co-workers proposed 
that an excess of ozone (ratio S 0,4 mol of hydrogen peroxide p:r mol of ozone) would be required if 
disinfection was the primary objective during water treatment whereas an excess of hydrogen peroxide (ratio 
between 0,7 and I mol of hydrogen peroxide per mol of ozone) would be required for the degradation of 
pesticides by reaction with hydroxyl radicals. Higher concentrations of hydrogen peroxide lead to the 
scavenging of hydroxyl radicals b)' hydrogen peroxide (Meijers et al. . 1993). 
A ratio of 2,6 mol of ozone consumed per mol of hydrogen peroxide was found to be optimal for the 
degradation of atrazinc. deethylatrazine and deisopropylatrazine for the experimental conditions used by 
Belu-an and co-workers (Beltrin et al., 1998). Hydrogen peroxide concentrations below 0,34 mg L·] did not 
significantly influence the degradation of atrazine for an initia1 atrazine concentration of 1,70 mg L·1, ozone 
concentration of 16 mg L· ' , pH of 7 and temperature of 20 °C. The 2,6 molar ratio of ozone to hydrogen 
peroxide corresponded to 3 hydrogen peroxide concentration in the range of 3,4 to 34 mg L+]: alrazine 
degradation decreased at highcr concentrations of hydrogen peroxide (>340 mg L+l) since the kinetic regime 
of ozone absorption changes and is controlled by the mass transfer of ozone. the rate of atrazine oxidation 
becomes inversely proportional to hydrogen peroxide concentration (Beltnin ct al ., 1998). 
Chapter 4 A TRA7..lNE CHEMISTRY 4-31 
The degradation products generated from the oxidation of atrazine by ozone and hydrogen peroxide are the 
same as those (shown in Figure 4. 11) produced from the ozonation of atrazine (PaiUard et al. , 1991). The 
rate constants (at 20°C and pH 1) for the direct reaction between ozone and the primary intermediates, 
deethylatrazine and deisopropylatrazine, are 0,19 and 7,51 M'I 5,1, respectively (Beltnin et al., 1998). 
Degradation of these compounds is due only to radical reactions (90 to lOO %) for hydrogen peroxide 
concentrations above 340 mg L'I (Beltr.ln et al., 1998). 
Paillard and co-workers investigated a pilot plant treating water from the river Seine in France with ozone 
and hydrogen peroxide (Paillard et al., 1991). The addition of hydrogen peroxide, under continuous flow 
conditions, reduced the time to achieve maximum atrazine degradation from 8 min for ozonation to 4 min 
for oxidation with ozone and hydrogen peroxide. An ozone dose of 4 mg L'I was required to reduce the 
atrazine concentration from 1,1 to below 0,1 ~g L" in the outlet water (the required EEC threshold); the 
optimal ratio of ozone to hydrogen peroxide was shown to be between 0,5 and 0,64 mole of ozone per mol of 
hydrogen peroxide. Atrazine degradation decreased with increasing water alkalinity, degradation was 76 % 
with 150 mg L'I of calcium carbonate (CaCOJ) and 55 % with 300 mg L'I of CaC(); for an initial atrazine 
concentration of 2 ~g L" , 2 mg L" ozone concentration, ozone to hydrogen peroxide ratio of 0,35 mol mol" 
and pH of 8 (Paillard et al , 1991). Ozone dosage used on an industrial scale (maximum of 5 mg L'I) does 
not result in complete degradation of atrazine concentrations above 0,25 ~g L'! (Meijers et al ., 1993 ; 
Paillard et al., 199 1). The combination of ozone and hydrogen peroxide is an effective and easy 
implemented method for the degradation of atrazine at higher concentrations (Paillard et al., 1991). 
Treatment of Thames river water in the United Kingdom (alkalinity bet\\'een 200 and 300 mg L'I as CaC();; 
total organic cartx)O of 3 to 6 mg L' I) has indicated that between 30 and 60 % of atrazine is removed with 
ozone doses of 2 to 4 mg L'I and a contact time of 5 to 15 min (Foster et al., 1992: Rachwal et al ., 1992). 
Addition of hydrogen peroxide increased the atrazine removal by a further 20 to 30 %; the improvement 
was not consistent or predictable and depended on the presence of natural organics and free radical 
scavengers (bicarbonates) in the Thames river source water (Foster el ai. , 1992; Rachwal et al.. 1992 ). 
Pesticides. such as atrazine, and natural organic carlxm compounds, such as humic and fulvic acids. are not 
completely oxidised by ozone or ozone and hydrogen peroxide during water treatment. Intermediate 
oxidation and hydrolysis by-products tend to be more biodegradable than the original compounds. a 
biological process is required after ozonation to prevent the increased assimilable organic carbon levels 
from supporung growth of biofllms and microorganisms in the distrirution system (Foster et al .. 1992). 
Activated carbon has been used to remove ozonation by-products (Foster et aI., 1992). 
Activated carbon is a porous material (used in either the powder or granulated form) with a large surface 
area and a high affinity for organic compounds (Foster et al ., 1992). It is produced by controlled 
Chapter 4 A'I'RA.ZIJ\'E COEMlSI'R.Y 4-32 
combustion of coal, peat, coconut shells or wood and can be regenerated tbennally when the adsorption 
capacity fills . Regeneration is usually perfonned for only high value activated carbon. Disposal of spent 
activated carbon can be a problem since it is laden with pesticides and can be a potentially hazardous 
material for landfill purposes (Foster et al., 1992). Activated carbon is used primarily in water treatment for 
the removal of taste, odour, colour, organic micropollutants (such as pesticides) and biodegradable organic 
matter. Activated carbon adsorbs most compounds from water that are not strongly hydrophiHic in nature 
(Orlandini et al. , 1996). 
The dosage of granular activated carbon (GAC) required to achieve a given treatment objective is 
determined by capacity, adsorption kinetics (the rate of adsorption), competitive and displacement 
adsorption and the pre-adsorption of naturally occurring organic matter (Orlandini et al., 1996). Natural 
organic matter (NOM) in surface water consists mostly of substances that are less adsorbable and are of 
lower diffusivity than target micropoHutants such as atrazine. The zone in the filter bed in which the 
natura1 organic matter adsorbs is larger and moves faster through the bed than the zone in which the 
micropollutants adsorb. Natura1 organic matter thus tends to adsorp onto granular activated carbon ahead 
of micropollutants and reduces the adsorptive capacity of the GAC bed for micropollutants (Huang and 
Banks, 1996: Orlandini et al .• 1997). Biodegradation also occurs within granular activated carbon filters 
and contributes 10 the removal of organic compounds. Microbial activity is supported. by the biodegradable 
portion of the natura1 organic matter in the GAC filter influent (Orlandini et al ., 1996). 
The pilot-plant study by Foster and. co--workers found that the breakthrough for atrazine of a GAC bed with 
an empty bed contact time of6 nUn was extended from 3,5 months, without pre--ozonation. to 5 months with 
pre--ozonation with an ozone concentration of2 to 3 mg L'\ and a contact time of 5 min (Foster et a1.. 1992). 
Orlandini and co-workers also demonstrated the extension of the lifetime of a GAC bed with pre--ozonallon 
(Orlandini et al.. 1996: Orlandini et aI. , 1997). Atrazine breakthrough at a bed depth of 0.35 m in a GAC 
ftIter that had treated 51 000 bed volumes of water was 66 % when receiving non-ozonated influent and 
39 % when receiving ozonated influent. The GAC filter had a bed depth of 1.1 m and an empty bed contact 
time of 20 nUn. ozone concentration was 0,7 mg L' \ and atrazine concentration was 2 J.lg L·' . Similarly, 
atrazine breakthrough. at a bed depth of 1.1 m, after 17 000 bed volumes of water had been treated was 
6 and 3 %. respectively. for filters receiving non--ozonated and ozonated influent (Orland.i.n.i et al .. 1996). 
Rachwa1 and co--workers found that atrazine was still being removed to less than 0,1 ~g L' \ after 2 yr of 
operation of a pilot plant consisting of pre--ozonation, chemical dosage, flocculation, filtration. main 
ozonation and GAC filtration. Atrazine removal (initial concentration between 0,2 and 0,4 J.lg L' \) was low 
in the pre--ozonation stage, between 30 and 60 % occurring during main ozonation and the final reduction to 
below 0.1 ~g L" was due to adsorption in the GAC filter CRachwal el al. , 1992). Pre-monation increased 
the mineralisation of ring-labelled atrazine in GAC fillers (Huang and Banks. 1996). Approximately 62 % 
Chapter 4 ATR.AZJNE CHEMISTRY 4-33 
of the ring_UL- 14C atrazine was convened to 14C~ in the GAC filter receiving ozonated atrazine and 
ozonated surface water, a 50 % conversion occurred in the filter receiving ozonated surface water and 
untreated atrazine and only 38 % conversion occurred in the filter receiving untreated water and untreated 
atrazine (Huang and Banks, 1996). 
The increased attazine removal in GAC filters receiving O2onated inDuent rather than non..ozonated 
influent is due to increased biodegradation of atrazine, decreased competitive adsorption of attazine and 
natural organic matter and lower natuzal organic matter pre-adsorption (OrJandini et al., 1997). Ozonation 
lowers the adsorbability of natural organic matter with respect 10 activated carbon and increases the portion 
of biodegradable natural organic matter that is removed in the filter via biodegradation rather than by 
adsorption (Or1andini et al., ) 997). The combination of ozone and granular activated carbon effectively 
removes pesticides from surface water since compounds that are not fuUy oxidised by the ozone are readily 
removed by the granular activated caroon (Rachwal et al., 1992). 
4.3.2 lTItraviolet radiation 
The degradation of a substrate during UV radiation can be due to both direct and indirect photolysis 
(Torrents et al., 1997). Direct photolysis is the adsorption of UV-visible light by a substrate which then 
undergoes transformation. Indirect photolysis is the adsorption of light by a species other than the substrate 
which then either transfers the energy directly 10 the substrate (called photosensitization), undergoes 
electron transfer with the substrate or causes a series of reactions producing oxidants such as singlet oxygen, 
hydroxyl radicals and alkyJperoxy radicals (Torrents et al., 1997). Direct photolysis ofatrazine proceeds via 
excitation of the molecule followed by dechlorination and bydroxylation to produce hydroxyatrazine (Khan 
and Schnitzer. 1978: Nick et at.. 1992; ?ape and Zabile., 1970; Torrents et al .. 1997). Between 0.95 and I 
mol of hydroxyatrazine is produced per mole of atrazine photolysed (De Laat et aI., 1995). The mechanism 
for the direct photoi)'sis of atrazine. shown in Scheme 4.3, proceeds via the homolytic scission of the C-Cl 
bond reaction [b]. followed by electron transfer, within the solvent cage, of the two resulting radicals. 
reaction (c). and reaction of the carbocation with water, reaction rdl (Nick et al. , 1992). The term. ACt in 
Scheme 4.3. represents the atrazine molecule, it indicates the bond between the Cl atom in position 6 and 
the rest of the molecule. 
ACI + hv -- ACI'" 
ACI' - A' + Cl' 
A' + Cl ' - A' + Cl ' 
A' + H, O - AOH + H 
ACI' - A' + CI-
 A- + CI- - ACI 
Scbeme 4.3 
la] 
Ibl 
[cl 
Id] 
le] 
[D 
Chapter 4 ATRAIDIE CHEMISTRY 4-34 
Nick and co-workers proposed that the heterolytic scission of atrazine in the excited state, reaction le), 
could occur as an alternative to reaction fb] (Nick et al. , 1992). The carbocation may react with the chloride 
ion within the solvent cage, reaction If], in competition with reaction Id] to reform atrazine and reduce the 
quantum yield of product formation (Nick et al. , 1992). The quantum yield of a photochemical reaction is 
defined as the amount of reactant degraded or product formed per amount of photons absorbed by the 
system (Braslavsky and Houk, 1988). 
Beltnin and co-workers detected low concentrations of hydrogen peroxide during the photolysis of atrazine 
(Beltran et al. , 1993a). Hydroxyl radicals are formed from the photolysis of hydrogen peroxide, as shown in 
Scheme 4.4, and the formation of hydrogen peroxide could indicate that atrazine degradation was also due 
to radical reactions (Beltran et al. , 1993a; De Laat et aI. , 1997). The quantum yield of hydrogen peroxide 
(CPH202) at 254 run was 0,5 mol per photon (Beluan et al. , 1993a). 
H, O, + 2hv ~ 2HO 
Scheme 4.4 
The degradation of atrazine was unaffected by the presence of hydroxyl radical scavengers such as 
bicarbonate and carbonate ions. Beltmn and co-workers concluded that the concentration of hydrogen 
peroxide formed during the direct photolysis of atrazine was too 10w to compete with atrazine for the 
incident UV radiation (Beltr.in et al., 1993a). Atrazine degradation decreased when humic substances were 
present in the water and competed for the incident radiation (Beltrin et al. , 1993a). 
Atrazine concentration was unaffected at UV radiation levels that are sufficient for water disinfection 
(Nick et aI. , 1992). Atrazine was degraded 2,4 % at a UV fluence of 250 J m·l which reduces E. CQIi count 
by four orders of magnitude where fluence is defined as the tota1 radiant energy per cross sectional area of 
target (Braslavsky and Houk, 1988; Nick et aI. , 1992). A UV fluence of 24 kJ m-2 was required to reduce 
atrazinc concentration by onc order of magnitude (Nick et al., 1992). 
The presence of other species such as hydrogen peroxide or nitrates at sufficient concentration to absorb the 
incident radiation and produce hydro:\-yl radicals increases atrazine degradation (Arantegui et aI., 1995; 
Behean et aI. , 1993a; De Laat et al., 1997; Torrents et al ., 1997), The UV degradation of atrazine in the 
presence of hydrogen peroxide follows first order reaction kinetics (Arantegui et ai. , 1995: Burkhard and 
Guth, 1976; Khan and Schnitzer, 1978). Atrazine degradation does not occur with hydrogen peroxide in 
the absence of light or UV radiation (Ar3ntegui et al ., 1995: BeltmD et al., 1993a). The reactions De Laat 
and co-workers used to model the degradation of atrazine with hydrogen peroxide are listed in Scheme 4.5 
(De Laat et al .. 1997). 
Cbapter 4 A TRAZlNE CHEMISTRY 4-35 
Photodecomposition of hydrogen peroxide and production of hydroxyl radicals 
H20 2 + H20,.,. HO; + H30~ pKa = 11 ,7 (a] 
H20 2 + 2hv -+ 2HO' 
HO' + H20 2 ,.,. HO; + H20 
HO' + HO; oF HOi + HO 
[PER] ~ [H,o,] + [HOi] 
HOi + HOi -+ H20 2 + O2 
2H20 2 + 2hv -+ 2H20 + O2 
Reactions with hydroxyl radicals 
HO' + A -+ products 
HO· + Si -+ products 
CO]" + H10 '" ,.,. HCO; + H20 
HCO~- + H30 '" ... H2CO) + H20 
HO' + HCOj - COj- + H,O 
HO' + COl- - CO; - + . 
[C-J ~ [HCO, ] + [COl-] 
Photodecomposition of atrazine 
A + hv -+ products 
Scbeme 4.5 
cDH,o, = 0,5 
k = 27 X 10' M- I , -I H,O, , 
h--- = 75 x 10' M-I S-I nu, , 
kPEJt (M.I 5") 
overall reaction 
kA (M"' S·I) 
ok; (MI s·,) 
pKa, = 10,38 at 20 QC 
pKa2 = 6,39 at 20 QC 
knco; = 8,5 X 10' M-' S- 1 
k~. = 39 x 10· M-I S- 1 
• • 
kciDcq (M"I s·,) 
(<I>..) 
[b) 
Icl 
Id) 
le) 
In 
Ig) 
(h) 
li] 
UI 
(k) 
nJ 
ImJ 
In] 
10J 
The photodecomposition of hydrogen peroxide, reaction [b] of Scheme 4.5, produces two hydroxyl radicals 
per molecule of hydrogen peroxide decomposed Hydroxyl racticals. as shown in Scheme 4,5. initiate a 
ractical chain decomposition of hydrogen peroxide and react with atrazine. bicarbonate or carbonate ions 
and other inorganic and organic solutes (S;) in the water (De Laat et al., 1997), 
Dc Laat and co-workers used equation 4.4 to represent the degradation of atrazine by hydrogen peroxide 
and UV radiation in a batch reactor (De Laat et aI., 1997), 
- alA] 
dt = + kHO.A[A][HO')ss 14.4] 
where lA] is atrazine concentration, [HO' Jss the steady state concentration of hydroxyl racticals. <llA the 
quantum yield of atrazine at 253.7 nm. V the volume of irradiated solution. I. the imensity of absorbed 
radiation and kHO.A the rate constant for the reaction of atrazine with hydroxyl radicals (De Laat et al.. 
1997), The first term on the right hand side of the equation in equation 4.4 represents the decomposition of 
atrazine due to ctirect photolysis and the second term the decomposition by reaction with hydroxyl radicals 
Chapter" A1'RA..'lI]\o"E CRE.MISTRY 4-36 
generated from the photodecomposition of hydrogen peroxide (De Laat et al ., 1997). De Laat and 
co-workers used the model to estimate the value of the quantum yield of the photolysis of atrazine and to 
predict the effects of variable such as bydrogen peroxide dose, pH and bicarbonate alkalinity on the 
degradation of 3trnzine (De Laat et al., 1997). The quantum yield of atrazine, as calculated by different 
authors, is listed in Table 4.9. 
Table 4.9: Quantum yield of atmine 
Reference 
Beltnln et al., 1993a 
De Laat et al., 1997 
Nick et al. , 1992 
Wavelength 
(nm) 
25' 
253,7 
254 
0,05 
0,04 
0,05 
Hydrogen peroxide increases the degradation of atrazme because of radical reactions and the formation of 
hydroxyl radicals (Beluin et al ., 1993a). However, at high concentrations of hydrogen peroxide, hydroxyl 
radicals are scavenged by the hydrogen peroxide; for an atraz.ine concentration of 10,4 mg L" , hydroxyl 
radicals are scavenged by concentrations of hydrogen peroxide above 340 mg L-1 (Beltnin et al., 1993a). 
The oxidation of atrazine, even at high concentrations of hydrogen peroxide, is due to hydroxyl radical 
attack since the relative concentration of hydrogen peroxide to atrazine indicates that most of the incident 
light is absorbed by the hydrogen peroxide which produces hydroxyl radicals (Beltnin et al.. I 993a). 
Photo-oxidation of atrazine, in the presence of hydrogen peroxide, leads to the fonnation of 
hydroxyatrazine, deethyiatrazine, deisopropylatrazine and deethyldeisopropylatrazine (De Laat et al., 1997). 
The degradation pathway of the photolysis of atrazine in the presence of hydrogen peroxide (indirect 
photolysis) is the same as that for the ozonation of atrazine as shown in Figure 4. 11 . Products such as 
2-acetamid0-6<hlor04-isopropylamino-s-triazine and 2-acetamid0-6-chloro-4-ethylamino-s-triazine have 
been detected; 3 % hydroxyatrazine. 6 % 2-acetamid0-6-chlorD-4-isopropylamino-s-triazine, 20 % deeth}'I 
atrazine, 3 % 2-acetamid0-6-chlor04-ethylamino-s-triazine, 10 % deisopropylatrazine. I % 4-amino-2-
 acetamid0-6-chloro-s-triazine and 16 % deethyldeisopropylatrazine were detected for an 87 % conversion of 
atrazine (Torrents et al ., 1997). 
The UV degradation of atrazine has also been investigated with photosensitizers such as acetone and 
photocatalysts such as titanium dioxide and zinc oxide (Burkhard and Guth, 1976; Hustert et al.. 199 1: 
Pelizzetti et al. . 1990). The use of acetone as a photosensitizer increases the photodegradation of atrazine. 
Degradation follows first order reaction kinetics. Burkhard and Guth found that after a 6 h ex"POSUTe period 
30 % of the initial atrazine remained unreacted and 16 % deethylatrazine. 5 % deisopropylatrazine, 15 % 
deethyldeisopropylatrazine and 10 % hydroxyatrazine had formed (Burkhard and Guth. 1976). The 
degradation of atrazine during UV radiation (A > 290 run) in an aqueous titanium dioxide suspension 
Chapter 4 ATRAlINE CIlDUSTRV 4-37 
(250 rng L-I ) was found to follow first order reaction kinetics. the atrazine half-life was less than lO min 
(Hustert et al ., ] 991). The degradation rate of deethylatrazine, the primary metabolite of attazine in soil, is 
similar to that of atrazine. The photocatalytic degradation pathway of atrazine passes through 
deethyldeisopropylatrazine which degrades slowly to cyanuric acid.. Attazine is degraded faster by UV 
radiation in the presence of titanium dioxide than in the presence of hydrogen peroxide. The atrazine 
half-life was approximately 6 miD in the presence of titanium dioxide and 180 min in the presence of 
hydrogen peroxide under comparable conditions, 20 rng L-t atrazine concentration, 55 rng L· t hydrogen 
peroxide concentration and A. > 290 nm (Hustert et al., 1991). Pelizzetti and cxrworkers investigated the 
photocatalytic degradation of atrazine under simulated solar-irradiation cxmditions (Pelizzetti et al., 1987; 
Pelizzetti et al .• 1990). The spectral distribution of a 1 500 W Xenon lamp above 340 om simulates solar 
radiation (Pelizzetti et al. , 1987). The half-life of atrazine was approximately 5 and 80 min. respectively, in 
the presence of titanium dioxide and zinc oxide for an atrazine concentration of 25 mg L-t and a 
semiconductor concentration of 500 mg L- t . Atrazine degradation in 4 different soil slurries (2 000 mg L- t ) 
with titanium dioxide (5 000 rng L- I ) bad a half-life, depending on the nature of the soil, of between 10 and 
40 min (Pelizzetti et al., 1990). Degradation in the soils was slow in the absence of titanium dioxide; in one 
soil, atrazine concentration was reduced to approximately 40 % of the initial concentration after 7 d of 
irradiation (Pelizzetti et al. , 1990). Complete degradation of an aqueous solution of atrazine (5 mg L-t ) with 
titanium dioxide (100 mg L- I ) occurred within 1 h (Pelizzetti et al., 1987). 
Kearney and co-workers found that ozone together with UV radiation decreased the concentration of a 
10 and 100 rng L-t atrazine solution by 90 % in 1 h and a I 000 mg L-I within 4 h (Keamey et al ., 1987). 
Beltr.in and co-workers investigated the effects of parameters such as pH, temperature and the presence of 
radical scavengers on the oxidation of atrazine by UV/ozonation (Beltr.in et al ., 1 994b). Atrazine 
degradation increased with decreasing pH and increasing temperature (3, 10 and 20 °C). The ozone partial 
pressure was adjusted so that the ozone solubility in water was the same at the three temperatures. Atrazine 
degradation also decreased in the presence of bicarbonate and carbonate ions because of the scavenging of 
hydroxyl radicals (Beltr.in et al. , 1994b). 
The degradation of atrazine during UV/ozonation is detennined by three processes. the direct photolysis of 
ozone, atrazine and hydrogen peroxide; direct reactions of ozone with atrazine and hydrogen peroxide; and 
reactions between atrazine and radicals generated from the decomposition of ozone and hydrogen peroxide 
(Beltnin et al ., 1994b). The mechanism of the oxidation of atrazine by UV/ozonation consists of the 
reactions listed in Scheme 4.2 and the reactions listed in Scheme 4 .6 (Beltnrn el al., 1994b: Zwiener et al., 
1995), 
Cbapter 4 ATRAlD'OE CUMlSI1ty 
0 3 + H20 + hv ..... H20 2 + O2 
H20 2 + hv ..... 2HO· 
A + hv ..... products 
H 20? "* HOi + HO. 
HOi + 0 3 ~ H0 2 + 0 ;-
 0 1 + H" ..... HO· + O2 
H20 l + HO· ~ HO; + H20 
HOi + HO· ~ HO; + HO-
 Schcme4.6 
pK: 11 ,7 
kl = 2 ,8 X la' M"I S·I 
kl = 2,7 X 10' M"I S·I 
k 3 = 7,5 X 109 MI S·I 
4-38 
[at 
[h) 
[c) 
[d) 
le) 
[n 
[g) 
[h) 
The direct photolysis reactions are listed in reactions la] to le1 of Scheme 4.6. Hydrogen peroxide in neutral 
solution dissociates to the hydroperoxide ion, reaction Id] of Scheme 4.6. The reaction initiates further 
decomposition of ozone to produce hydroxyl radicals, reactions le] and [I]. Reactions ofbydrogen peroxide 
with h)'droxyJ radicals, reactions [g) and [bI. are of low imponance because of the low concentration of 
hydrogen peroxide and high pK value (Beltnin et al., 1994b). 
The primary quantum yields of ozone, hydrogen peroxide and atrazine, reactions [a]. [b] and [cl of 
Scheme 4.6, are 0,62; 0 ,5 and 0,05 mol per photon, respectively (Beltrin et al., 1994b). The extinction 
coefficients al254 run of hydrogen peroxide and ozone, 18,6 and 2900 L mor l cm·I , respectively, indicate 
that since hydrogen peroxide is the least light absorbing species (lowest extinction coefficient) the formation 
of hydroxyl radicals from reaction [b) of Scheme 4.6 would be significantly lower than that formed through 
reactions [d] to [f] (Bcltrin et al ., 1994b; Zwiener et al ., 1995). Atrnzine degradation is due mainly to 
reaction with hydroxyl radicals. followed by direct photolysis and the lowest contribution due to direct 
ozonation (Belu-an et al.. I 994b). 
Both Beltnin and co-workers and Zwiener and co-workers represented the degradation of atrazine by a ftrst 
order differential equation with three terms representing each of the contributing processes. ozo03tion, 
photolysis and photolytic oxidation (Beltcln et aI., 1994b; Zwiener et al. , 1995). The equation proposed by 
Bcltrlln and co--workers is listed in equation 4.5. 
- dCA 
dl [4.5) 
The first term on the right hand side of the equation represents degradation due to direct reaction with 
ozone in the liquid phase, the third term the photolytic oxidation due to reaction with hydro",-")'I radicals. 
The second term lA represents degradation due to direct photolysis and is calculated using equation 4.6. 
Chapter 4 ATRAnNE CHEMISTRY 4-39 
(4.6( 
where <l>A is the quantum yield of atrazine, 10 the effective intensity of the incident radiation in water, L the 
effective path of radiation, &; and Cj the extinction coefficient and concentration of a species i present in the 
water andfA the fraction of radiation absorbed. by atrazine (Beltrtn et al. , 1994b). 
Beltran and co-workers compared various oxidation technologies for the degradation of atrazine, the 
degradation rate increased in the order: direct photolysis < ozonation < UV radiation with hydrogen 
peroxide < UV radiation with ozonation (Beltdn et al., I 994b). Atrazine degradation in the presence of 
hydroxyl radical scavengers, such as bicarbonate or carbonate ions, is the same for direct photolysis, 
ozonation and UV radiation with hydrogen peroKide and is greatest for UV radiation with ozone (Beltran 
et al., 1994b). 
Zwiener and co-workers investigated the degradation of atrazine and deethylatrazine by UV radiation with 
ozone (ozone doses of 0 ; 2,2; 4,4; 6,7; and 8,9 mg L-1 and a UV radiant power of 42,5; 54,6; 72,25 and 
85 W L- 1) in a large-scale water treatment plant (Zwiener et al. , 1995). Atrazine of concentration 
0,28 Jlg L·I was degraded in a 70 m3 h-I flowrate of raw water to below the EEC threshold limit for 
pesticides (0,1 Jlg L·I ) at ozone doses of 4,4; 6,9 and 8,9 mg L·1 in combination with a radiant power of 
between 42,5 and 85 W L·1• A radiant power greater than 72,25 W L·1 was required to reduce the ozone 
concentration in the outlet water to the allowed threshold limit for ozone (0,05 mg L- I ) in clrinking water. 
The concentration of deethylatrazine, at a flow rate of70 m3 h-1, was reduced from 0,6 to 0,25 Jlg L·1 for an 
ozone dose of 8.9 mg L-1 and a UV radiant power of 85 W L- 1• The outlet concentration of deethylatrazine 
was only reduced to the EEC limit of 0, I Jlg L·1 at an ozone oose of 16 mg L-1 and a radiant power of 
85 W L'\ when the hydraulic residence time in the UV radiation unit was increased by decreasing the water 
flow from 70 to 30 m3 h·1 (Zwiener et al. , 1995). 
4.3.3 Ultrasonic degradation 
The advanced oxidation tc:x::hnology, ultrasound, has also been investigated for application in water 
treatment. Investigations by Koskinen and co-workers and ?etrier and co-workers have indicated that 
atrazine is degraded by ultrasound (Koskinen et al., 1994; petrier et al., 1996). 
Atrazine degradation followed first order reaction lcinetics with a rate constant of 2,10 x 10'3 min,1 for 
sonication of a 0,67 mg L'\ atrazine solution, at 30 0 C, with a Heat-Systems Ultrasonics model W-225R 
sonicator operating at a ]X>wer ofbehveen 70 and 80 Wand a frequency of 20 kHz (Koskinen et al. , 1994). 
A 200 mL sample volume was sonicated in a 250 mL beaker. The solution pH did not change significantly 
Chapter 4 A TRA11NE CIlEMlSTRY 4-40 
during sonication, after 120 min, pH had changed from 7,64 to 7,45 (Koskinen et al., 1994), The rate of 
atrazine degradation increased with increasing frequency~ the initial degradation rate was 7,8 times greater 
at 500 kHz than at 20 kHz for sonication at 20 QC with an acoustical JX)wer of 18,5 W (Petrier et al., 1996). 
Hydrogen peroxide formation was also greater at 500 kHz than at 20 kHz. The concentration of an atrazine 
solution (22 mg L'\ initial concentration) was reduced within 80 mm, during sonication of a lOO mL sample 
at an acoustical JX)wer of 18 W and frequency of 500 kHz, to below the detection limlt (0,02 mg L"\) of the 
system (Petrjer et al ., 1996). 
Sonochemlca1 degradation products (deethylatrazine, deisopropylauazine and deethyldeisopropylatrazine) 
resulted from dealkylation of atrazine similarly to that occurring during the oxidation of atrazine by ozone 
or photocatalysis with titanium dioxide (PWier et al ., 1996). The intermediate, 2,4-diamin0--6-<:hloro 
N-aceto-N ' -{l-methylethyl)-1 ,3,5-triazine, was also detected. This intermediate precedes the fonnation of 
deethylatrazine and characterises the reaction of hydroxyl radicals with the carbon atom adjacent to a 
secondary amlne (De Laat et al., 1995). Sonochemical degradation of atrazine is due to reaction with 
hydroxyl radicals, the presence of n-pentanol (a radical scavenger) during sonication totally inhibited the 
degradation of atrazine (Petrier et al., 1996), The higher hydrogen peroxide formation at an ultrasonic 
frequency of 500 kHz accounted for the greater atrazine degradation at that frequency since more hydroxyl 
radicals would be present (Pettier et al. , 1996). The reactions reported by Petr1er and co-workers to occur 
during the ultrasonic degradation of atrazine are listed in Scheme 4 .7" 
Initiation step 
H20 m H" + HO" 
O2 !?l 20· 
Reactions in the cavitation bubble 
H· + O2 .... HOi 
0+ H,O ~ 2HO· 
2HO· ~ H,O + 0 
20· ~ 0, 
HO; + HO' .... H20 + O2 
Reactions in the interfacial area of a cavitation bubble 
2HOi .... H20 2 + O2 
2HO· ~ H,O, 
A + HO· .... products 
Scheme 4.7 
la] 
[b] 
Ic] 
Id] 
le] 
ID 
Ig] 
Ih] 
liJ 
til 
The degradation of atrazine is initiated by the homol)1ic scission of water and oxygen, reactions la] and (b] 
of Scheme 4 .7. Water and oxygen are refonned from the radical reactions (reactions (CI IO Ig]) occurring in 
Cbapter 4 ATRAZIN£ CUMlSTR.Y 4-41 
the cavitation bubble. Hydrogen peroxide and products from the reaction of hydroxyl radicals and atrazine 
are produced in the interfacial area of a cavitation bubble and in the bulk solution. Atrazine degradation is 
related to the availabiJity of hydroxyl radicals escaping the cavitation bubble (Pettier et al. , 1996). 
4.3.4 Biological treatment 
Atrazine is not readily degraded. in activated sludge systems during water treatment (Dries et al., 1987; 
Hogrefe et al., 1986; Meakins et al., 1994; Nsabimana et al, 1996). Meakins and co-workers investigated 
the behaviour of atrazine and its primary degradation products, deethylatrazine and deisopropylatrazine, 
during bench-scale primary and secondary wastewa1er treatment (Meakins et al., 1994). The triazine 
compounds remained in solution and were not removed during primary sedimentation by adsorption onto 
settleable solids. Secondary treatment with biologically active mixed liquor (of different suspended solids 
concentration) removed between 9 and 39 % of the ttiazine compounds. The reduction in concentration 
was not a biological process since similar results were achieved with mixed Jiquor in which the microbial 
activity had been suppressed by the addition of a 1 % (w/v) sodium azide solution and reduction in 
temperature to 4 °C. The percentage removal of triazine herbicides with mixed liquor, as sho\\ll in 
Table 4.10, increased with increasing suspended solids concentration (Meakins et al., 1994). Reduction in 
triazine concentration during secondaJy wastewater treatment was thus due to physical adsorption onto 
suspended solids (Meakins et al., 1994). 
Table 4.10: Removal of triazine herbicides with mixed liquor with different concentrations of 
Alspended solids (MewD. et aL, 1994) 
Suspended solids (mg L-I ) 
atrazine 
deethylatrazine 
deisopropylatrazine 
o 
~.9 
4,9 
-6.5 
Triazine berbicide removal (%) 
1900 2200 2700 
12,5 18,6 35.5 
9,5 16,5 21,0 
15,4 17,2 39. 1 
The maximum removal of deisopropylatrazine (39 %) indicated that the bulk of triazine compounds would 
pass through a sewage treatment process and be discharged in the effluent into the environment . A 
conventional two-stage sewage ueatmem process is thus ineffective in substantially reducing the 
concentration of triazine compounds (Meakins et al., 1994). 
Nsabimana and co-workers investigated the effects of atrazine on the microbiology of the activated sludge 
process since atrazine is toxic to organisms such as bacteria and fungi in soil and algaes and shrimps in 
aquatic ecosystems (Nsabimana et al. , 19%). Six bacterial strains were isolated from a sludge sample as 
representative of activated sludge microflora. Growth of the bacterial strains was unaffected by atrazine 
concentrations of I and 10 mg Lol; an atrazine concentration of]OO mg L-t retarded the growth of one of the 
Chapter 4 A'TRA.2'.D"'I'E CHEMISTRY 4-42 
bacteria] strains, The total bacterial count and biomass (measured as mixed liquor volatile suspended 
solids) were reduced in a laboratory pilot plant that was fed continuously with an atrazine concentration of 
either I mg L'\ for 6 d or 20 mg L'\ for 7 d Chemical oxygen demand (COO) was increased after 22 h for 
an atrazine concentration of 20 mg L,1 , Sludge metabolic activities such as dehydrogenation, proteolyis, 
nitrification and denitrification were unaffected by atrazine (Nsabimana et al ., 1996). 
Volskay and co-workers investigated the toxicity of various organic compounds (including atrazine) to 
activated sludge microorganisms by measuring the concentration that reduced the oxygen consumption of 
the organisms by 50 % in comparison to a control (Volskay and Grady, 1988; Volskay et al" 1990). 
Oxygen consumption was reduced by 5 % for an atrazine concentration (33 rng Lol) equal to its solubility 
limit (Volskay and Grady, 1988; Volskayet al., 1990). 
Hogrefe and co-workers developed a small-scale system that was capable of degrading 80 % of triazine 
compounds in real wastes (Hogre.fe et al ., 1985). The process consisted of a mixed bacterial culture 
(aerobic) maintained at 37 DC in an open, unsterilised system. An added source of carbon, such as glucose, 
in a molar ratio of carbon to nitrogen of ID was required for microbial growth, Microbial growth and thus 
atrazine degradation was retarded by salt concentrations abo\'e 3 % . A residence time of 40 h (i.e. a 
dilution rate of 0,025 hot) was necessary to prevent washout. The dilution rate was too low for industrial 
applications and a fluidised bed reactor, in which the organisms were immobilised on a layer of sand. was 
developed (Hogrefe et al., 1986). The biomass concentration, of between 12 and 18 g L'I volatile suspended 
solids, in the reactor was ID-fold higher than that in the continuous culture system; a lower residence time 
(25 h versus 40 h) was also achieved. The system recovered readily from process perturbations. Aerobic 
conditions and an added carbon source were required to achieve maximal removal (approximately 80 %) of 
triazine compounds (Hogrefe et al ., 1986), 
A mixed microbial culture has also been used to degrade atrazine in soil and water systems (Grigg et al., 
1997), Degradation in soil. after 100 d. was 78 and 21 %. respectively. for atrazine concentrations of 
10 and 50 g L'I ; degradation in water, after 80 cl, was 90 and 56 %. respectively. Lower degradation of the 
50 g L't atrazine concentration was attributed to phosphorous depletion. Degradation of these atrazine 
concentrations indicated that mixed microbial cultures could be used for atrazine bioremediation at 
cont.amination concentrations, typically bern'een 0,02 and 4ID mg L'I , repon.ed at agrochemical mixing and 
loading facilities (Gngg et al., 1997), 
Microorganisms, in the biological transformation of pesticides and other compounds in natural waters. use a 
variety of electron acceptors such as oxygen, nitrate. sulfate and carbon dioxide (Wilber and Parkin, 1995), 
The dominant electron acceptor condition can affect the rate of transformation of compounds. Wilber and 
Chapter 4 ATRAZlI'o'E CHEMISTRY 4-43 
Parkin investigated the biotransformation of atrazine uncler aerobic, nitrate-reducing, suIfate-reducmg and 
methanogenic conditions (WiJber and Parkin. 1995). Biotransformation occurred under all electron 
acceptor conditions. The relative rates of transformation decreased in the order methanogenic 
> sulfate-reducing > aerobic > nitrate-reducing conditions, however, the differences in rates were not 
statistically significant at a 95 % confidence limit. Atrazine was transfonned as a secondary co--metabolic 
substrate; acetate was fed as a primary substrate and. transformation, under aerobic and. nitrate-reducing 
conditions, stopped once the acetate bad. been depleted (Wilber and. Parkin, 1995). 
Ozonation has been used as a pretreatment for soil biodegradation of atrazine because of the enhanced 
biodegradability of the oxidation by-products (Kearney et al. . 1988). Keamey and co-workers developed a 
two-chamber unit consisting of ozonation and subsequent circulation through a biologically active soil 
column for on-site treatment of herbicide wastewater (Keamey et al., 1988; Somich et al., 1990). 
Degradation of atrazine in the soil column after SO d, as shown in Table 4.11, increased with increasing pH 
of ozonation. 
Table 4.11: Effect of ozone pretreatment on atrazine biodegradation in soil columns after SO d 
(Kearney et aL, 1988) 
without ozonation 
ozonation at pH 6,5 (unbuffered) 
ozonation at pH 8 
ozonation at pH 10 
Percentage atrazine degradation 
(%) 
< 20 
55 
65 
75 
Atrazine solutions, after ozo08tion at pH 10, were neutralised. before being added to the soil colwnn of the 
unit developed by Kearney and co-workers. The microbial population of the soil column was fortified with 
Pseudomonas strain A since different strains of Pseudomonas have been identified that degrade atrazine 
(Bebki and Khan, 1986; Shapir and Mandelbaum, 1997). Atrazine metaoolism was rapid in the first 3 d, 
slower over the following 7 d and levelled off at approximately 60 % after 15 d of treatment. Indigenous 
soil microorganisms, without the addition of Pseudomonas strain A, degraded atrazine more slowly, a 40 % 
degradation was achieved after 15 d (Keamey et aI., 1988). 
Ozonation and metabolism in a biologically active soil column were used to treat the pesticide waste and 
rinsate, a mixture of several pesticides, from a small farm (Somich et al .. 1990). Degraclation kinetics of 
atrazine in solution with other pesticides did not follow first order reaction kinetics as did degradation in 
pure solution. The lag in degradation was attributed to preferential reaction of ozone with other compounds 
such as formulating agents (Somich et al ., 1990). Atrazine concentration, in 114 L batch experiments, 
Chapter 4 A TRAZINE CHEMISTRY 4-44 
decreased from 17,2 mg L-1 in the untreated pesticide waste to 8,8 mg L-1 after O2Onation and 5, 1 mg L-1 
after ozonation and circulation through the soil column (Somich et al. , 1990). 
Bioassays were conducted to determine the toxicity of by·products produced from treatment of the pesticide 
waste and rinsate (Somich et al., 1990). Wheat, a monocotoyledon. and soybean, a dicotyledon. were 
treated 2 wk post~mergence with either distilled water, untreated pesticide waste, ozonated pesticide waste 
or pesticide waste that had been O2Onated and circulated through the soil column. Leaves of the soybean 
plants that had received untreated pesticide waste were withered and discoloured after 1 wk; all other plants 
appeared normal. Wheat plants showed no visible differences after 3 wk; soybean plants that had received 
untreated pesticide waste were stunted relative to the control, while the plants that had received either of the 
treated solutions were larger and healthier. lbis was attnbuted to the bioavailable nitrogen in the oxidation 
by· products. Toxic effects were not observed in wheat or soybean plants that received untreated or treated 
pesticide waste (Somich et al., 1990). Ames tests using several strains of Salmonella typhimurium and one 
strain of Escherichia coli , with and without metabolic activity, showed no evidence of mutagenic activity in 
untreated pesticide waste, ozonated pesticide waste or pesticide waste that had been O2onated and circulated 
through the soil column (Somich et al ., 1990). 
Oxidation of atrazine in the binary treatment system consisting of ozonation and biomineralisation 
produced chlorodiamino--s-triazine as the fina1 ozonation product (Leeson et al .. 1993). This compound is 
used as a nitrogen source by the microbial species in the soil column. however, agricultural wastes usually 
contain other nitrogen sources such as ammonia fenilisers . An ammonia concentration of 0.14 g LO ] 
inhibited the biodegradation of atrazine by Pseudomonas strain A The microbial strain Klebsiella 
terragena (strain DRX·l) which preferred an organic nitrogen source and was able to tolerate high 
ammonia concentrations was isolated from a sewage sludge (Hapef!Wl et al., 1995; Leeson et al.. 1993). 
Degradation of chlorodiamino--s·triazine by Klebsiella terragena required an added carbon source (corn 
syrup) and occu.rred regardless of the presence of ammonia. Complete mineralisation of 
chloroctiamino--s· triazine occurred within 3 d in the presence of either a 0.07 or 0, 14 g L-1 ammonia 
concentration: 40 % degradation occurred within 10 d in the presence of a 14 g L' ] ammonia concentration. 
The complete utilisation of chlorocliamim)+triazine in the presence of ammonia concentrations of 0.07 and 
0, 14 g L' ] indicated that Klebsiella te"agena had a strong preference for chlorodiamino--s ·triazine as a 
nitrogen source and that both the nilrogen atoms of the amino groups and the triazine ring structure were 
being utilised as nitrogen sources (Leeson et aI. , 1993). Degradation of chlorocliamino--s·triazine occurred 
via the intermediate aminochlorohydroxy-s·triazine (Hapeman et al.. 1995). Pilot-scale trials of the 
ozonation and biomineralisation process of pesticide wastes were performed in a 200 L system (Hapeman 
et al .. 1995). Formulated atrazine was mineralised by Klebsiella terragena with and without ammonium 
nitrate (NH..NDJ) , a fertiliser often found in agriculturaJ waste. with respective ozonation efficiencies of 
0.0088 and 0.0108 mol of atrazine per mol of ozone (Hapeman et al.. 1995). 
Chapter 4 ATRAZINE CHEMISTRY 
4.4 CONCLUDING REMARKS 
Atrazine is a herbicide that acts due to the inhibition of photosynthesis and is thus less toxic to animals than 
to plants. The effect of atrazine on bwnans has not been fully investigated though it appears to be 
minimally toxic during inhalation, eye and skin oontact and for ingestion in small quantities. 
The major ooncem with atrazine is its persistence in the environment, its half-life in soil varies between 
14 and 109 et. it has been found to persist from year to year in soil and water. It is the most frequently 
detected herbicide in ground and surface waters in Europe and the United States. The presence and 
persistence of atrazine in potable water sources has resulted in various countries regulating or banning the 
use of atrazine. Health standards in the US have been set for atrazine at 0,3 IJ.g L·1 and in Europe at 
0, 1 ~gL·I . 
Atrazine in soil is degraded or adsorped. into soil particles. Degradation is through hydrolysis to form 
hydroxyatrazine or via N-dealkylation to form deethylatrazine and deisopropylatrazine. Dealkylation is 
usually due 10 rnicrobial action whereas hydrolysis is an adsorption-cata1ysed chemica1 reaction. Further 
degradation is via deamination and ring cleavage. The degradation products are more biodegradable than 
atrazine. 
Groundwater contamination occurs due to atrazine leaching through the soil. Deethylatrazine and 
deisopropylauazine are the main degradation products thal leach because of their lower retention in soil. 
Fwther degradation does not continue in groundwater due to the lower temperatures and microbial 
popuialions. Alrazine accumulation in surface water is due to runoff from agricultural fields during rainfall 
and irrigation. Deethylatrazine and deisopropylatrazine are also formed from the degradation of other 
triazine herbicides such as simazine, cyanazine and propazine. 
Chlorine, chlorine dioxide and chloramine have no effect on atrazine during water treatment. Atrazine is 
also not readily degraded in an activated sludge system or affected. by Oocculation and coagulation in a 
clarifier. Nanofiltration removes between 80 and 90 % of atrazine. Atrazine concentration is reduced by 
about 83 % during ozonation and between 95 and 97 % with the use of granular acth'ated carbon. 
Pretreatment with ozone improves atrazine degradation in an activated sludge system (due to the enhanced 
biodegradabili~' of the degradation products) as well as the length of time before atrazine breakthrough 
occurs in activated carbon filters. Atrazine degradation during ozonation is improved with the addition of 
hydrogen peroxide. Atrazine is also degraded by ultrasound 
The nature of the ozonation degradation products depends upon reaction conditions. although. ring cleavage 
does not occur. Degradation is due to both reaction with ozone and hydroxyl radica1s. The direct reaction 
with ozone produces onJy dealkylated products whereas dealkylated and. hydrolysed products are fonned 
Chapter 4 ATRA2lNE CBEMlSTRY 
from tbe reaction witb hydroxyl radicals. Hydrogen peroxide is formed during the production of dealkylaled 
products. The higher rate CO'II5tant indicates that the reaction witb hydroxyl radicals is faster than the direct 
reaction with ozone. Both reactions occur in the bulk solution and are not Limited by mass ttansfer. 
5 
EXPERIMENTAL DESIGN 
The advanced oxidation pnx:esses, ultrasound and ozone, have been investigated. The equipment designed. 
and used during the investigation, as well as, the analytical procedures followed are described in this 
chapter. The ultrasonic equipment is detailed in Section 5.1 , ozonation equipment in Section 5.2, 
experimental procedures in Section 5.3, analytical procedures in Section 5.4 and the statistical software 
program in Section 5.5. 
5.1 ULTRASOUND EQUIPMENT 
The investigation of the sub-processes occurring during sonication and the degradation of atrazine was 
performed in an ultrasonic laboratory reactor (ultrasonic cell). The cell was designed and operated using an 
ultrasonic horn for an energy source. The ultrasonic horn is descnbed in Section 5.1.1 and the ultrasonic 
cell in Section 5,1.2. 
5. t.1 Ultrasonic horn 
A 20 kHz Ultrasonic process system, model PlOOI2-20, was obtained from Sonic Systems (postal address: 
Unit 3. Monks Dairy, Isle Brewers. Taunton, Somerset. TAJ 6QL, UK). The system. shown in Figure 5. 1. 
consists of a generator (model PJOO-20), transducer assembly (model 23820-AS) and an high intensity 
ultrasonic horn (pan no. A93655) with replaceable tip (part no. A93656-2). 
~Ie 
ultrasonic 
generator 
selecto,~, -t':::-;:;:;--( 
~ -I-~ 
analogue I=.!=---':~V 
mete< variable power 
control 
Figure 5.1: Ultra.50nic PI"'OCe5S system 
"'""<iuc" 
housing 
ultrasonic 
hom 
Cbapte r 5 EXPERIMENTAL DESIGN 5-2 
The system operates at a nominal frequency of 20 kHz. The vibration amplitude of the transducer and the 
electrical power to the transducer are detennined by means of a variable control mounted on the front panel 
of the generator assembly. Amplitude and electrical output by means of a selector switch are read from an 
analogue meter, amplitude off the green scale and electrical power off tbe red scale. The amplitude of the 
peak to peak transducer displacement is calibrated in micrometres, the meter range is 0 to 15 }lm. 
Electrical power is measured in Watts. the meter range is 0 to 100 W. The mass of the generator assembly 
is 4,5 kg, the width, depth and height dimensions are 321 , 175 and 110 mm, respectively. The generator 
assembly is designed for continuous operation without fo rced air cooling (Sonic Systems, 1994). 
The pre-stressed piezoelectric transducer is mounted in a sealed housing and is designed for continuous 
operation at ambient temperature (5 to 40 QC). The mass of the transducer assembly is 0,332 kg. A 
schematic diagram and a picture of the transducer assembly and ultrasonic horn are presented in Figure 5.2. 
dimens 
I 59 I I' 124.1 .[ 
87 
57 
38 
71 
ions in mm 
-
 ....... 
30 
<a) 
;- 12 ,5 
Figure S.2: Ultrasonic horn 
transducer 
housing 
upperfLxed 
horn element 
ultrasonic 
horn 
replaceable 
tip 
(b) 
The horn and replaceable tip are manufactured from a titanium 3110y (Ti-64) containing 6 % of an 
aluminium alloy and 4 % vanadium (Perkins, 1997b). The horn is coupled to the transducer assembly by a 
stud that is screwed into the transducer. The working surface of the horn is eroded during sonication. 
Erosion of the horn is prevented with the use of a replaceable tip screwed into the end of the horn. 
Replacement of a tip due to erosion is less expensive than replacement of the entire horn. Approximately 
Chapter 5 EXPt:RIMENTAL Dt:SIGN 5-3 
I mm of the tip can be machined without significantly changing the total length of the born and thus 
affecting the resonant frequeDcy of the system (Perkins:, 1997a). The surface of the tips used during the 
investigation were machined once the electrical power drawn by the transducer for a preset vibrational 
amplitude began to fluctuate. A tip was replaced when the machined surface had also eroded. The effect of 
machining the surface of a born tip on the performance of the ultrasonic system was evaluated by measuring 
the formation of hydrogen peroxide generated by a new tip and comparing it to that generated after 
machining (Section A 1. 1 of Appendix A). The data presented in Figure A 1 indicates that machining the 
surface of a horn tip does not affect the performance of the system. 
The Ultrasonic process system supplied by Sonic Systems has a powr-by-<kl1lQlJd characteristic, the 
generator automatically delivers the required electrical \X)Wer to maintain a preset transducer disp/iacement 
amplitude irrespective of the load conditions (Sonic Systems, 1994). Tbe acoustic power uansferred to a 
sample is calculated (equation 2. 1) from the difference between the electrical power supplied to the 
transducer when the born is pIaoed in a sample and the electrical power supplied to the transducer when the 
horn is without load (Sonic Systems, 1994). The born is defined as being witbout load when the tip rests in 
air and is not submersed in any sample. 
A . (W) ( electrical power supplied to ) ( electrical power supplied to ) cousuc power '" -transducer when loaded transducer when unloaded 12.11 
The calculation of acoustic power is based on the assumption that the losses in the system are only 
mechanical and displacement-related. electrical losses are low and can be ignored (Perkins., 1997a). The 
losses in the system are mechanical hysteresis losses in the transducer stack (both in the piezoceramic and at 
the interface j oints), dynamic frictional losses in the transducer compression bolt threads and mechanical 
losses in the transducer and horn materials (Perkins, 1997a). The total electrical power (as indicated on the 
power meter) supplied to the transducer when it is unloaded is used to overcome these mechanical losses at 
a set transducer displacement amplitude. The mechanical losses are unchanged when the transducer is 
loaded and operated at the same displacement amplitude, the measurement indicated on the power meter is 
thus the sum of the acoustie mechanical losses and the acoustic power supplied to the reaction solution. It 
follows that the acoustic: power supplied to a treatment sample is. the difference between the loaded 
transducer power and the unloaded transducer power (Perkins, 1997a). 
The efficiency of the energy transformation of the ultrasonic: horn was determined by comparison of acoustic 
power with electrical power. Acoustic: power is the energy delivered into a reaction solution, electrical 
power is the energy drawn at the wall. The electrical power supplied to the transducer of the ultrasonic: 
horn was read from the meter on the face of the ultrasonic generator when the transducer was loaded. The 
acoustic power was calculated from the difference in electrical power su~jed to the transdlK:er when 
loaded and without load (Sonic Systems, 1994). The calculation of acoustic power was confirmed using 
Cbal,ter 5 EXPERIMENTAL DESlGN 5-4 
calorimetry (described in Section A1.2). The average acoustic power measured during sonication at each 
transducer displacement amplitude is compared in Table 5.1 with the average electrical power. 
Table 5.1: Conversion of eledrical power to acoustic power 
Transducer Average acoustic Average electrical Energy 
displacement power power conversion 
(»m) (W) (W) (%) 
5 24,2 28,2 86 
8 39,3 46,3 85 
11 56,7 68,7 83 
Acoustic powers generated during sonication with tbe ultrasonic horn at different transducer displacement 
amplitudes are recorded in Table 5.1. The conversion of electrical to acoustic power by the ultrasonic horn 
is approximately 85 %. 
5.1.2 Ultrasonic cell 
An ultrasonic laboratory reactor (ultrasonic cell) was designed and constructed to investigate the free radical 
reactions occurring during sonication and to apply in a case study of the ultrasonic degradation of a model 
organic compound The u1trasonic cell is shown in Figure 5.3. 
Figure 5.3: Ultrasonic cell 
Chapter 5 ExpEJUMENTAL DESIGN 
The ultrasonic cell (manufactured from 316 stainless steel) is 195 mm high (from the base to the top of the 
lid) and the external diameter of the water jacket is 114 mm. Other dimensions of the cell are presented in 
Figure 5.4. A 500 mL sample can be sonicated in the cell. Temperature is measured by a thermocouple in 
a well centred at the bottom of the cell and is regulated by the circulation of cooling water in the water 
jacket. A schematic diagram of the cell is shown in Figure 5.4. 
139 
74 
60 
seal for 
trBruidUCC2' 4 J 
bo"""", +~~~-j 
to rest on 
"",ling 
wole< 
inkt 
114 
""'pm>< """"" 
sample gas port for gas sampling 
feed inlet 15 e:ntnmces ....--
 6 "g~ -!. 0b -born 
= ...=:I:!:--T I. 139 .1 "'.....,. 
~ 71 
Wet 
12 
"",ling 
WBl<r 
outlet 
-
 sample 
dreID 
195 
dimensions in mm 
ultnl90nic oeI.llid 
Figure 5.4: Schematic diagram of ultrasonic cell 
The sample inlet pan, gas inlet and outlet parts and a neoprene septum for gas sampling are positioned, as 
shown in Figure 5.4, in the lid of the cell. A liquid sample can be saturated with gas during sonication by 
bubbling the gas through the stainless steel diffuser in the cell. 
The parameters that affect cavitation and thus the outcome of a sonochcmica1 exp!riment are listed in 
Table 2.5 in Section 2 .5.5. These parameters include the characteristics of the reaction medium (viscosity, 
vapour pressure, presence of solid particles), reaction conditions (pressure, temperature, gas sparging) and 
the type of ultrasonic system (power, frequency, reactor size and geometry). These parameters were taken 
into account during the design of the ultrasonic cell and are discussed in the following subsections. 
Cbapter 5 ExrERIMENTAL DESIGN 
5.1.2.1 Frequency 
Ultrasonic experiments are usually performed at frequencies of between 20 and 100 kHz. the frequencies 
used in different ultrasonic investigations are reported in Section 2.5.5 . The ultrasonic intensity required to 
initiate cavitatioo increases as frequency increases (Mason, 1993; Pestman et al ., 1994). Commercial 
ultrasonic equipment operate at a fixed frequency (determined by the transducer) and optimisation in 
regards to frequency cannot be performed The frequency of the Ultrasonic process system supplied by 
Sonic Systems is 20 kHz (Sonic Systems, 1994). 
5.1.2.2 Acoustic power 
Ultrasonic effects are determined. primarily by the amount of acoustic power delivered to a system. Acoustic 
power is also IqX)rted in literature as intensity (or power density), either, in terms of the surface radiating 
area of the wtrasonic source or the volume of liquid sonicated. (fable 2.6 in Section 2.5.5). Perkins defined 
intensity as the electrical power supplied to the transducer divided by the mating surface area of the 
ultrasonic horn (Peridns., 1990). 
The acoustic power delivered by the Ultrasonic process system from Sonic Systems was varied by changing 
the displacement amplitude of the transducer from 0 to 12 j.UD.. The wtrasonic effects, in 500 mL of water, 
increased with increasing power. Experiments were performed at a transducer displacement of 11 }.UIl and 
not at 12 j.UD. because the maximum displacement vibration began to decrease as the horn tip eroded 
Experiments were also performed at displacement amplitudes of 5 and 8 f.Ull when the effect of acoustic 
power was being investigated.. Typically, 69 W of electrical power was required by the transducer to 
generate a displacement of It IllD when the wtrasonic horn was loaded with 500 mL of water. An electrical 
power of 12 W was required by the transducer to generate a displacement of 11 j.UD. when the horn was 
without load Hence. the acoustic power delivered to 500 mL of water for a transducer displacement of 
amplitude of 11 j.UD. was 57 W. Ultrasonic intensity calculated in terms of the radiating surface area of the 
horn (1,25 cm diameter: 1.23 cm2 radiating surface area) was 46, 1 W cm·2. The radiating surface area was 
calculated from re x (radius)2. Ultrasonic intensity calculated in terms of the volume of liquid sonicated 
(500 mL) was 0.114 W cm"l. Acoustic power and ultrasonic intensity (in terms of both area and volume) of 
the Ultrasonic process system for sonication of a 500 mL sample at transducer displacement amplirudes of 
5, 8 and 11 jim are reported in Table 5.2. 
Chapter 5 EXPERIMENTAL DESIGN 5-7 
Table 5.2: Acoustic power and iDtensity of the Ultrasonic process system 
Transducer displacement Acoustic power Ultrasonic intensity Ultrasonic intensity 
(j>m) (W) (W cm-1) (W cm"') 
5 24,2 19,7 0,048 
8 39,3 32,0 0 ,079 
11 56,7 46, 1 0.114 
The ullrasonic intensity at the railiating horn surface of the Ultrasonic process system, shown in Table 5.3. 
is comparable with that of other investigations using an ultrasonic horn. 
Table 5.3: Comparison of tbe ultrasonic intensity of the Ultrasonic process system with that of 
other sonocbemical investigations using an ultrasonic horn 
Reference 
(Pugin, 1987) 
Investigation 
enhancement of reaction between 
lithium and l-bromopentane 
(Donaldson et al., 1979) polymerisation of nitrobenzene 
(Price el al., 1994) degradation of aromatic compounds 
this in\'estigatioD degradation of atrazine 
(Suslick et al., 1989) effect of ultrasound on nickel and 
copper powders 
(Serpone et al. , 1994) degradation of chlorophenol 
compounds 
(Ratoarinoro et al ., 1992) enhancement of reaction between 
ethyl malonate and chalcone 
(Kolronarou et al.. 1992b) degradation of hydrogen sulfide 
.5.1.2.3 Horn sbape 
Frequency Ultrasonic intensity 
(kHz) (W cm-1) 
20 24 
20 25 
22 39 
20 46 
20 50 
20 52 
20 57 
20 75 
The vibrating motion generated by a transducer is magnified. by the horn (Goodwin., 1990; Perkins, 1990). 
Different shapes of ultrasonic horns and the magnification obtained with each shape is discussed in 
Section 2.5 .3. Materials of construction are also discussed in Section 2.5.3. Two stepped ultrasonic horns 
manufactured from a titanium alloy were obtained with the Ultrasonic process system. The low intensity 
horn had an upper and lower diameter of 3,0 and 2,0 cm, respectively, and the high intensity horn an upper 
and lower diameter of 3,0 and 1,25 cm, respectively. The magnification of transducer displacement 
calculated for the low and high intensity ultrasonic horns was 2,25 and 5.76, respectively. Magnification 
for stepped horns is calculated from the square of the ratio of upper to lower horn diameter (Perlcins, 1990). 
Cbapter 5 ExPERlMr.NTAL DI.SIGN 
The low intensity born was only used during preliminary experiments (reported in Section A 1.3) to 
determine acoustic intensity in sample volumes from 100 to I 000 mL. The high intensity horn was 
incorporated in the design of the ultrasonic cell all experiments in the cell were performed with the high 
intensity horn. 
5.1.2.4 Volume 
PreliminaJy experiments using the low intensity ultrasonic born (Section AI.3 of Appendix A) indicated 
that the initial rate of hydrogen peroxide formation i.ncreased with increasing ultrasonic intensity and hence 
decreasing sample volume for a range of water volumes from 100 to 1 000 mL. Gondrexon and co-workers, 
in an investigation of the sonocbemica1 degradation of chlorophenol, found that the rate of degradation 
decreased with increasing sample volume for the range of 100 10 400 mL when the acoustic power input 
was 60 W (Gondrexon et al., 1993). Sonochemica1 effects decrease with increasing sample volume for a set 
acoustic power input. The volume of a sample to be sonicated in the ultrasonic ceU was fixed at 500 mL. 
The volume bad to be large enough so that the kinetic information would be applicable for process sca1e-up 
but also small enough so that the ultrasonic intensity would be sufficient to cause sonocbemicaJ effects. The 
sample volume could also not be too big since as discussed in Section 2.5.3 a small zone of high intensity is 
created around the tip of an ultrasonic born and intensity decreases with increasing distance from the horn 
tip (Goodwin, 1990; Pugin. 1987; Suslick. et al., 1984). The internal diameter of the ceU is 8 , I cm and the 
height of a 500 mL sample in the ceU is thus 9,7 cm. The diameter was chosen so that the sample volume 
would not be flat and wide which would have created deadspots in the liquid on either side of the horn tip. 
The height of the cell was ca1culated. such that the tip of the born wou1d be approximately 1 cm below the 
surface of the liquid and would not be exposed even with the vigorous liquid movement and splashing that 
occurs during sonication. The total volume of the ultrasonic ceU is approximately 900 mL. 
The volume of a sample (500 mL) that can be sonicated in the ultrasonic cell is compared in Table 5.4 with 
sample volumes used in other sonochemical investigations. The comparison of ultrasonic intensity in 
Table 5.3 indicates that the Ultrnsonic process system operates with a comparable intensity at the radiating 
surface area of the horn with that of other sonochemical investigations, however, as shown in Table 5.4 the 
greater sample volume in the ultrasonic ceU results in a lower volume-based intensity. 
Chapter 5 EXPERIMENTAL DESIGN 5-9 
Table 5.4 : Comparison of the sample ,'o)ume that CaD be sonicated in tbe ultrasonic cell with 
sample volumes of otber sonocbemical investigations 
Reference Investigation Sample volume Ultrasonic intensity 
(mL) (W cm"') 
(Cost et al., 1993) degradation of p-nitrophenol 25 2 
(Kotronarou et al., 1992b) degradation of hydrogen sulfide 25 1,7 
(Ratoarinoro et aI. , 1992) enhancement of reaction between 57 1.3 
ethyl malonate and cha1cone 
(Serpone et aI., 1994) degradation of chlorophenol lOO 0,5 
compounds 
(Pugin, 1987) enhancement of reaction between lOO 0,3 
lithium and l-bromopentane 
(Price et al., 1994) degradation of aromatic 120 0.25 
compounds 
(de Visscber et al., 1996) degradation of monocyclic 150 0,097 
aromatic compounds 
(Petrier et al., 1994) degradation of phenol 200 0,15 
(Donaldson et al., 1979) polymerization of nitrobenzene 250 0,28 
this iovestigation degradatioo of atraziDe SOD 0,11 
(Lorimer et al., 1991) decomposition of potassium I 200 0,025 
persulphate 
(Bhatnagar and Cheung, degradation of chlorinated C I and 2000 0,1 
1994) C2 volatile organic compounds 
5.1.2,5 Vessel type 
Ultrasonic horns are used. to sonicate samples, described in Section 2.5.3. in beakers. cuphoms or flow cells. 
Sample volumes sonicated by ultrasonic horns, indicated in Table 5.5. typically range from 25 to 1 400 mL. 
Flow cells incorporating an ultrasonic horn as the source of ultrasonic energy allow for sonication in both 
balch and continuous mode, the sonicated zone of the system used by Cheung and Kurup to study the 
degradation of chlorofluorocarbon compounds was 35 mL when used in batch mode and 250 mL when used 
in continuous mode (Cheung and Kurup. 1994). Lin and co-workers used a circulating system in the stud)r 
of the degradation of 2<hloropbenol, the 1 L sample volume ",as circulated at a flow rate of 0.5 L min'l 
through the sonication cell (Un et al., 1996). Flow cells are manufactured. indicated in Table 5.5, from 
either glass or 316 stain1ess steel. The ultrasonic cell was designed as a closed batch system (to allow for 
saturation with different gases) that could be modified later to be used in continuous mode. It was 
manufactured from stainless steel so that the material of construction in a scale-up exercise ~'Ould be the 
Chapter 5 EXPERIMENTAL DESIGN 5-10 
same to that used in a pilot-plant system. The ultrasonic ceU is compared in Table 5.5 with vessel types and 
volumes of other sonochemica1 investigations using an ultrasonic horn. 
Table 5.5: Comparison of tbe ultrasonic cell witb vessel types and volumes of other SODocbemical 
investigations using an ultruooic born 
Reference Investigation 
(Kotronarou et al. , 19913) degradation of parathion 
(Serp:me et al., 1994) degradation of chlorophenol 
compounds 
(Price et al. , 1994) degradation of aromatic 
compounds 
(de Visscher et al. , 1996) degradation of monocyclic 
aromatic compounds 
tbis investigation degradation of atrazine 
(Lorimer et aI., 1991) decomposition of potassium 
persulphate 
Sample 
volume 
(mL) 
25 
100 
120 
150 
1200 
V .... I 
volume 
(mL) 
V .... ltype 
50 stainless steel cell with 
150 
500 
200 
900 
1400 
water jacket 
glass cell with water 
jacket 
centrifuge tube immersed 
in cooling bath 
glass vessel with water 
jacket 
stainless steel cell with 
water jacket 
flange top, round 
bottomed glass flask 
The connection between an ultrasortic horn and the wall of a flow cell (Figure 2.13 in Section 2.5 .3) can be 
sealed with either an O-ring or by screwing the cell onto an external thread at me nodal point on the horn 
(Goodwin. 1990). Suslick and co-workers used flow cells that were sealed with a Tenon collar and 
gas-tight O-rings in various sonochemica1 investigations (Suslick et aI., 1983; Suslick, 1990). The flow cell 
used by Cheung and Kurup in the investigation of me degradation of chlorofluorocarbon compounds was 
sealed by threading the reactor body onto an external thread on the transducer housing of the ultrasonic 
horn (Cheungand Kurup, 1994). 
The ultrasonic horn was initially inserted through a hole in the lid of the ultrasonic cell that was sealed with 
an D-ring. The contact between the horn and the lid of the cell dampened the movement of the horn and 
reduced the transfer of acoustic power to a sample in the cell. Acoustic power transferred to a water sample 
in the cell was calculated to be approximately 17 % of that transferred to a 500 mL water sample in a beaker 
(detailed in Section A 1.4). 
The cell lid was changed such that the transducer housing fitted into a raised support, shown in Figure 5.4, 
to prevent contact between the horn and the lid. A rubber seal coated wim Vaselinee was used to ensure a 
Chapter 5 EXPERIMENTAL DESIGN 5011 
gastight fit between the transducer housing and the cell suppon. Elastic bands were also used to hold the 
transducer housing down and exen a slight pressure OD the seal . The connection between the transducer 
housing and cell suppon was checked using soapy water for leakage when gas experiments were perfonned 
in the cell. 
5.1.2.6 Pressure 
External applied pressure is one of the paramelers thal determines the level of cavitation in a system 
(Section 2.2. 1). Henglein and Gutierrez found that the yield of iodide oxidation decreased with increasing 
pressure from 100 to 300 kPa at acoustic powers of 30 and 50 Wand reached a maximum before decreasing 
at powers of 80 to 150 W (Henglein and Gutierrez, 1993). Chendke and Fogler found that the intensity of 
sonoluminescence from an aqueous carbon tetrachJoride solution varied non-linearly with pressure, local 
maxima 0CCll1'J'ed at 608 and 1 216 kPa (Chendke and Fogler, 1983a). 
The connection between the transducer housing of the horn and the lid of the uluasonic cell, as described in 
Section 5.1.2.5. did not allow for pressure control in the cell . Pressure is not usually controlled in Oow 
cells. Kotronarou and co-workers during the investigation of the ultrasonic oxidation of hydrogen sulfide 
noted that pressure was not controlled in the Oow cell and that initial headspace pressure was atmospheric 
(Kotronarou et aI., 1992b). 
5.1.2.7 Mixing 
Shock \\"ave formation during sonication, (discussed in Section 2.3.2), results in efficient mixing. Goodwin 
noted that high-intensity uJuasound delivered by ultrasonic horns may provide sufficient agitation so as 10 
eliminate the need for stirring (Goodwin, 1990). 
The agitation induced in 500 mL of \\"ater by the ultrasonic horn was investigated (described in 
Section A.I .5) using IXltassium permanganate crystals. A comparison of the mixing in the water after 5 s of 
sonication and after 10 min of the control without sonication is shown in Figure 5.5. The efficient mixing 
shown in Figure 5.5 indicated that additional facilities for stirring in the ultrasonic cell did not have to be 
included. 
Chapter 5 EXPERIMENTAL DESIGN 5-12 
<a) (b) 
Figure 5.5: Mixing of potassium permanganate crystals in 500 mL of wa.ter during sonication 
and in the control without sonication; (a) after 5 s of sonication and (h) the control after 10 min 
5.1.2.8 Temperature 
Ultrasonic experiments are affected by temperature, generally, the lower the temperature the greater the 
sonochemical effect (Mason et aI. , 1992). The temperature of a sample also increases duriog sonication and 
must be controlled for repeatable and consistent results. The intensity of ultrasonic cavitation in water is a 
maximum at 35 °C, cavitation intensity is within 70 to 100 % of the maximum in the temperature range 
20 to 50 °C (Niemczewski, 1980). Sonication in the ultrasonic cell was performed at 25 ± 1 QC. Ultrasonic 
experiments are typically performed at 25 QC (Kotronarou et al., 1992b; Lin et aI. , 1996; Mead et aI., 1975; 
SehgaJ et al., 1982), other temperatures reported in literature include 15°C (SusLick et aI., 1989), 20Q C 
(Donaldson et aI., 1979~ Price et al., 1994) and 30°C (Kotronarou et aI. , 1991 ; Kotronarou et aI., 1992a). 
Temperature was regulated in the ultrasonic cell by circulating cooling water through a water jacket. Flow 
cells used in ultrasonic experiments, as noted in Table 5.5, usually incorporate water jackets for the control 
of temperature. The cold water bath from which the cooling water was circulated was maintained below 
17 °C (usually between 10 and 15 0c) by a Labotec refrigeration unit (model no. FTC 300). The cooling 
water bad to be cooler than 17 QC for the temperature in We cell to be maintained at 25 QC. The 
refrigeration unit and pump (Linle Giant pump, catalogue no. 582013) that circulated the water through the 
refrigeration unit were set on a timer (National TB 179) that typically switched the unit on for 2 h and off 
for 3 h when overnight experiments were being perfonned. A schematic diagram of the cooling system is 
shown in Figure 5.6. 
Chapter S Exrf.ItIMI..NTAL DI'.SIGN 
, 
"""""nic 
,ell 
thermocouple 
[!~[}------------j 
__ 
controller 
cold water 
baIh 
refrigeration 
unit 
Figu.re 5.6: Schematic diagram of the cooliag system of tbe uIt:ruODic cell 
S-u 
The temperature in the ultrasonic cell was measured by a thermocoupl.e positioned in the centre. just above 
the base of the cell. The pump (Gormann Rupp Industries) that circulated the cooling water through the 
water jacket was controlled by a Fuji PYZ 7 digital controller. The pump was automatically switched on if 
the temperature in the cell rose above 25°C and switched off if the temperature fell below 25 cc. 
5.1.2.9 Gas saturatiOll 
The presence ofa gas during sonication, as discussed in Section 2.2.1, provides additional nucleation sites 
for cavitation. The nature of the gas, as discussed in Section 2.3 . 1.2, influences the radical reactions that 
occur during sonication. A diffuser was included in the ultrasonic cell to allow for saturation of liquid 
samples with different gases. The gas inlet pipe. as shown in Figure 5.4. entered through the lid of the cell. 
The diffuser l\"35 screwed into the bonom of the gas inlet pipe and was positioned in the lower half of the 
liquid sample. The diffuser was made from an appro,omate 1,5 cm piece of porous steel pipe. Gas "''as 
sparged continually through the liquid samples in the ultrasonic cell during sonication since liquids are 
degassed by ultrasound (Mason, 1990). To ensure saturation, gas "'"35 also sparged for 10 min prior to 
sonication through the liquid samples. Saturation periods reported in literature arc 5 min (Misik et al.. 
1995; Young, 1976). 10 miD (Donaldson et al., 1979; Mead et al., 1976; Sehgal and Wang, 1981), 15 min 
(Hart and Henglein, 1985) and 20 miD (Gutierrcz et al .. 1987; Rassokhin et al., 1995). A schematic 
diagram of the gas lines is shown in Figure 5.7. 
Chapter 5 ExPUlIMUlTAL DESIGN 
gas 
cylin<ler 
ro""""", 
ultrasonic 
cell 
Figure 5.7: Schematic diagram of'the gas IiDes 
5-14 
conical flask 
Gas Oow rate was oontroUed by the rotameter shown in Figure 5.7. The calibration of the rotameter is 
descnbed in Section A. 1.6. A flow rate of 6.0 mL S·l was used to ensure gas saturation in the liquid samples 
but without excessive splashing. Gas flowrates reported in literature vary between 0.25 and 31,7 mL S·I 
depending on the volume of sample (Hua et al .• 1995a; Hua et al., 1995b; Rassokb.in et al .• 1995; WeissJer. 
1959). 
5.1.2.10 Sampliag 
TIle ultrasonic ceU was designed such that samples could be taken for analysis in three ways. A neoprene 
septum, as shown in Figure 5.4, allowed for gas sampling during the sonication of volatile compounds. 
Liquid samples (up to 5 mL) were also taken through the septum using a syringe with a long need1e. This 
allowed for continuous sampling during an experiment without disturbing the gas atmosphere in the cell. 
Liquid samples (up to to mL) were sucked up using a pipette through the sample inlet port in the lid of the 
cell. Larger samples (SO to 200 mL) were coUected through the outlet drainage port at the bottom of the 
cell . The fina1 two methods of sampling could not be performed continuously but only once an experiment 
had been completed and the ultrasonic ceU was opened. 
5.2 OWNE EQUIPMENT 
Ozone was generated using a Sorbios generator model GSG 001.2 that was loaned by Umgeoi Water. The 
generator produces ozone according to the corona discharge method, described in Section 3.2.2. and is air 
cooled. The ozone·genefating element consists of an AL-:03-cerantic tube with an outerg.rounded electrode 
and an inner net·like counter electrode (Sorbios, 1993). The oxygen used as feed gas has to be free of 
bydrocarbon compounds and have a dew point below -40 °C (Sorbios. 1993). 
The mass of the generator assembly is 8 kg, the width, height and depth ' dimensions are 360, 200 and 
360 mm. respectively. A rotameter with regulating valve ror gas flow adjustment and a variable diaJ to 
Chapter 5 ExrERIMENTAL DESIGN ~15 
regulate power inpJt are mounted in the front panel of the generator assembly. Analogue meters to record 
current and pressure are also mounted in the front panel of the generator assembly. Ozone production is 
determined by gas flow (production increases with decreasing flow) and by power inrut that can be varied 
between 0 and 220 V. The generator produces ozone at a nominal rate of I g b·\ with a concentration of 
50 g m,3 for an oxygen flow rate of I L b'\ (Sorbios, 1993). 
Ozone experiments were performed in a fumebood due to ozone toxicity. The ultrasonic cell was used as 
the ozonation chamber and was moved into the fumebood. Ozone was bubbled through a sample using the 
existing gas diffuser in the cell. Typical volumes of ozonation chambers recorded. in literature are 200 mL 
(Andreozzi et al , 1989), 250 mL (Upbam et al., 1997), 500 mL (Siddiqui, 1996; SoIeIo et al., 1987) and 
1000 mL (Beltrin, 1995; Perkows1O et al., 1996). 
The gas lines, shown in Figure 5.7, were modified to allow for both ozonation and non-ozonatioD gas 
experiments. A schematic diagram of the complete experimental setup is shown in Figure 5.8. 
I!iIS 
cylinder 
rotameter 
ozonation 
experimemal setup 
ozone 
generator 
OOD-Ozone gas experiments 
ultrasonic 
cell 
to annospb:::re 
oonical flask 
ulnasonic 
cell 
Figure 5.8: Schematic diagram of the ozonatioo experimental setup 
to veI1l 
2%Kl 
trap 
to vent 
2%Kl 
uaps 
Thc experimental setup, shown in Figure 5.8, is similar to that used in the investigation by Singer and Zilli 
of the ozonation of ammonia (Singer and Zilli, 1975) and in the disinfection investigation by Upham and 
Cbapter 5 ExPERIMENTAL DESIGN S-16 
co-workers (Upbam et al., 1997). A bypass line allowed for the ozooe-containing gas to be divened when 
the ultrasonic cell was opened for a sample to be taken. The bypass line and the exhaust gas line from the 
ultrasonic cell were passed through 500 mL gas washing bottles containing a 2 % (20 000 mg L·I ) 
potassium iodide solution. The p:>taSSium iodide reacted with any remaining ozone before the gas lines 
ended near the top of the fumebood The solutions changed from colourless to yellow with the iodine that 
was liberated during the reaction with ozone. A 2 % potassium iodide solution was used to degrade 
unreacted ozone in the investigations by Smith and co-workers (Smith et al. , 1992) and Yolk and 
co-workers (Volk et al., 1997). 
The Dow rate of oxygen (medical grade; 99,5 %) was controlled by the rotameter in the gas-line and the 
rotameter in the ozone generator. Oxygen Dow rate was maintained at 2,4 mL S·I, oxygen DOWl3tes 
recorded in literature vary between 1,7 and 17 mL S·I (Andreozzi et al., 1989; K.oga et al., 1992; Singer and 
Zilli, 1975; Sotelo et al. , 1987; Trapido et al., 1997). Oxygen Dow was always started prior to the generator 
being switcbed on at the start of a day and stopped after the generator bad been switched off at the end of a 
day. The generator was switched on and allowed to run for over an hour 10 warm up and stabilise before 
experiments were staned 
The ozone generator was characterised, described in Section A2, by measuring ozone production at 
different voltages (voltage couJd be adjusted between 0 and 220 V). A voltage setting of 130 Y was used as 
a standard voltage for ozone experiments. The concentration of ozone generated in the oxygen stream at 
voltages between 100 and 200 Y is shown in Figure 5.9. 
40 
2 E3J -
 c 
Jl 
~ 20 -
 • g 
8 
• 6 10 r 
Ii 
o 
80 
• • 
100 120 
• 
• 
• 
• 
140 
"" 
lOO 220 
Voltage (V) 
Figure 5.9: Ozone concentration in a 2,4 mL Si oxygen gas stream generated by a Sorbios 
O'LOne generator at voltages between 100 and 200 V (11 "" 3 per voltage) 
Chapter 5 Expt:RIMENTAL DESIGN S-17 
The average ozone concentration, the weight percentage of ozone in the oxygen/ozone gas stream and the 
ozone production rate generated by the Sorbios ozone generator at voltages of 100, 130, 150 and 170 V for 
an oxygen flow rate of 2,4 mL S-I are presented in Table 5.6. 
Table 5.6: Average ozone rooct:ntratioa and weigbt percentage of ozone iD a 2,4 mL S·I oxygen gas 
stream generated by the Sorbios ozone geaerator at voltages of lOO, 130, ISO and 170 V 
Voltage 
(V) 
100 
130 
150 
170 
Ozoae 
coocentratioo 
(mg L-l) 
1,2 
5,7 
14,6 
20,7 
Ozone weigbt {}rone 
percentage productioo rate 
(0/.) (mg sol) 
0,1 0,003 
0,4 0,014 
1,1 0,030 
1,6 0,047 
The % error in the daily rate of ozone production during experimentation (at 130 V) was calculated from 
the ozone mass balances reported. in Section D.4 to be 7.5%. 
5.3 EXPERIMENTAL PROCEDURES 
Hydrogen peroxide, dissolved oxygen, ozone and atrazine concentrations were measured during the 
investigation . Hydrogen peroxide and dissolved oxygen concentrations were measured in the 
characterisation of the ultrasonic cell. Hydrogen peroxide is formed during sonication and was thus used. as 
a qualitative measure of the free radical reactions initiated by ultrasound Dissolved oxygen concentration 
was determined since the presence of oxygen enhances free radical reactions. Hydrogen peroxide and ozone 
concentrations were measured in the characterisation of the ozone system. Atrazine concentration was 
measured in the atrazine degradation experiments. The analytical procedures used to perform each of the 
measurements are detailed in Section S.4. 
An 8 mL sample was required for the hydrogen peroxide determination. Samples were sucked up using a 
pipette through the sample inlet pon in the lid of the ultrasonic cell. Each data point in an experiment was 
thus measured in a separnte experiment since the ultrasonic cell had to be opened to take the measurement, 
thus altering the gas atmosphere in the cell. Separate experiments were also performed for each data point 
in the measurement of dissolved oxygen and ozone concentrations. Dissolved oX)'gen concentration was 
measured by lifting the ultrasonic horn out of the ultrasonic reU and inserting the oxygen electrode into the 
reaction solution immediately after sonication was stopped. Larger sample volumes were required in the 
determination of ozone concentration, volumes of between 15 and 90 mL were used in the indigo method 
and 50 mL in the potassium iodide method Samples were collected through the outlet sample drain at the 
Chapter S ExrERIMENTAL DESIGN 5-18 
oonom of the ultrasonic cell. The ozone gas was diverted through the bypass line to the potaSSium iodide 
traps so the ultrasonic c:d..I c:ou1d be opened for a sample to be taken. 
Alrazine samples for determination were withdrawn oontinuaUy during an experiment since samples of only 
up to 2 mL were required for HPLC analysis (25 j.Ll... sample injection volume). Atrazine samples were 
taken through a syringe with a long needle through the neoprene scpnun in the ultrasonic ceU lid The total 
reaction volume (500 mL) was not significantly affected by the three samples « 10 mL in total) that were 
taken during the experiment. 
Experiments were performed in either duplicate or triplicate to ensure data reproducibility. The standard 
deviation in data was calculated at eacb time period of an experiment though only the maximum value is 
reported where the data is presented in a figure or table. Statistical analysis (mean. 95 % oonfidence limits, 
standard deviation) is reported in Appendix F. 
S.4 ANALYTICAL PROCEDURES 
The analytical procedure used to measure hydrogen peroxide ooncentration is described in Section 5.3. 1, to 
measure ozone ooncentration in Section 5.3.2, dissolved oxygen in Section 5.3.3 and atrazine concentration 
in Section 5.3.4. 
5.4.1 Hydrogen peronde measurement 
A method based on the recWction of copper(ll) ions by hydrogen peroxide in the presence of an excess of 
2.9-dimethyl·I , 1O-phenanthroline (DMP) was used to measure the concentration of hydrogen peroxide in 
solution (Saga et al., 1988). A copper(l)-DMP oomplex, Cu(DMPh ". is formed according to the reaction 
presented in Scheme 5. 1. 
2Cu" + 4DMP + H,O, - 2Cu(DMP)i + 0 , + 2H' 
Scbeme 5.1 
The product is stable in air- and oxygen-saturated solutions and is measured spectrophotometrically at a 
wavelength of 454 om. The method is accurate to measure hydrogen peroxide concentrations in the range 
0,030 to 4,08 mg L·1 (Baga et a1., 1988). 
A sample mixture consisting of I mL of a coppe:r(I1} sulphate solution, I mL of an ethanolic OMP solution 
and 8 mL of the solution of which the hydrogen peroxide: concentration 'was to be detennined was prepared 
in a 10 mL volumetric flask.. Absorbance of the sample mixture was measured at a wavelength of 454 nm 
in a 10 mm glass cuvette with a Pharmacia Biotech Ult.rospec 2000 (model 30-2106-00) spectrophotometer. 
Absorbance was measured for solutions of known hydrogen peroxide concentration and a calibration curve 
Chapter 5 ExPI'.RIMJ:.NTAL DESIGN 5-19 
was drawn by plotting absorbance at 454 om versus bydrogen peroxide concentration. The concentration of 
a hydrogen peroxide solution of unknown concentration could thus be determined. using the calibration 
curve and the measured absorbance value of the solution. 
The experimentaJ procedure used to construct a calibration curve and the calibration curve drawn for each 
batch of chemical reagents used during the investigation are presented in Section B. I of Appendix B. The 
error associated. with determining hydrogen peroxide concentration using the three calibration curves 
reported in Appendix B were 11. 14 and 13 0/0" respectively. 
5.4.2 Ozone measurement 
Aqueous ozone concentration was detennined using an iodometric (Eaton et al., 1995a) and indigo 
colorimetric method (Bader and Hoigni, 1981; Eaton et al., 1995b). The iockmctric method could detect 
concentrations greater than 1 mg L·1 whereas the indigo colorimetric method oould measure ozone 
concentrations down to 2 IJ.& L' \ (Eaton et al. , 1995a; Eaton et al., 1995b). The iodometric method was used 
for the daily characterisation to determine the rate of ozone generation as well as to detennine the amount 
of ozone that had passed through the ultrasonic ce.ll and reacted with the potassium iodide in the gas traps. 
The indigo colorimetric method was used to measure residual ozone in solution in the ultraSonic cell after 
an experiment had been completed. Both methods are reported in literature for ozone determination, the 
iodometric method was used in the investigations by Caprio and co-workers (Caprio et al., 1982), Smith and 
co-workers (Smith et al., 1992) and Singer and Zilli (Singer and ZilIi, 1975) and the indigo method in 
investigations by Roebe and co-workers (Rocbe et al., 1994; Roehe and Prados, 1995), Yolk and co-workers 
(Yolk et al ., 1997) and Hoigne and Bader (Hoigne and Bader, 1983). 
5.4.2. 1 lodometric method 
The iodometric method is based on the liberation of free iodine from potaSSium iodide (K.I) solutions by 
reaction with ozone (Eaton et al., 1995a). The liberated iodine is titrated with a standard solution of sodium 
thiosulfatc (NazS:z03) using starch as an indicator. The titration is performed at a pH of 3 to 4 since the 
reaction is not stoichiometric at a neutral pH due to the partial oxidation of thiosulfate to sulfate. The 
chemical reactions of the iodometric method are shown in Scheme 5.2 (Hargis, 1988). 
0 3 + 21- + H20 -+ 20H- + O2 + 12 
2S 20~- + 12 ... S40~- + 21-
 Scbeme 5.2 
(a( 
(h( 
The iodometric method can be used to determine ozone concentrations above 1 mg L·\. The procedure of 
the potassium iodide reaction and titration with sodium thiosulphate was performed as recorded in Standard 
Methods (Eaton et al ., 1995a). The daily characterisation to determine the rate of ozone generation is 
described in Section 8 .2. 
Chapter 5 EXPERlMI.NTAL DESIGN 5-20 
5.4.2.2 Indigo coIorimetric metbod 
The indigo oolorimetric method is based on the decolourisation of indigo in acidic solution by ozone (Bader 
and Hoigne, 1981 ; Eaton et al. , 1995b). Potassium indigo trisulfonate is used to prepare the indigo stock 
solution. Indigo trisulfonate absorbs light strongly at a wavelength of 600 n.m and has a molar absorptivity 
of E600 = 23 800 Ml cm·1 (Bader and Hoigne, 1981). A potassium indigo trisulfonate molecule oontains 
one carbon-aubon double bond that will react with ozone. Thus, as shown in Scheme 5.3, one mole of 
ozone reacts with one mole of sulfonated indigo. The reaction produces sulfonated isatin and similar 
products that do not absorb light at a wavelength of 600 DID (Bader and Hoigne, 1981). 
00,' 
indigo trisulfonic acid 
SchemeS.3 
so,' 
0, 
• 
so,' 
2 
o 
othe, 
+ products 
isatin sulfonic acid 
The indigo oolorimeLric method can be used to detect ozone ooncentrations down to 2 ~g L -I (Eaton et al., 
1995b). The relative error is less than 5 % without special sampling setupS and in laboratory testing may be 
reduced to I % (Eaton et al. , 1995b). Hydrogen peroxide and organic peroxides decolourize the indigo 
reagent very slowly; no interference is recorded for hydrogen peroxide if the abiorbance at 600 n.m is 
measured within 6 h of adding the reagents. The indigo method cannot be used to determine ozone 
ooncentration in the presence of bromine, the interference of chlorine can be masked by malonic acid 
(Eaton et al ., 1995b). The procedure of the indigo oolorimelric method (Section 6 .2) was performed as 
recorded in Standard Methods (Eaton et al., 1995b). 
5.4.3 Dissolved oxygen measurement 
The dissolved oxygen ooncentration of water samples was measured using a WTW 90 oxygen electrode wjth 
a wrw OXl 91 oontroller. Dissolved oxygen ooncenttation was measured immediately after sonication and 
care was taken so as to not trap oxygen bubbles below the electrode. The oontroller was adjusted daily for 
changes in atmospheric pressure. Atmospheric pressure was read from a Negretti & ZambIa barometer 
(model number MM 4 188/43). The oontroUer ... :as calibrated daily by submersing the oxygen electrode in 
oxygen-saturated water at 25 QC and setting the standard measurement to 40,32 mg L-\ the solubility 
concentration of dissolved oxygen in water at 25 QC (International Critical Tables, 1928). The solubility 
ooncentration of oxygen in water at temperatures between 0 and 40 QC is recorded. in Section B.3. 
Cbapter S ExrI'.RlMl'.NTAL DESIGN 5-21 
5.4.4 Atrazine measurement 
Atrazine concentration during sonication and ozonation was measured using high performance liquid 
chromatography (HPLC). The HPLC system consisted of a Waters TM 600 controUer, a Waters 486 
tunable absorbance detector, a Waters U6K injector system. a Demark Technology coDlfJUter system and a 
Seikosha SL-75 printer. The solvent was a 70:30 (v/v) mixture of methanol to water and. was pumped at a 
flow rate of 0,6 mL min- I through a reverse phase column, LiChrospber* RP-setect B (250 x 4 mm., 5 j.I.Dl 
panicles). The sample injection volume was 25 jJl.. Absorbance was measured at a wavelength of 230 nm. 
A calibration curve relating alrazine concentration to absorbance measured with the HPLC al a wavelength 
of 230 nm is presented in Section B.4. The error in the alrazine concentrations calcu1atcd using the 
atrazine calibration curve is 7 %. HPLC methods, as reported in literature, have been used 10 detect 
at.raz.ine concentrations from 0,4 10 over 50 rng t ol (Brambilla et al ., 1995; Chan et al., 1992; Petrier et al., 
1996; Stucki et al., 1995). Solvents are usuaUy 5O:SO 10 80:20 (vlv) mixtures of methanol and water 
(Hustert et al., 1991 ; PCtrier et aL, 1996) or 30:70 10 55:45 (vlv) mixtures of acetonitrile and water 
(Beltr.in et al., 19930; Beltr.in et al., 1998; Mirgain et al., 1993). Columns = typically !eVe"" phase CI8 
columns (Acl.ntegui et al., 1995; Cban et al. , 1992; Grigg et al. , 1997) and absorbance is measured al a 
wavelength between 214 and 235 om (Arantegui et al., 1995; Beltn\n et al., 1998; Mirgain et al ., 1993). 
An atrazine stock solution was prepared in alL volumetric flask by dissolving 100 mg of atrazine (97 %; 
Sanachem) in 60 mL methanol and diluting with ultrapure water to tota1 volume. Methanol was used to 
dissolve the alJ3Zine since atrazine solubility in water is 33 mg Lol at 22 °C (Tomlin, 1997). The ultrasonic 
cell was rinsed between experiments with both acetone and water due to the low solubility of atrazine in 
water. Waste atrazine solution was collected in buckets and as recommended in the Material Safety Data 
Sheet (MSDS) was reacted with a IO 000 mg L-1 sodium hydroxide solution before dilution and disposal to 
droin (Bradfield. 1997). 
The methods used to measure ozone and hydrogen peroxide concentration were tested using controls for 
interference due to atrazmc. Atrazine was shown not to interfere with the methods. 
5,5 STATISTICAL MODELLING 
Statistical analysis of the data was performed using the software programme SfATISfICA that was 
Qeveloped bl' Statsof\, lnc (Statsof\, Inc., 2(00). STASTlCA is 0 oomprcbensiv" statistical data analysis, 
graphics, data base management and custom application development system. SI'ATISfICA consists of 
modules in which the different types of analyses are perfonned (Statsoft, inc., 2000). Statistical analysis 
used in the investigation included descriptive analysis to calculate means and standard deviations, 
regression analysis, analysis of variance, non·1inear estimation and experimental design to generate surface 
resp:mse diagrams. 
6 
ULTRASOUND PROCESS INVESTIGATION 
The implementation of a new industrial process, as shown in Figure 1.1, progresses from the discovery or 
demonstration of a process to a laboratory reactor SO that reaction rate and yield data can be measured and a 
mathematical reactor model be formulated, to a pilot plant reactor for the evaluation of the reactor model 
and investigation of process control and hence to the design and construction of a large~sca1e reactor. A 
sound fundamental knowledge of the process should be obtained during the investigation in the ultrasonic 
laboratory reactor. An ultrasonic latx:lratory reactor (the ultrasonic cell) was designed (Chapter 5) for 
laboratory experimentation to demonstrate and evaluate the potential application of ultrasound to degrade 
organic pollutants during water treatment and to investigate the implications for scale-up. 
This chapter details the process investigation to characterise the ultrasonic cell and to investigate the 
mechanistic sub-processes occurring during sonication. Experiments investigating dissolved oxygen 
concentration are reponed in Section 6.1, hydrogen peroxide fonnation in Section 6.2. bydrogen peroxide 
degradation in Section 6.3, interval experiments in Section 6.4 and commercial hydrogen peroxide 
experiments in Section 6.5. A summary of the experimental programme of the ultrasonic pnx::ess 
investigation is presented in Table 6. 1. 
Table 6.1 : Experimental programme or tbe ultrasonic process innstigatioD 
Stttion Title Measured Variables Data Statistics 
parameters Appendix Appendix 
6.1 Dissolved oxygen [Ozl Oz concentration C. I F.3.1 
concentration gas (Nz; air; Oz) 
6.2 Hydrogen peroxide [11,0z1 Oz concentration C.2 F.3.2 
fonnation acoustic power 
6.3 Hydrogen peroxide [H,OzI gas (Nz: air; 0.2) c.J F.3.3 
degradation 
6.4 Interval experiments [H,OzI sonication intervals C.4 F.3.4 
gas intervals 
6.5 Commercial hydrogen [H,OzI H20z concentration C.5 F.3.5 
peroxide e"-periments gas (Nz: air; 0.2) 
Chapter 6 ULTRASOUND PROCESS INvEsTIGATION 6-2 
The experimental data discussed in this chapter is reported (as indicated in Table 6.1) in Appendix C. The 
statistical analysis (mean. 95 % confidence limits, standard. deviation) is reported in Appendix F. Standard 
deviation was calculated for each time period of an experiment though only the maximum value is reported 
in this chapter. Each data point in experiments measuring hydrogen peroxide or dissolved oxygen 
concentration was measured in a separate experiment since the ultrasonic cell had to be opened during 
sampling (described in Section 5.3), thus altering the gas atmosphere in the cell. 
6.1 DISSOLVED OXYGEN CONCENTRATION 
The chemical reactions that take place during sonication are reported in Section 2.3. The reaction schemes 
are summarised in Appendix H on A3 paper that can be folded out. Free radicals react to produce hydrogen 
peroxide and to oxidise organic compounds in water. The reactions that take place during sonication are 
affected by the presence of a gas since certain reactions may be inhibited or other reactions initiated 
depending on the nature of the gas. The presence of carbon dioxide or hydrogen, explained in 
Section 2.3.1.2, inhibits the sonochemical formation of hydrogen peroxide whereas the presence of oxygen 
enhances oxidation taking place during sonication since perhydroxyl radicals (OHOz) are produced from the 
additional reactions shown in Scheme 2.3 and Scheme 2.4 (see Appendix H). Cavitation intensity is also 
enhanced, explained in Section 2.2.1 , when a gas is present during sonication since additional nucleation 
sites are available for cavity formation. 
A gas-saturated liquid, however, is degassed during sonication. Schwikkard demonstrated that the o":ygen 
concentration of an o":ygen-saturated solution was reduced to that of an unsaturated solution \\ithin 3 h of 
sonication in a ultrasonic bath (Schwikkard, 1995). Dissolved oxygen concentration is an important 
parameter to quantifY and control in an ultrasonic system in order to obtain reproducible results. 
The dissolved oxygen concentration in oxygen·saturated water was measured during sonication over 3 h. 
The water was saturated with oxygen by sparging medical grade oxygen through the water at a flow rate of 
6 m.L S·l for 10 min prior to sonication: o>..)'gen sparging was stopped as the ultrasonic horn was switched 
on. O~.:ygen concentration was measured at time periods of 0; 5: 10: 20; 30; 45; 60: 90: 120: 150 and 
180 min. Triplicate experiments were perfonned for each time period. The dissolved ox)'gen concentration 
in water with no sonication, once oxygen sparging was stopped, was measured as a control . The reduction 
in dissolved oxygen concentration during nitrogen sparging of oxygen-saturated water (without sonication) 
is also shown in Figure 6.1: nitrogen sparging was started as the oxygen sparging was stopped. 
Chapter 6 ULTRASOUND PRocESS boVESTJGAll0N 
OO f 
i 40 r -e--~ -e--- - --e- - -e-- - -e- - -e-- -
 • 
• 
• • • 
• • • • • • 
• 
• • • 
- ~- - -.()- - -.()- - - -<)- - - -- - - - ~- - - - - - - -~- - - - - - - - &-
 Time (min) 
• control 
• sonication 
lit nitrogen 
~ (P = 20 kPa) 
-e (P = 100 kPa) 
6-3 
Figure 6.1: Reduction in dissolved oxygen concentration in oxygen-saturated water during 
sonication, without sonication (control) and during nitrogen sparging without sonication; the 
dissolved oxygen concentrations in water saturated with oxygen at partial pressures of 20 and 
]00 kPa are also shown (n = 3 per time period) 
Sonication, shown in Figure 6_1, reduces the oxygen concentration of oxygen-saturated water faster than the 
reduction in concentration in the absence of ultrasound, demonstrated by the control. The maximum 
standard deviation in the oxygen concentrations measured for the control and during sonication are 1,65 
and 1,85 mg L -I, respectively. Oxygen concentration decreased during sonication from that of saturation at 
a panial pressure of 100 kPa (approximately 40 mg L-1 at 25 °C, the long dashed line in Figure 6.1 denoting 
water saturated with pure oxygen) to that of water saturated with oX'ygen at a partial pressure of 20 kPa (the 
shon dashed line in Figure 6.1 denoting water saturated with air). The oxygen concentration of water 
saturated with air was measured in all samples before experimentation, oxygen concentration varied 
between 8.3 and 12,3 mg L-1, the average was 10,0 mg L·1 (Table C.2 in Appendix C). 
The rapid decrease in oxygen concentration in water saturated with pure oxygen during nitrogen sparging, 
shown in Figure 6.1 , is a convenient method to de-oxygenate a solution so as to investigate chemical 
reactions occurring in the absence of oxygen. Oxygen concentration decreased within 5 min from 40 mg L·1 
to below 3 mg L·1 and remained below 2 mg L-1 from 15 min onwards. The maximum standard deviation 
in oxygen concentrations measured during nitrogen sparging is 2,15 mg L-1, the standard deviation in data 
at each time period is recorded in Table F.II in Appendix F. It is thus possible to operate the ultrasonic cell 
as an anaerobic environment by sparging nitrogen through the system. 
The degassing characteristic of ultrasound resulted in the oxygen concentration of oxygen-saturated water to 
decrease during sonication, as shown in Figure 6.1 , however, the dissolved oxygen concentration must be 
Chapter 6 ULTRASOUND PROCESS INvFsnGA1l0N 
kept constant during sonication if the effect of oxygen on radical reactions is to be investigated A constant 
oxygen concentration, shown in Figure 6.2, was obtained by continuing the oxygen sparging during 
sonication. Pure oxygen was sparged through water at a flow rate of6 mL S-I for 10 miD prior to and during 
sonication. Oxygen concentration was measured at time periods of 0; 5; 10; 20 and 40 min. Duplicate 
experiments were performed for each time period Oxygen concentration was also measured over 40 min in 
water saturated with air and nitrogen; saturation was achieved by sparging air or nitrogen through the water 
for 10 min prior to and during sonication at a flow rate of 6 mL S-I . A control was perfonned by measuring 
dissolved oxygen concentration during sonication in water that bad been saturated with air and subsequently 
had no gas sparging through it. 
50 
~"" • • 1 1 §. 
f c 0 3) • control ii r 
" 
.. nitrogen 
• l • air ~ 0 20 
• oxygen 0 l c • • a • , • ?< 10 • 0 I 
r 
0 
0 10 20 3) 
"" 
50 
Time (min) 
Figure 6.2 : Dissolved oxygen concentration in nitrogen-, air- and oxygen-saturated water 
during sonication in tbe ultrasonic cell (n = 2 per time period) 
The dissolved OX1'gen concentmtion in water during sparging with nitrogen, air or oxygen, as shown in 
Figure 6.2, remained constant during the 40 min experiment. Dissolved oxygen concentration in the 
control and during air and ox)'gen sparging was approximately 10,7; 12,5 and 39,6 mg L-t. respectively. 
The saturation concentration of pure oxygen (at 100 kPa paniai pressure) in water is approximately 
40 mg L· 1• Oxygen concentration during sonication with nitrogen sp<uging was below 1.2 mg L-1 
throughout the ex-periment. The maximum standard deviation in oxygen concentrations measured for tbe 
control and during nitrogen, air and oxygen saturation are 0,28; 0,21 ; 0,71 and 0.96 mg L- I • respectively. 
There is not a significant difference between air sparging and the conU'ol with out air sparging. 
The ultrasonic cell is thus able to be used to investigate the effect of oxygen concentration on reactions 
occurring during sonication. Oxygen concentration is kept constant by continuing the gas sparging through 
Chapter 6 ULTRASOUND PROCESS b."VES11GATION 6-5 
the water during sonication. Different oxygen concentrations can be maintained during sonication by 
sparging gases such as nitrogen, air and oxygen or no gas through the solution in the ultrasonic cell . 
6_2 HVDROGEN PEROXIDE FORMATION 
Free radicals are highly reactive chemical species and have a very short lifespan. Radica1 concentrations 
can be measured using spin trapping compounds and electron spin resonance techniques. The chemistry of 
free radicals is summarised in Appendix G. The radical reactions occurring during sonication, as shown in 
Scheme 2.2 in Section 2.3 .1.1 and Scheme 2.3 in Section 2.3. 1.2, lead to the production of hydrogen 
peroxide (summarised in Appendix H). Various routine methods are available to measure hydrogen 
peroxide concentration in water. Thus the detection of hydrogen peroxide can be used as a qualitative 
measure of the radical reactions occurring during sonication. A higher hydrogen peroxide concentration 
indicating process conditioDS that enhance the occurrence of the radical reactions. The effect of ultrasound 
on a system is indirect, as discussed in Section 2.3. 1.1, in that it is due to the formation of highly reactive 
radical species and not due to direct attack on solutes present in solution. The measurement of hydrogen 
peroxide was used in this investigation as an indirect surrogate measure of the degree of radica1 reactions 
occurring during sonication and as a tool to judge under which process conditions the most radical reactions 
occurred in the ultrasonic cell. 
6_2_1 Effect of dissolved gas 
The fonnation of hydrogen peroxide in water in the ultrasonic cell Vt'3S measured using the DMP method 
detailed in Section 5.4 . 1. The effect of dissolved oxygen concentration on hydrogen peroxide formation 
(and hence on the occurrence of radical reactions) during sonication was investigated by measuring 
hydrogen peroxide concentration in nitrogen •. air· and oxygcn·saturated water. The gases were sparged 
through the water at a flow rate of6 mL S·I for 10 min prior to and during sonication to maintain sanuation. 
The rate of formation of hydrogen peroxide under saturation with each gas was calculated from the 
measurement of hydrogen peroxide concentration at time periods of 0: 4: 8; 12; 16~ 20; 30; 60: 120; 240: 
480 and 960 min (1 6 h). Triplicate experiments were performed for each time period for each gas. A 
control was performed by measuring hydrogen peroxide concentration during sonication of water with no 
gas sparging. The formation of hydrogen peroxide in nitrogen·, air· and oxygen·saturated water is 
presented in Figure 6.3. 
Chapter 6 ULTRASOUND PRocESS iNvEsnGATION 
0,12 
'ii 
.s 0,00 f 
~ 
~ 0 ,06 
• ~ 
N 
~ 0,03 
::t: 
, ,5 
:::; 
.. 
E-
 o 
0 
~ 
• b 
0 
• g 
0 
0 0,5 
N 
0 
N 
::t: 
0 
o 
0 
• 
• I 
4 B 
4 B 
'2 
Time (min) 
(a) 20 miD 
'2 
Time (h) 
(b) 16 h 
,. 
• 
, 
* 
• 
20 
'6 
24 
20 
• control 
... nitrogen 
:Ill air 
... oxygen 
• control I 
't' n~trogen l 
,.. air 
... oxygen 
Figure 6,3: Hydrogen peroxide formation in nitrogen-. air- and oxygen-saturated water during 
sonication in tbe ultrasonic cell at an acoustic power of 57 W (n = 3 per time period) 
The formation of hydrogen peroxide over 20 min is shown in Figure 6.3(a) and over 16 h in Figure 6.3(b). 
The type of gas present during sonication is shown to influence the concentration of hydrogen peroxide in 
solution, Figure 6.3 (a). Hydrogen peroxide concentration was the greatest during ox-ygen saturation, the 
least during nitrogen saturation and similar in the control and during air saturation. The maximum 
standard deviation in the hydrogen peroxide concentrations measured. over ° to 20 min for the control and 
during nitrogen, air and oxygen saturation are 0,010: 0.004; 0,006 and 0,008 mg L'\ respectively (the 
maximum standard deviation for the experiment is quoted, though, the standard deviations calculated for 
each of the time periods are listed in Appendix F). The initiaJ offset in Figure 6.3(a) in the hydrogen 
peroxide concentration during ox-ygen sparging after the 10 m.in saturation period is unexplained but was 
Chapter 6 ULTRASOUND PRocrss iNvEsnCATION 6-7 
present during the 3 replicate experiments. The ver)' low concentrations, however, lie at the minimum 
detection limit of the DMP method and the offset could be due to experimental error. Hydrogen peroxide 
concentration (and hence the occurrence of radical reactions) during sonication is directly detennined by the 
dissolved oxygen concentration which, as reported in Section 6.1, is greatest for oxygen saturation, the least 
during nitrogen saturation and similar in the control and air saturation. 
The rate ofbydrogen peroxide formation for the different gases was calculated over the initial 20 min period 
from the regression of the data using the linear regression model 
y=a+bx (6.1( 
where the coefficient b, the gradient of the regression line, represents the rate of hydrogen peroxide 
formation. The calculated rates of hydrogen peroxide formation in nitrogen-, air- and oxygen-saturated 
water during sonication (and the standard error of the calculated values) are recorded in Table 6.2. 
Table 6.2: Rate of hydrogen peroxide formation in nitrogen-, air- and oxygen-saturated water 
during sonication in the ultrasonic cell for 20 min at an acoustic power of 57 W 
Gas 
control 
nitrogen 
air 
o~\:ygen 
Rate of B2<h formation 
(mg L- I min-I ) 
0,0043 ± 0,000 I 
0,0028* 0,0001 
0,0044 ± 0,0001 
0,0050 ± 0,0003 
0,968 
0,962 
0,971 
0,928 
The statistical F-test at a 95 % confidence level indicated that there was a significant difference in the 
gradients of the regression model for the different gases. Multiple comparison using the Bonferroni method 
(Coetzer, 2(00) indicated that the significant difference was between the gradient of the nitrogen regression 
line and the other regression lines, there was no significant difference between the gradients of the control. 
air and oxygen regression lines. Thus, although the increased dissolved oxygen concentration during 
ox)'gen saturation increased hydrogen peroxide concentration in solution_ as shown in Figure 6.3 (a). the rate 
of formation was not significantly greater than that of the control and during air saturation. The significant 
difference between nitrogen saturation and air or oxygen saturation indicates the importance of the presence 
of oxygen to enhance radical reactions and thus the peJfonnance of an ultrasonic process. U1trasonic 
systems should not be run under anaerobic conditions if the primary goal is chemical oxidation of solutes. 
The difference in hydrogen peroxide formation over longer time periods (16 h), as shown in Figure 6.3(b), 
between the control and oxygen and air saturation was not as significant as during the shorter periods of 
sonication although hydrogen peroxide concentrations were still greater during oxygen sparging. The data 
Chapter 6 ULTRASOUND PROCESS lNvFsnGAll0N 
is more scattered after the longer sonication periods and has a greater standard deviation than that measured 
at shorter periods of sonication. The regression lines in Figure 6.3(b) are plotted from the correlation 
(equation 6.3) of bydrogen peroxide as a function of time, acoustic power and oxygen concentration. The 
maximum standard deviation in the hydrogen peroxide concentrations measured over 0 to 16 h for the 
control and during nitrogen. air and oxygen saturation are 0,060; 0,012; 0,097 and 0,108 mg L-1, 
respectively. The lack of oxygen present during sonication with nitrogen saturation, shown in 
Figure 6.3(b), caused the hydrogen peroxide concentration after 20 min of sonication to remain relatively 
constant and to fluctuate between 0,06 and 0,09 mg L·1• Whereas hydrogen peroxide concentration under 
nitrogen saturation reached steady state after 20 miD, the hydrogen peroxide concentrations during air and 
oxygen saturation only started to level off and approacb steady state conditions after 16 h. 
6.2.2 Acoustic power 
Hydrogen peroxide fonnation is not only dependent on the presence of oxygen during sonication but also on 
acoustic power. A lower acoustic power resu1ts in fewer water molecules being split into hydro~1'1 and 
hydrogen radicals (as shown in Scheme 2.1 in Section 2.3.1.1) and thus fewer radical reactions taking place 
during sonication. Hydrogen peroxide was again used as an indicator of the radical reactions occurring 
during sonication at a lower acoustic JXlwer. A lower acoustic JXlwer in the ultrasonic cell was achieved by 
sonicating at a transducer displacement amplitude of 5 ~ whereas the previous experiments investigating 
the fonnation of hydrogen peroxide under saturation with different gases during sonication had been 
performed at a transducer displacement amplitude of 11~. Transducer displacement amplitudes of 5 and 
11 ~. as recorded in Table 5.2 in Section 5.1.2.2, correspond to acoustic JXlwers of 24 and 57 W, 
respectively, and ultrasonic intensities of 0,048 and 0, 114 W cm·3• respectively. 
The effect of dissolved oxygen concentration was also investigated during sonication at a lower acoustic 
power. Hydrogen peroxide formation during the sonication of water saturated with nitrogen (an 
oxygen-free atmosphere) was compared with that in a control (water without any gas sparging through it). 
The oA1'gen concentration in a comrol, as reponed. in Section 6.1 , is 10,7 mg L·1. Water saturated with 
nitrogen was prepared by sparging nitrogen through the water for 10 min prior to and during sonication at a 
flow rate of 6 mL S·l. A control was performed by measuring hydrogen peroxide concentration in water 
with no gas sparging. Hydrogen peroxide was measured at time periods of 0; 4; 8; 12; 16; 20: 30; 60; 120: 
240: 480 and 960 min (16 h). Triplicate experiments were performed for each time peri<Xl. A comparison 
of the hydrogen peroxide fonnation during sonication at acoustic powers of 24 and 57 W for the control is 
presented. in Figure 6.4 and for the nitrogen-saturated water in Figure 6.5. 
Chapter 6 ULTRASOUND PROC&SS iNvEsnCA110N 6-9 
0,12 
:J ~ • ~ 0,00 0 • • 0 ~ 
r • • \ 0 24W I ~ 0,06 0 
.57 W • g 
0 
• • 0 0 
• 11 0,03 N 
:z: 
0 \' 
0 • 8 12 I. 20 2. 
Time (min) 
(a) 20 min 
1,. 
:J 
~ 1,2 
0 
0 
ii 
• I 0 24W I 
" 
0,8 0 
·57W • g 
0 
0 
N 
0 0,4 N 
:z: 
0 
0 • 8 12 I. 
Time (h) 
(b) 16 b 
Figure 6.4: Hydrogen peroxide formation in the control during sonication in tbe ultrasonic cell 
at acoustic powers of 24 and 57 W (n = 3 per time period) 
The rate of hydrogen peroxide formation in the control, as shown in Figure 6.4(a), was lower when 
sonication was performed at an acoustic power of 24 W than at 57 W. The maximum standard deviation in 
the hydrogen peroxide concentrations measured over 20 min at acoustic powers of 24 and 57 W for the 
control are 0,006 and 0.010 mg L't. respectively. A lower acoustic power results in fewer water molecules 
being degraded into hydro:\.")'l and hydrogen radicals and thus fewer radical reactions occurring in solution 
during sonication. 
The difference in hydrogen peroxide formation in the control due to acoustic po .... 'Cr input. as shown in 
Figure 6.4(b), became insignificant after I h. The regression lines in Figure 6.4(b) are ploned from the 
Cbapter 6 ULTRASOUND PROCESS INvIsnGATION 6-10 
correlation (equation 6.3) of hydrogen peroxide as a function of time, acoustic power and oxygen 
concentration. The maximum standard deviation in the hydrogen peroxide concentrations measured during 
sonication at acoustic powers of 24 and 57 W for the control over 16 h are 0,026 and 0,060 mg L-I, 
respectively. 
The hydrogen peroxide formation in nitrogen-saturated water during sonication at acoustic powers of 
24 and 57 W is presented in Figure 6.5. 
0 ,12 
0.1S 
"'012 -
 .s ' 
~ 
~ o ,CS -
 ~ 
N 
~ 0,04 
I 
" J i 
K' 
- . 
" 
° ~--------------------------------~ 
° 
2 3 4 5 
Time (h) 
(b) 4 h 
fV24Wl 
~ 
-:- 24W 
" 57 W 
Figure 6.5 : H)'drogen peroxide formatioo in nitrogeo-saturated 1\'ater during sonication in the 
ultrasonic cell at acoustic powers of 24 and 57 W (n = 3 per time period) 
Chapter 6 ULTRASOUND PROCESS bo'VESTIGATION 6-11 
Hydrogen peroxide formation in a nitrogen-saturated solution, as shown in Figure 6.5(a), was the same 
irrespective of the acoustic power input. The maximum standard deviation in the hydrogen peroxide 
concentrations measured at acoustic powers of 24 and 57 W during nitrogen saturation are 0,004 and 
0,004 mg L'., respectively. Whereas in Figure 6.5(a) acoustic power was shown not to affect hydrogen 
peroxide formation in nitrogen-saturated water during 20 min of sonication, Figure 6.5(b) indicates that 
acoustic power did not influence hydrogen peroxide formation for up to 4 h of sonication. The regression 
lines in Figure 6.5(b) are plotted from the correlation (equation 6.3) of hydrogen peroxide as a function of 
time, acoustic power and oxygen concentration. The maximum standard deviation in the hydrogen 
peroxide concentrations measured during nitrogen saturation over 16 h at acoustic powers of 24 and 57 W 
are 0,004 and 0,007 mg L· 1, respectively. 
The rates of hydrogen peroxide formation for the control and nitrogen-saturated water were calculated from 
the regression of the data presented in Figure 6.4(a) and Figure 6.5(a) using the linear regression model 
y:: br (6.2( 
where the coefficient b, the gradient of the regression line, represents the rate of hydrogen peroxide 
formation. The calculated rates of hydrogen peroxide formation at acoustic powers of 24 and 57 W for the 
control and during nitrogen saturation (and the standard error of the calculated values) are recorded in 
Table 6.3. 
Table 6.3: Rate of hydrogen peroxide formation in nitrogen-saturated water and the control 
during sonication in the ultrasonic cell at acoustic powers of 24 and 57 W for 20 min 
Gas 
control 
control 
nitrogen 
nitrogen 
Acoustic power 
(W) 
24 
57 
24 
57 
Rate of H:O: formation 
(mg L-l min-I ) 
0,0025 ± 0,000 1 
0.0043 ± 0,0001 
0.0028 ± 0.0001 
0.0028 ± 0,0001 
0,%2 
0,%8 
0.968 
0,962 
The statistical F~test at a 95 % confidence level indicated that there was a significant difference in the 
gradients of the regression model for sonication at acoustic powers of 24 and 57 W for the control but no 
significant difference in the gradients of the 24 and 57 W regression model for nitrogen saturation. 
Hydrogen peroxide fonnation was shown to be affected by acoustic power in the presence of ox)'gen (as 
during the control) but not in the absence thereof (as during nitrogen saturation). Thus, under these 
experimental conditions, the presence or absence of oxygen had a greater effect on the rate of hydrogen 
Chapter 6 ULTRASOUND PROCESS INvrsnGA110N 6-12 
peroxide formation than the change in acoustic power from 24 to 57 W. The formation of hydrogen 
peroxide during sonication at both acoustic powers is an indication that the acoustic powers are greater than 
the threshold power value. as discussed in Section 2.2.1 of Chapter 2, required to initiate cavitation. 
Oxygen sparging can thus be replaced with air sparging in ultrasonic systems of similar design and power 
that require the presence of oxygen and have a long sonication period without sacrificing performance as 
demonstrated in Figure 6.3(b). A lower acoustic power energy source can also be used under both 
oxygenated and de-oxygenated process conditions in ultrasonic systems with an extended sonication period 
as demonstrated in Figure 6.4(b) and Figure 6.5(b). Maximum performance in ultrasonic systems with a 
short sonication period is obtained with oxygen sparging and a high acoustic power input as demonstrated 
in Figure 6.3(a) and Figure 6.4(a). The most appropriate mode of operation should be investigated for 
potential ultrasonic applications, however. as a guide, short experiments should be pe.rfonned with oxygen 
sparging and a high acoustic input and extended sonication experiments with air sparging and a lower 
acoustic energy source. 
6.2.3 Regression analysis 
The measurement of hydrogen peroxide concentration has been used as a tool to investigate process 
conditions (dissolved oxygen concentration and acoustic po'\\o'Cr) that enhance radical reactions occurring 
during sonication. A statistical software package STAT/SF/CA, detailed in Section 5.5 of Chapter 5, was 
used to model hydrogen peroxide concentration. the dependent variable, as a function of dissolved oxygen 
concentration and acoustic po\\'Cr, the independent variables. Surface response diagrams, gradient profiles 
of hydrogen peroxide concentration as a function of the two independent variables. were generated fOT time 
periods of 20, 120 and 960 min. The gradient profiles presented in Figure 6.6 indicate the conditions of 
oxygen concentration and acoustic power that lead to the highest (dark grey) and lowest (light grey) 
hydrogen peroxide concentrations. 1be circular symbols indicate the experimental data points. 
~ 
i 
~ 
= 0 • 
=0.02 ~ 
lE!] 0.04 0 
_0.06 
_0.06 
_0.1 
-.-
 (a) 20 min 
Chapter 6 ULTRASOUND PRocESS iNvEsnGAllON 
i 
1 
c:::J 0 .1 50 
CJ 0.1. ~ 
I!!I 0.22 cS 
_0.28 
_0.34 
.0.4 
-..... 25 ,. 35 .. 
Acoustic power (W) 
(b) 120 min 
35 40 45 
Acoustic power (W) 
(c) 960 min 
6-13 
55 
50 55 .. 
Figure 6.6: Surface response diagrams of bydrogen peroxide concentration as a function of 
acoustic power and dissolved oxygen concentration during sonication of water in the ultrasonic 
cell 
Surface response diagrams, as in Figure 6.6, indicate under which conditions a maximum response of the 
dependent variable is obtained or under what conditions future experiments should be performed to improve 
the response. Hydrogen p::roxide concentration was not measured at a low acoustic power and high oxygen 
concentration (top left co-ordinate in the diagrams presented in Figure 6.6), measurement of hydrogen 
peroxide concentration under these oonditions would improve the accuracy of the predicted gradient profiles 
as the 4 corners of the experimental programme would be mapped oul The trends of hydrogen peroxide 
concentration as a function of acoustic powcr and oxygen concentration are still illustrated. Hydrogen 
peroxide concentration during short sonication experiments is shown in Figure 6.6(a) to be a maximum 
with oxygen sparging and a high acoustic power input and during long sonication experiments, as in 
Chapter 6 ULTRASOUND PRocESS iNvrsnGA110N 6-14 
Figure 6.6(c), is independent of acoustic p>wer inpIt and that air sparging can be used in place of oxygen 
sparging. 
Oxygen concentration, shown in Figure 6.6, was varied from a de-oxygenated concentration to that of water 
saturated. with pure oxygen, si.milarly, an acoustic power of 57 W is the maximum output of the ultrasonic 
born in the ultrasonic cell. Oxygen concentration and acoustic power can thus not be increased in the 
present equipment to increase bydrogen peroxide concentration during short sonication experiments as 
recommended in Figure 6.6(a). The treock in hydrogen peroxide concentration (and hence radical reactions 
occurring during sonication), however, have been demonstrated and can be used to guide the choice of 
process conditions in larger·sca1e ultrasonic systems that deliver a wider range of acoustic powers. 
The change in the surface response profiles of hydrogen peroxide at 20, 120 and 960 miD,. as sbo\\lD in 
Figure 6 .6, indicates that the dependence of hydrogen peroxide concentration on the independent variables, 
oxygen concentration and acoustic power, change with time. There is thus a significant interaction between 
the variables time and oxygen concentration, and time and acoustic power. The independent variable time 
should thus be included in a regression model of hydrogen peroxide formation in water during sonication in 
the ultrasonic cell. Possible non·linear relationships between the dependent and independent variables 
sbou1d also be taken into account, a non·linear response of hydrogen peroxide as a function of dissolved 
oxygen concentration is shown in Figure 6.6 by the curved gradient lines in the surface response profiles of 
hydrogen peroxide. 
A correlation was generated for the prediction of hydrogen peroxide concentration as a function of oxygen 
concentration, acoustic power and time using tbe non·linear estimation modu1e of the software 
SfATISfICA . The oorrclation equation is 
[H202] = -{), 5420[rime ] o,~ + 0, 5516[time ]o,m4[uIS ]--tI,OCII UI [0 2]0.00491 1 (6.3 ( 
where IH..lhJ is measured in mg L'!, (u/s] in W, I~J in mg L' \ and (time) in min. The coefficient of 
variance (R2) of the correlation is 0,973 . This indicates that approximately 97 % of the variance in the data 
is explained by the model . The correlation has been used to plot the solid regression lines in Figures 6.3(b), 
Figure 6 .4(b) and Figure 6 .5(b). The fit with experimental data is good for water saturated with oxygen and 
air and with no gas bubbling (control) but not that good for water saturated with nitrogen. This correlation 
can be used in the design of ultrasonic equipmen~ openlling at the same frequenc)' of 20 kIIz, to predict 
changes in hydrogen peroxide concentration so as to guide the selection of appropriate process conditions. 
Batch operation mode is recommended for processes with a relatively long residence time (Shah et al., 
1999d), this correlation by predicting the hydrogen peroxide concentration over time can indicate whether a 
long or sbon residence time is of greater benefit and thus whether a process should be operated in a batch or 
in a continuous mode. 
Chapter 6 ULTRASOUND PRocESS lNvEsnGATION &-15 
6.3 HYDROGEN PEROXIDE DEGRADATION 
Hydrogen peroxide, as demonstrated in Section 6.2, is produced by radical reactions occurring during 
sonication. The hydrogen peroxide that is formed can react further with the radical species in solution, can 
dissociate 10 form ions or due 10 bond cleavage break down to hydroxyl and perbydroxyl radicals. The 
concentration of hydrogen peroxide measured in solution is thus the combined effect of the formation and 
degradation of hydrogen peroxide. 
The overall rate of hydrogen peroxide formation, as shown in Figwe 6.3 , was greatest during the initia1 
period of sonication and decreased over time until hydrogen peroxide cona:ntration levelled off and 
approached a oonstant value. The initial rate of degradation is significantly lower than the roue of formation 
as the amount of hydrogen peroxide in solution is still small. the rate of degradation increases as bydrogen 
peroxide concentration builds up. The overall concentration of hydrogen peroxide in solution begins to 
reach an equilibrium as the rate of degradation approaches that of the rate of formation. EquihDrium is 
established when the rates of formation and degradation are equal and the overall measured hydrogen 
peroxide concentration in solution remains constant. The equilibrium in hydrogen peroxide formation and 
degradation, as reported in Section 6.2, was establisbed within 30 min during sonication in a de-oxygenated 
atmosphere and only after 16 h in the presence of oxygen. 
The rate of hydrogen peroxide degradation was determined by sonicating water for 16 h at an acoustic 
power of 57 W (to produce hydrogen peroxide) and measuring peroxide concentration over 3 h after 
sonication was stopped.. Hydrogen peroxide concentration was measured at time periods of 0; 5; 10; 20; 30; 
60; 90; 120; 150 and 180 min. Nitrogen, air and oxygen were sparged through the water during the 3 h 
degradation period to investigate their effect up:m the rate of hydrogen peroxide degradation. A control was 
performed by measuring hydrogen peroxide degradation in water with DO gas sparging. Triplicate 
experiments were performed for each time period Di1Jerent hydrogen peroxide concentrations were 
produced during the 16 h period of sonication thus resulting in different initia1 concentrations for the 
degradation experiments. 1be average initia1 concentrations of the control and during nitrogen, air and 
o),:ygen saturation were 0,43; 0,99; 0,95 and 0,37 mg L-1, respectively. The variation in initial 
concentration was removed. for comparison puT}X)SCS, by normalising the data and dividing each 
concentration by the initia1 concentration for that particular e""periment. 1be degradation of bydrogen 
peroxide in nitrogen., air· and oxygen.-.saturated water after sonication in the ultrasonic ceU for 16 h is 
shown in Figure 6.7. 
Cbapter 6 ULTRASOUND PRocESS lNvEsnGAnoN 
1,1 
Cl 
N 
:;. 09 ;a , 
,. 
-N 0,8 
~ 
;!;. 
0 ,7 
0,. 
o 1'"' 
Time (min) 
6-16 
• control 
• nitrogen 
~ air 
• oxygen 
''''' 
2«> 
Figure 6.7 : Hydrogen peroxide degradatioa in nitrogeft-, air- aDd uygea:-saturated water 
after sonication id the uhl"UOl1ic: ceU for 16 b (11 = 3 per time period) 
The maximum standard deviation in the normalised hydrogen peroxide concentrations, as shown in 
Figure 6.7, for the oontrol and during nitrogen, air and oxygen satur.ltion are 0,038; 0,046; 0,049 and 
0,036 (mg L·I)I(mg L ·I) respectively. The rate of degradation over 3 h for each of the gases and the control 
was calculated from the regression of the data presented in Figure 6.7 using the linear regression model 
y = a+bx (6.4( 
where the coefficient b, the gradient of the regression line, represents the rate of hydrogen peroxide 
degradation. The calculated rates of hydrogen peroxide degradation after sonication for 16 b for a control 
and during nitrogen. air and oxygen saturation (and the standard errors of the calculated vaJues) are 
recorded in Table 6.4 . 
Table 6.4 : Rate of bydrogen peroxide degradation o\'er 3 bin nitrogen-. air- and oxygen-saturated 
",'ater after sonication in the ultnuoDic ceU for 16 b 
Gas 
control 
nitrogen 
ai, 
oxygen 
Rate of BIOz degradation 
(mio' l ) 
- 0,0003 ± 0,000 1 
- 0,00 12 ± 0,0001 
- 0,00 \3 ± 0,000 I 
-O,OOll ± O,OOO l 
0,543 
0,9\3 
0,945 
0,933 
Chapter 6 ULTRASOUND PRocESS lNvEsnGA110N 6-17 
Hydrogen peroxide degradation, as shown in Figure 6.7, in nitrogen-, air and oxygen-saturated water 
followed the same trend with almost identica1 rates within experimental standard deviation. The statistical 
F -test at a 95 % confidence level with multiple comparison using the Bonferroni method (Coetzer, 2000) 
indicated that there was no significant difference between the gradients of the nitrogen, air and oxygen 
regression lines wbereas the slope of the control regressjon line was significantly different from the other 
regression lines. 
The significantly lower rate of degradation for the control than that measured during saturation with any of 
the gases indicatcs that bydrogen peroxide degradation is increased by the presence of a gas but unaffected 
by the type of gas. Degradation is thus increased due to the pbysical phenomena such as agitation and 
mi..xing provided by a sparging gas but not influenced according to any chemical mechanism dependent on 
the dissolved oxygen concentration in solution. 
The fonnation of hydroxyl and hydrogen radicals in collapsing cavities and the formation of hydrogen 
peroxide from the radicals are stopped when sonication is ended. The remaining hydrogen peroxide in 
solution either decomposes to water and oxygen or is available to react with solutes in the water according 
to the reactions listed in Section 3.3.1. Decomposition to water and oxygen is slow as demonstrated by the 
control in Figure 6.7, hydrogen peroxide is thus available to react with organic compounds in solution. The 
implications for an industrial ultrasonic system is that equipment based on a flow system allows for 
ultrasonic degradation of organic pollutants and hydrogen peroxide fonnation to occur within the sonicated 
zone and further oxidation of pollutants by hydrogen peroxide to occur outside the sonicated zone. A gas 
such as oxygen may be used in a sonicated zone to increase dissolved oxygen ooncentration to enhance free 
radical fonnation. however, gas sparging beyond the sonicated zone has a negative impact leading 10 the 
decomposition of the available hydrogen peroxide. 
6.4 INTERVAL EXPERIMENTS 
The design of some ultrasonic reactors, as discussed in Section 2.5 .4 of Chapter 2. is based on a flow system 
in which a small volume. of limited diameter. is sonicated and the reaction solution is pumped through the 
sonicated zone. Some systems include a flow loop that oonsists of a sonicated zone and a holding tank or 
reservoir so that the reaction solution is circulated through the system. An example of a flow loop system is 
the Homefl sonochemical reactor shown in Figure 2.14 and Figure 2. 15 in Section 2.5.4. Scale-up of a 
flow loop system is favourable if the Limiting reagent of a reaction is retained in the sonicated zone or if the 
reagents exhibit a memory effect and remains activated after passing through the sonicated zone. 
The interval experiments performed to monitor hydrogen peroxide ooncentration when sonication was 
stopped for a cenain time period and then restaned arc reponed in Section 6.4.1. The interval experiments 
performed to monitor the effect of a change in gas during sonication are reported in Section 6.4 .2. 
Cbapter 6 ULTRASOUND PRoc:£ss iNvEsnGA110N 6-18 
6.4.1 Sonication period intervals 
Interval experiments were performed to monitor hydrogen peroxide concentration when sonication was 
stopped. for a certain time period and then restarted SO as 10 simulate a flow loop system with the solution 
moving in and out of a sonicated zone. Hydrogen peroxide concentration was measured in an 
oxygen-saturated solution at time periods of 0; 5; 10; 15; 20; 25; 30; 35; 40 and 45 min. Triplicate 
experiments were performed. for each time period.. SaturatiOD was achieved by sparging pure oxygen (at a 
100 kPa partial pressure) through the water for 10 min prior to and during the 45 min experiment at a flow 
rate of 6 mL S-I . The experiment was structured so that the water was sonicated for 15 ~ sonication was 
stopped for the following 15 min and then restarted for a further 15 miD.. Sonication in the ultrasonic cell 
was performed at an ac:oostic power of 57 W. A control was performed in water with no gas sparging. 
Hydrogen peroxide formation during a 15 min ):Criod of sonication followed by a 15 min period without 
sonication and a further 15 min period with sonication is presented in Figure 6.8. 
0,25 
~ 0,2 
.s 
c 
,g 0 ,15 
~8; 0 ,1 
S 
~ 0,05 
o ~ 
0 10 
.y • control • ... oxygen 
• I 
20 3J 50 
Time (min) 
Figure 6.8: Hydrogen peroxide formation in oxygen-saturated water in tbe ultrasonic cell 
during 15 min of MKlicatiou at an acoustic power of 57 W. 15 min witbout sonication and a 
furtber 15 miD witb sonication (11 = 3 per time period) 
The maximum standard de\<iation in the hydrogen peroxide concentrations measured in the control and 
oxygcn·saturated water, as shown in Figure 6.8, are 0,006 and 0,008 mg L-I , respectively. The rate of 
hydrogen peroxide formation in the control and oxygen·saturation solution during each of the 15 min 
periods was determined from the regression of each set of 15 min dala using the linear regression model 
y = a + bx 16.5) 
Chapter 6 ULTRASOUND PRocESS 1Nv£sncATION &-19 
where the coefficient b, the gradient of the regression line, represents the rate of hydrogen peroxide 
formation. The calculated rates of hydrogen peroxide formation in the control and oxygen-satura1ed water 
during interval experiments (and the standard error of the calculated values) are recorded in Table 6.5. 
Tab&e 6.5: Rate of bydrocea peroxide formatioe ill oxygeo-saturated water aDd a control iD the 
ultruoDk cell cIu.riJIg 15 IDi.a: of IOIlieatioD at .. acoustic power of 57 W, 15 m.i.a without IOIlic:atioD 
and a further 15 min with toaicatioa 
Gas 
control 
oxygen 
TIme 
(miD) 
o to 15 
15t030 
30 to 45 
o to 15 
15 to 30 
30 to 45 
\JItruou.od 
on 
off 
on 
on 
off 
on 
Rate of BlOt fonwioa 
(mg L-1 mia-t ) 
0 ,0051 ±-0,0003 
0,0000 ± 0,0000 
0,0044 ± 0,0002 
0 ,0070 ± 0,0004 
- 0,000 1 ±- 0,0002 
0 ,0054 ± 0,0002 
0,971 
0,000 
0,975 
0,971 
0,022 
0,981 
The rates of hydrogen peroxide formation presented in Table 6.5, as discussed in Section 6.2, were greater 
in an oxygen-saturated solution than in the control. Hydrogen peroxide concentration for both the 
oxygen-saturated solution and control increased during the periods of sonication and remained constant or 
decreased slightly during the period without sonication. The rates of hydrogen peroxide formation for both 
the control and oxygen-saturate<1 solution were lower in the second period of sonication than in the first 
period. 
The statistical F -test at a 95 % confidence level indicated that the rates of hydrogen peroxide formation in 
the control between 0 to 15 min and 30 to 45 min (detennined. from the regression model) were not 
significantly different whereas the peroxide formation rates in the oxygen-saturated solution between 0 to 
15 min and 30 to 45 min were significantly different. The lower peroxide formation during the second 
sonication period of oxygen-saturated water could be due to the increasing hydrogen peroxide degradation 
relative to fonnation, as discussed in Section 6.2 and Section 6.3, because of the greater peroxide 
concentration in solution. 
The peroxide concenttation profiles presented in Figure 6.8 show that ultrasound initiates radical reactions 
to produce hydrogen peroxide, hydrogen peroxide concentration remains COIlSlaDt in the absence of 
ultrasound and then increases again with further sonication. The slow degradation of hydrogen peroxide, as 
discussed in Section 6.3, indicates that in a flow loop system hydrogen peroxide fonned in a sonicated zone 
remains a,'3ilabIe for the oxidation of organic compounds C\'en as the reaction solution moves through the 
holding tank or reservoir. 
Chapter 6 ULTRASOUND PROCESS I.NvTsncATlON 6-20 
The solid lines shown in Figure 6.8 were obtained using the regression model (equation 6.4) that accounts 
for step changes in hydrogen peroxide concentration 
[H 20,:] = a + b{lime1 + c[lime -15][time > 15] + d[time - 30][time > 30] (6.41 
where (H~] is measured in mg L- I and time in mi.n. Estimates of the coefficients for the regn::ssjon model 
of the control and oxygen-saturated solution data and. the calcula1ed SUlDdard errors of the coefficients are 
recorded in Table 6.6. The terms of the regression model are significant at a confidence level of 95 %. The 
coefficient of variance (If) for the oxygen and control regression models were both 0,981. The regression 
models thus account for 99 % of the variance in the data. 
Table 6.6: Regressioa coetracats for the model of bydrogea peroxide fOl'1llation in oxygen 
saturated water ud a coatrol duriag 15 mi.n of 9ODicatioa at an acoustic power of 57 W. 15 min 
without sonication Dd a further 15 min with IODkatiOD 
estimate 
standard error 
Coatrol 
• b c d 
0,00539 0,00495 -0,00506 0,00458 
0,00230 0,00023 0,00039 0,00039 
Oxygen-saturated solution 
III b ed 
0,01208 0,00686 -0,00740 0,00599 
0,00291 0,00029 0,00050 0,00050 
An empirical model for hydrogen peroxide formation of the type presented in equation 6.4 can be used in 
the study of an ultrasonic loop reactor where the reaction solution is circulated through a small sonicated 
zonc. Equipment parameters such as the optimal time length between sonication periods and the presence 
of gas can be incorporated during sonication but not beyond require further investigation. 
6,4.2 Gas intervals 
Interval c:\.-periments were also performed to monitor the effect of a cbangc in gas during sonication at an 
acoustic power of 57 W. An oxygen-saturated solution was prepared by sparging pure oxygen through the 
solution for 10 min prior to and during sonication at a flow rate of 6 mL 5.1• Sonication was performed with 
a 15 min period of oxygen saturation followed by a 15 min period with nitrogen saturation and a further 
15 min period with oxygen saturation. Hydrogen peroxide concentration was measured at time periods of 
0: 5: 10: 15; 20; 25: 30; 35; 40 and 45 min. Triplicate experimcnts were performed for each time period. 
Hydrogen peroxide formation during sonication with a 15 min period of oxygen saturation followed by a 
15 min period with nitrogen saturation and a further 15 min period with oxygen saturation is presented in 
Figure 6.9. 
Cbapter 6 ULTRASOUND PRocESS lNvuncA110N 
0,25 
~ 0,2 
E 
c 
,g 0 ,15 
~ 
~ 0 ,1 
Z 
~ O,OS 
o 
o 10 
'" 
Time (min) '" 
6-21 
• nitrogen 
~ oxygen 
Figure 6.9: Hydrogen peroxide formatioa iD water during SOIlicatioe id the uJtruooic ceH at 
ao acoumc power ol57 W witb 15 mid ol oI)'ICII saturatioo, 15 mid ol Ditrogeo saturatioD add 
a further 15 miD of oxygen saturatioD (It - 3 per time period) 
The maximum standard deviation in the hydrogen peroxide ooocentrations measured in water saturated 
alternately with oxygen and nitrogen., as shown in Figure 6.9, is 0,004 mg L'I . The rate of hydrogen 
peroxide fonnation during the different 15 nUn periods (0 to 15 min; 15 to 30 min; 30 to 45 min) was 
detennined from the regression of each set of 15 ruin data using the linear regression model 
y = a + bx [6,7] 
where the coefficient b. the gradient of the regression line, represents the rate of hydrogen peroxide 
formation. The calculated rates of hydrogen peroxide fonnation during 15 min periods of oxygen and 
niuogen saturation (and the standard error of the calculated values) are recorded in Table 6.7. 
Table 6.7: Rate of bydrogen peroxide formation io ""ater duriog sonication in the ultrasonic cell at 
an acoeutic power of 57 W with 15 min of o~gen satunlltion, 15 min of oitrogen gturation and a 
further IS min of oxygen saturatioo 
Time 
(min) 
o to 15 
15 to 30 
30 to 45 
G .. 
oxygen 
nitrogen 
oxygen 
Rate of H2G,: formation 
(mg L' t min' l) 
0,0071 * 0,0002 
0,0025 * 0,0002 
0,0035 * 0,0003 
0.992 
0.924 
0,950 
Chapter 6 ULTRASOUND PRocrss INv!snGATION 6-22 
The rates of hydrogen peroxide fonnation recorded. in Table 6 .1 indicate that peroxide formation decreased 
significanUy when the gas was changed during sonication from oxygen to nitrogen. 10e rate of peroxide 
formation during the period with nitrogen sparging was comparable, within the standard deviation. to the 
formation rate, as recorded in Table 6.1 in Section 6.2. in a nitrogen-saturated solution. The rapid 
depletion of dissolved oxygen due to nitrogen sparging prevented the formation of peroxide being enhanced 
by the earlier saturation with oxygen. The rate of peroxide formation during the second period of oxygen 
saturation, as shown in Figure 6.9, was lower than that during the first period due to the overall rate of 
peroxide formation decreasing with increasing concentration. The swistical F-test at a 95 % confidence 
level indicated that the rates of hydrogen peroxide formation during oxygen saturation between 0 to 15 min 
and 30 to 45 min (determined from the regression model) were significantly different. 
A regression line (using the model in equation 6.4) was fined to the data presented in Figure 6.9 (the solid 
line in Figure 6.9) to account for step cbanges in the data. The coefficients for the model are recorded in 
Table 6 .8. Each term is significant at a confiQeooe level of95 % and the coefficient of variance (!il) for the 
regression model is 0,995. 
Table 6.8: Regression coeff"lcients for tbe model of' bydrogen peroxide formation in water during 
sonication in the ultraJOll)C cell at u acoustic power of ~1 W with l~ miD of oxygen saturation, 
IS min of nitrogen saturation and a further 1~ miD of oxygen saturation 
estimate 
standard error 
• 
0,00606 
0,00217 
b 
0,00725 
0,00022 
c 
~,oo505 
0,00037 
6,5 COMMERCIAL HYDROGEN PEROXIDE EXPERIMENTS 
d 
0,00118 
0,00037 
Hydrogen peroxide is used commercially for oxidation. Examples of hydrogen peroxide oxidation in water 
treatment are given in Section 3.3 .2 of Chapter 3. Hydrogen peroxide. as reported in Section 6 .2. is formed 
during sonication, however, peroxide concentration in the ultrasonic cell was onJy approximately 1 mg L·1 
after 16 h of sonication. The low sonocbentical formation of hydrogen peroxide can be supplemented with a 
commercial source of hydrogen peroxide. Ultrasound acts as a promoter for the formation of hydroxyl 
radicals from hydrogen peroxide. The use of hydrogen peroxide to enhance the ultrasonic degradation of 
pollutants, such as 2<hloropbeool, is reported in Section 2.4 .4. 1 of Chapter 2. Ultrasound is described as 
being more effective when used in combination \'\itb other conventionaJ water treatment processes than as a 
a stand~alone process. 
Hydrogen peroxide concentration was monitored during sonication of solutions prepared from a commercial 
hydrogen peroxide source (AR grade; Saarthem). Hydrogen peroxide formation was measured in accurately 
Chapter 6 ULTRASOUND PRocISS lNvuncATlON 
prepared hydrogen peroxide: solutions of awroximate 0,25; 0,50 and 0,75 mg L'I concentration during 
sonication in the ultrasonic cell for I h at an acoustic power of 57 W. The effect of gas on hydrogen 
peroxide: formation was investigated. by sparging nitrogen, air or oxygen through the solutions for 10 miD 
prior to and during sonication at a flow rate of 6 mL S'I . A conuol was performed in water with no gas 
sparging. Hydrogen peroxide: concentration was measured at time periods of 0 ; 5; 10; 20; 30 and 60 min. 
Triplicate experiments were performed for each time period for each gas. The hydrogen peroxide 
concentrations measured during the sonication of a 0,28 mg L'I hydrogen peroxide solution is presented in 
Figure 6. 10. 
~ 0,8 
g 
:8 0,6 
i ~ 0,4 ~ ,,-=*=~S==:t======: 8 
... 0.2 
Time (min) 
• control 
., nitrogen 
-. air 
... oxygen 
Figure 6,10: Hydrogen peronde formation iD aitrogcu-, air- &Dd oxygcu-saturated 0,28 mg L'I 
hydrogen peronde 9OIutions during sollicatiou iD the ultnuonic cell for 1 h at an acoustic pollt'cr 
of 57 W (If - 3 per time period) 
The maximum standard deviation in the hydrogen peroxide concentrations measured during sonication of a 
0,28 mg L'I peroxide solution for the control and during nitrogen, air and oxygen saturation are 0,009; 
0,008: 0.009 and 0,009 mg L'\ respectively. 
The hydrogen peroxide concentrations measured during the sonication of a 0.43 mg L' I hydrogen peroxide 
solution is presented in Figure 6. 11 . 
Chapter 6 ULTRASOUND PRocI"SS lNvunGA110N 
~ 0.8 
E-
 o 
0 0,. ~ 
• 
" 0 • g 0,4 0 
0 
N 
0 
N 
:I: 0.2 
0 
0 10 
Time (min) 
so 70 
• control 
.. nitrogen 
Ill; air 
~ oxygen 
Figure 6.11: Bydrogea peroxide formation in nitrogeo-, air- and oX}"IeIHI.turated 0,43 mg L'\ 
bydrogen peroxide lOIutioas during SOIlication in tbe ultruoak cell for 1 b at an acoustic power 
of 57 W (n = 3 per time period) 
The maximum standard deviation in the hydrogen peroxide concentrations measured during sonication of a 
0,43 mg LO \ peroxide solution for the control and during nitrogen. air and oxygen saturation are 0,017; 
0,010; 0,014 and 0,010 mg LO\ respectively. The hydrogen peroxide concentrations measured during the 
sonication of a 0,72 mg Lo1 hydrogen peroxide solution is presented in Figure 6.12. 
0 
~ 0,. -
 " 
I 0 , 
• ~ g 0.4 0 0 
N 
0 
N 0,2 :I: -
 L 
0 
0 10 70 
Time (min) 
• control j 
.. nitrogen 
'" air 
~ oxygen 
Figure 6.12: Hydrogen peroxide formation iD nitrogeD-, air- and oxygeD-saturated 0,72 mg LO \ 
bydrogen peroxide solutioos during sookatioo id tbe ultrasooic ccU for I b at an acoustic power 
of 57 W (n = 3 per time period) 
The maximum standard deviation i.n the hydrogen peroxide concentrations measured during sonication of a 
0.72 mg LO) peroxide solution for the control and during nitrogen. air and oxygen saturation are 0,020; 
Chapter 6 ULTRASOUND PRocrss INVI'SIlGA11ON 
0,009; 0,009 and 0,015 mg L- 1, respectively_ The rates of hydrogen peroxide formation during sonication of 
hydrogen peroxide solutions were calculated from the regression of the data presented in Figure 6.10, 
Figure 6.11 and Figure 6_12 using the linear regression model 
y=a+bx 16.81 
where the coefficient b, the gradient of the regression line, represents the rate of hydrogen peroxide 
formation_ The calculated rates of hydrogen peroxide formation in peroxide solutions during sonication 
with nitrogen, air and oxygen saturation (and the standard enor of the calculated va1ues) are recorded in 
Table 6.9_ 
Table 6.9: Rate of hydrogen pero:ride formation in 0,28; 0,43 aDd 0,72 DIg L-1 bydrogen pero:ride 
solution! saturated with nitrogeu, air Dd oxygea. during sonication in the ultruonic ceU at an 
acoustic power of 57 W 
Gas Rate of B2(h formation (mg L-1 min-I ) 
0,28 mg L-t If 0,43 mg L- l If 0,72 mg L-' If 
control 0,OO22± 0,0001 0,980 0,0024 ± 0,000 I 0,972 0,0017 ± 0,0002 0,866 
nitrogen - 0,0001 ±-0,0001 0,073 - 0,0007± 0,0001 0,787 - 0,0014 ± 0,0001 0,871 
ai, 0,0024 ± 0,0001 0,977 0,0025 ±- 0,0001 0,979 0,0015 ± 0,0001 0,906 
oxygen 0,0046 ± 0,0001 0,993 0,0029 ± 0,0001 0,986 0,0018 ± 0,0001 0,837 
The trends of hydrogen ~xide formation were similar in all of the hydrogen peroxide solutions, formation 
was greatest during oxygen saturation, the lowest during nitrogen saturation and similar during air 
saturation and the control. The negative rate of formation shown in Table 6.9 for nitrogen saturation 
indicates that hydrogen peroxide was being degraded since the initial concentrations of the commerciaJ 
hydrogen peroxide solutions were all greater than the equilibrium concentration (0,09 mg L- I ) of hydrogen 
peroxide during nitrogen sparging as shown in Figure 6.3(b) in Section 6.2 . L Hydrogen peroxide is both 
fonned and degraded during sonication from chain radical reactions such as that shown in Scheme 6. 1. 
H20 2 -+ HO· + HO· 
HO· + HO· - H20 + O· 
O· + H2 0 -+ H20 2 
Scheme 6.1 
la1 
Ibl 
Icl 
Hydrogen peroxide is degraded, as shown in reaction laJ of Scheme 6. 1, to (orm hydroxyl radicals that react 
to produce an oxygen radical. Hydrogen peroxide is reformed from the oxygen radical reaction with water, 
Chapter (; ULTRASOUND PRocESS 1NvEsnGA'OON 6-26 
reaction (c) of Scheme 6. 1. The reactions presented in Scheme 6. 1 are drawn from those presented in 
Scheme 2.4 in Section 2.3.1.2 and Scbeme 3.9 in Section 3.3.1. 
Oxygen radicals and molecules are formed, as sbown in Scbeme 2.3 and Scheme 2.4, from the radical 
reactions taking place during sonication (summarised in Appendix Hl, The radicals and molecules react 
further to continue the cycle of reactions. lbis was demonstrated with the oxygen radical formed in 
reaction [b) in Scheme 6. 1. However, as described in Section 6.1, oxygen is removed from solution during 
nitrogen saturation. Oxygen radicals thus formed from the degradation of hydrogen peroxide, such as in 
reaction (a] and rh] of Scbeme 6. 1, are scavenged by the nitrogen. The cycle of hydrogen peroxide 
degradation and reformation is broken in that peroxide degrades but does not reform since the intermediate 
oxygen species have been scavenged by the nitrogen. Peroxide concentration was thus found to decrease, as 
recorded in Table 6.9, during the sonication of nitrogen -saturated hydrogen peroxide solutions. 
The rate of hydrogen peroxide degradation during nitrogen saturation was found to be related to the initial 
hydrogen peroxide concentration of the solution. The rate of degradation increased from 0,000 1 to 
0,(X)}4 mg L" min'! as the concentration oftbe solution increased from 0,28 to 0,72 mg L'I . The number of 
peroxide degradation and reformation reactions taking plate is determined <1irectJ:y by the peroxide 
concentration, more reactions take place in more concentrated solutions. Thus during nitrogen saturation 
when all reformation reactions. are prevented from taking place, hydrogen peroxide concentration decreases 
quicker in a more concentrated solution since more degradation reactions are taking plate. The increasing 
rate of hydrogen peroxide degradation is shown in Figure 6. 13 to be linear as the initial concentration of the 
peroxide solution increases. 
0,006 
~ 
E " ~ 0,004 
-E-
 c 
0 
• control ~ 
• • § 0 ,002 L nitrogen 
" 
,.. air 
N 
.. oxygen 0 
N 
I 
'0 o • 
'" 
0: 
-0,002 
o 0,2 0,4 0,. 0,. 
H202 concentration (mg/ L) 
Figure 6.13 : Rates of hydrogen peroxide formatioD iD 0,28; 0,43 and 0,72 mg L' I commerciaJ 
hydrogen peroxide 5OIutions saturated with nitrogen, air and oxygen during soniutioD in tbe 
ullrasonic cell for 1 h at .lID acoustic power of 57 W 
Chapter 6 ULTRASOUND PRocESS iNvrsnGATION 6-27 
The hydrogen peroxide concentrations of the commercial peroxide solutions (0,28; 0,43 and 0,72 rng L'I) 
were lower than the equilibrium peroxide concentrations in the control and. for air and oxygen saturation 
(I to 1,5 rng L'I as shown in Figure 6.3(b) in Section 6,2,1), The concentration ofbydrogen peroxide in the 
commercia1 peroxide solutions thus increased during sonication although the rate of peroxide formation 
(shown in Figure 6 .13) decreased with increasing hydrogen peroxide concentration. The decreasing rate of 
hydrogen peroxide formation during oxygen saturation with i.ncreasing (xlllcentration of the peroxide 
solution (a power law relationship) is shown in Table 6.9 and. Figure 6.13. The rate of bydrogen peroxide 
formation decreased from 0,0046 to 0,0018 mg L- I min-I as the concentration of the peroxide solution 
increased. from 0,28 to 0,72 mg L-I . 
The rate of hydrogen peroxide formation in the 0,28 mg L'I hydrogen peroxide solution was significantly 
greater during oxygen saturation than in the control and during air saturation. Whereas the rate of 
hydrogen peroxide formation in the 0,72 rng L'\ hydrogen peroxide solution was the same (within standard 
deviation) in the control and during air and oxygen saturation. Thus the enhancing effect due to oxygen 
saturation. shown in Figure 6. 13, decreases with increasing concentration of a hydrogen peroxide solution. 
The implications for an industrial ultrasonic system are that o~gen or air sparging do not significantly 
enhance free radical reactions when hydrogen peroxide is added during sonication and that the appropriate 
hydrogen peroxide concentration should be investigated for a particular process. 
6.6 PROCESS CHEMlSTRY 
The sonochemical benefits of wtrasound. are not derived as in UV radiation due to the direct effect of energy 
on solutes in solution but rather due to highly reactive radicaJ species generated in the cavitation cavities. 
The high temperatures and pressures created in the coUapsiog gas cavities cause the thermal dissociation of 
.... '<Iter vapour to produce hydroxyl and hydrogen radicals. RadiaJ recombination reactions lead to the 
formation of hydrogen peroxide, bydrogen and water. Volatile poUutants arc degraded by pyrol)1ic 
reactions that occur in the gas cavity. The processes occurring in and around a collapsing gas cavity are 
iUustrated in Figure 2.6 in Section 2.3 .1.1. A comparison of bond dissociation energies is presented in 
Table 6. 10. 
Chapter 6 ULTRASOUND PRocr.ss IxvI:snGATION 6-28 
Table 6.10: Bood diuociatioa eaergies (Grec.wood ud Earnsbaw. 1984a; Buug et al., 1974; 
Kircbeaer.I981) 
HO-H - HO- + W 
HO-OH - 2HO 
HOO-H - HO; + H-
 0-0 - 20-
 H-H - 2W 
Boad diSlOCiatioa eDergy 
(kJ mort) 
497,9 
209,2 
376,6 
493,4 
435,1 
The bond dissociation energies listed in Table 6.10 indicate that if the water molecule is split under the 
conditions in the collapsing gas cavities to initiate all sooocbcmical reactions, the other molecules with 
lower bond dissociatioa eoergjes will also be split if pesent in the gas cavities. Thus. hydrogen peroxide 
formed from the reaction between two hydroxyl radicals can be split back to the radical species. Reactions 
also take place with volatile solutes in the collapsing gas cavity. 
The radical recombination reactions producing hydrogen. water and hydrogen peroxide from hydroxyl and 
hydrogca radicals also take place at the gas/liquid. interface when the species diffuse out of the cavity. The 
species may diffuse out of the cavity as they are formed or be released into the surrounding liquid on cavity 
collapse (Shah et al., 199ge). Hyd.ropbobic pollutants accumulate and react at the gasIliquid interface, 
degradation at the interface is due to both pyrolysis and radical reactions. The concentration of hydroxyl 
radicals and hydrogen peroxide at the interface is significantly higher than in the bulk solution 
(Neis, 2(00). 
Hyd.rophilic pollutants in the bulk solution are degraded by reaction with free radicals or hydrogen peroxide 
molecules that have diffused from the gas cavities. It is estimated that approximately 10 % of the hydro).)'1 
and hydrogen radicals genemted in the gas cavities escape to the bulk solution (Thompson and 
Doraiswamy, 1999). Reaction kinetics in the bulk solution arc analogous to that of radiation chemistry. In 
addition to the radical reactions, macromolecules and solid particles are degraded in the bulk solution clue to 
the shock wave and microjet formation triggered by the collapse of the gas cavities (Neis, 2(00). 
The concentration of hydrogen peroxide and hydrogen in the bulk solution increase linearly at ftrst, and 
then approach a limiting concentration due to further reaction or degradation in the gas cavities back to 
radicaJ species (Chanerjee et al .. 1983). The concentration of radicaJ species in solution rapidly approaches 
a stationary value due to their high reactivity. The steady state approximation states that the concentration 
or free radicals will be very low at all times, essentially remaining constanl since they act as intermediates 
in a ehain reaction (Chatterjee et al., 1983; Huang et al., 1974). 
Chapter 6 ULTRASOUND PRocESS lNVtSTIGATlON 6-29 
The radical species and products in solution can also dissociate to fonn ions that take part in further radical 
reactions. Perhydroxyl radicals dissociate at a pH greater than 4 (Scheme 6 .2), bydrogen peroxide and 
hydroxyl radicals dissociate in basic solutions at a pH greater than 11 (Spinks and Woods, 1976). 
HOi ..= Oi + H+ 
H20 2 ~ HOi + H+ 
HO· ~ 0 - + H+ 
Scbeme 6.2 
pK~4,88 
pK~ 11,6 
pK ~ 11,9 
lal 
Ib] 
le] 
A detailed list of radical reactions (from radiolysis and photolysis chemistry) thaI. may potentially take place 
during sonication (due to similar reaction conditions) are listed. in Appendix G. 
6,7 CONCLUDING REMARKS 
An ultrasonic laboratory reactor was designed and constructed to investigate the sub-processes occurring 
during sonication and to make recommendations for the scale-up of ultrasound processes. 
The formation of hydrogen peroxide during sonication indicated that the ultrasonic source in the ultrasonic 
ceU was sufficiently powerfuJ to induce cavitation in solution. The collapsing gas cavities being the sites of 
sonocbemistry initiation. Hydrogen peroxide concentration was used as a measure of the degree of radical 
reactions occurring during sonication and as a tool to judge under which process conditions the most radical 
reactions occurred in the ultrasonic cell. 
The effects of dissolved oxygen concentration and acoustic povier on hydrogen peroxide formation were 
investigated. De-()X)'genation was achieved by sparging nitrogen through a solution. The gases (oxygen. 
air and nitrogen) were sparged continuously during sonication to keep the dissolved oxygen concentration 
constant. Hydrogen peroxide formation increased with increasing dissolved oxygen concentration due to 
the additional radical reactions being initiated. As hydrogen peroxide ooncentration in solution built up the 
rate of degradation increased until an equilibrium was established and the rate of degradation equalled the 
rate of formation. Equilibrium was reached during sonication in the absence of oxygen within 30 miD and 
in the presence of o>""Ygen after 16 h. Changes in the rate of hydrogen peroxide formation with changes in 
aooustic power were only observed during the first 30 min of sonication. and only in the presence of oxygen. 
Ultrasonic processes with a short reaction time, should thus, as a guide, be performed with oxygen sparging 
and a high acoustic power for optimum effectiveness whereas long ultrasonic processes should be operated 
with air sparging and a lower acoustic energy source. 
Cbapter 6 ULTRASOUND PRocISS 11tvrsncA11ON 6-30 
A correlation was developed for bydrogen peroxide as a function of time, acoustic power and dissolved 
oxygen concentration to guide the choice of process conditions and mode of operation of an ultrasonic 
process. 
The rate of degradation of hydrogen peroxide after sonication was slow in the absence of any gas sparging. 
A sparging gas, irrespective of the type. significantly enhanced peroxide degradation due to mixing and 
agitation. The post-sonication pesence of hydrogen peroxide indicated the availability of peroxide in an 
ultrasonic loop system to oxidise organic compounds beyood the sonicated zone. Gas sparging with oxygen 
and air may be used in an ultrasonic reactor to enhance free radical formation, however. beyond the reactor 
in a flow loop system gas sparging has a negative effect and leads to the decomposition of hydrogen 
peroxide. Interval experiments demonstrated the formation of hydrogen peroxide during sonication, the 
constant peroxide conocntration in the absence of sonication and a further increase in peroxide 
concentration with renewed sonication. 
Sonication of commercial hydrogen peroxide solutions indicated that the enhancement in hydrogen 
peroxide formation, due to oxygen and air sparging, decreased. with increasing concentration of the added 
commercial hydrogen peroxide. Thus the occurrence of free radical reactions during sonication should 
either be enhanced with added hydrogen peroxide or gas sparging but DOt with both. 
To summarise, guidelines for the scale-up of ultrasonic processes: 
• 
• 
• 
• 
A gas (oxygen and air) or added hydrogen peroxide should be used to enhance radical reactions 
during sonication. If a gas, sparging should be continued throughout sonication. 
Ch..-ygen sparging and a high acoustic power should be used for processes with a short retention 
time, air sparging and a lower acoustic energy source for longer processes. 
The measurement of hydrogen peroxide concentration is a convenient tool to investigate parameters 
that enhance process operation and reactor performance. 
A flow loop system should be considered to maximise the oxidation of organic compounds both in 
and out of a sonicated zone. However. gas sparging beyond the sonicated zone should be avoided. 
7 
OZONE PROCESS INVESTIGATION 
Ozone is widely used in water treatment for colour, taste and odour removal, disinfection, oxidation of 
pollutants and flocculation enhancement. The oxidation of )X)lIutants during ozonation is enhanced by the 
addition of hydrogen peroxide. Other advanced oxidation processes, such as ultrasound, when used in 
conjunction with ozone may also enhance the performance of ozone in water treatment. This chapter details 
the investigation of ozone chemistry and the effects of combining ultrasound and hydrogen peroxide with 
ozone. The implications of including ozone in the scale-up an ultrasound process are also discussed. 
Ozonation experiments were performed in the ultrasonic cell (with and without ultrasound), Ozone was 
generated from oxygen in a Sorbios ozone generator and sparged through a solution in the ultrasonic cell 
similar 10 the gas e:\-'periments reported in Chapter 6. The measurement of dissolved ozone concentration is 
reported in Section 7. 1, ozone decomposition in Section 7.2, the formation of hydrogen peroxide during 
ozonation in Section 7.3 and ozone mass balances in Section 7.4. A summary of the experimental 
programme of the ozone process investigation is presented in Table 7.1 . 
Table 7.1 : Experimental programme of the ozone process investigation 
Section Title Measured Variables Data Statistics 
parameters Appendix Appendix 
7.1 Dissolved ozone [0,] ulttasound 0 .1 FA.I 
concentration hydrogen peroxide 
7.2 Ozone decomposition [0,] ozone 0 .2 F.4.2 
[H,o,] ultrasound 
7.3 Hydrogen peroxide [H,o,] ozone 0.3 FA.3 
formation ultrasound 
7.4 Mass balances [0,] ozone 0 .4 F.4.4 
[H,o,] ultrasound 
hydrogen peroxide 
The ex-perimental data discussed in this chapter is reponed (as indicated in :rable 7.1) in Appendix D. The 
statistical analysis (mean, 95 % confidence limits, standard deviation) is reponed in Appendix F. Standard 
deviation was calculated for each time period of an experiment though only the maximum value is reponed 
Chapter 7 O7.O!'l"£ PRocESS lNvrsnGATIO:-i 7-2 
in this chapter. Each data point in experiments measuring hydrogen peroxide or dissolved ozone 
concentration was measured in a separate experiment since the ultrasonic cell had to be opened during 
sampling (described in Section 5.3), thus altering the gas atmosphere in the cell. 
7_1 DISSOLVED OZONE CONCENTRATION 
Oxidation of organic poUutants by ozone during water treatment can take place due to direct reaction with 
ozone molecules or by reaction with free radicals fonned from the decomposition of ozone. The reaction 
scheme proposed. by Hoigne and co-workers for the decomposition of ozone is presented in Scheme 3.3 in 
Section 3. 1.2, the reaction scheme for the combination of ozone and hydrogen peroxide is presented in 
Scheme 3.6 in Section 3. 1.2. These reaction schemes are summarised in Appendix H on A3 paper that can 
be folded out. PoUutant oxidation is enhanced with the addition of hydrogen peroxide during ozonation. 
Hydrogen peroxide, as shown in Scheme 2.3 in Section 2.3.1.1 (repeated in Appendix H), is fonned during 
sonication. The combination of ultrasound with ozone may thus also enhance pollutant oxidation during 
ozonation. The amount of ozone available in solution for either direct reaction with pollutants or for the 
formation of free radica1s is determined by the transfer of ozone into solution. Parameters that affect ozone 
transfer are discussed in Section 3.1.2 and Section 3.2.2. 
Ozone chemistry was investigated by sparging the oxygen/ozone gas through water in the ultrasonic ceU 
and measuring the dissolved ozone concentration in solution. The ozone concentration in the 2,4 mL s·] gas 
stream was 5,7 mg L'] and the ozone production rate was 0,014 mg s·] (fable 5.6 in Section 5.2). Dissolved 
ozone concentration in water was measured during ozonation at time periods of 0; 1: 2.5: 5; 10; 20; 40 and 
60 min. Experiments were perfonned in duplicate for each time period. Dissolved ozone concentration was 
also measured during ozonation when the water was sonicated at an acoustic power of 57 W. when 
40 mg L ,] hydrogen peroxide 'was added prior to ozonation and when ozonation was combined with ooth 
hydrogen peroxide and sonication. Selection of the hydrogen peroxide concentration is described in 
Section E.2.3 of Appendix E. 3trazine degradation was used as the selection criterion. Atrazine degradation 
was greatest over a 4S min period during ozonation with 3 hydrogen peroxide concentration of IO mg L·1• 
hydrogen peroxide ooncentration was thus increased 4·fold fo r a 3 h ozonation period. Dissolved ozone 
concentration in water during ozonation. ozonation combined with sonication. ozonation combined with 
hydrogen peroxide and ozonation combined with ooth sonication and hydrogen peroxide are presented in 
Figure 7.1. 
Chapter 7 OZOl'l'E PROCESS iNvEsnCAll0N 7-3 
2 
~ 
.§. 1,5 
• c 
• i • • · 03 • • • 03/ uls ~ • • • • • • ~ 03/ H202 ~ 
0 
•• :lit 03/ ulsIH2021 u • • • 
* 
" 
• • • N 0,5 
r 0 
0 
0 10 20 3l 40 
'" 
EO 70 
Time (min) 
Figure 7.1: Dissolved ozone concentration in water during ozonation (0,014 mg S-l), ozonation 
combined with sonication (57 W), ozonation combined with hydrogen peroxide (40 mg L-t ) aod 
ozonation combined with soDic.don Dd hydrogen peroxide (If - 2 per tim~ period) 
The maximum standard deviation in the measured dissolved ozone concentrations in water during 
ozonation, ozonation combined with sonication, ozonation combined with hydrogen peroxide and ozonation 
combined with sonication and hydrogen peroxide are 0,108; 0,054; 0,030 and 0,001 mg L-1, respectively. 
The concentration of dissolved ozone in water during ozonation, as shown in Figure 7.1, increased to a 
maximum after 10 min and remained relatively constant, within standard deviation, after 20 min. This 
trend in dissolved ozone concentration is similar 10 thal reponed by Beltr:in and co-workers in which 
dissolved ozone concentration \\'3.5 found to increase rapidly during ozonation. reaching a stationary value 
after 10 nUn (Beltran et al .• 1994b; Beltrin et aI. , 1998). Dissolved ozone concentration after 10 min during 
ozonation combined with sonication was about 50 % (shown in Figure 7.1) of that during ozonation a lone. 
Ultrasound thus reduces the amount of dissolved ozone in solution during ozonation. 
The reduction in dissolved ozone concentration caused by sonication is due to reduced ozone transfer from 
the gas bubbles into the bulk solution as well as to the greater decomposition of ozone in solution. It is 
proJX>Sed that the ozone-containing bubbles act as nucleation sites for cavitation during the combination of 
ozonation with sonication. The ozone in the bubbles is thus subjected to the high temperatures and 
pressures created during bubble collapse in the compression stages of a sound wave. The slow 
decomposition of ozone in the gaseous phase 10 oxygen (at room temperature) is significantly enhanced by 
heat (Peleg, 1976). Ozone molecules under the conditions in the collapsing cavitation bubbles are destroyed 
by conversion to o"1'gen, reaction la] of Scheme 7.1 (repeated from Scheme 3. 1), and degradation to oxygen 
radicals. reaction [b) of Scheme 7.1. T he decomposition of ozone to o"1'gen in the gas bubbles reduces the 
ozone that is available to diffuse into the bulk solution. Oxygen radicals are also fonned according to 
Cbapter 7 OZONE PRocESS iNvEsnGATION 7-4 
reaction [cl of Scheme 7.1 from molecular oxygen present in the cavitation bubbles because the bond 
dissociation energy of oxygen is lower than that of water which is dissociated during cavitation (discussed 
in Section 6 .6). Oxygen is formed according to reactions [a] and [b] of Scheme 7.1 oot it is present in the 
cavitation bubbles predominantly since it is the carrier gas in which Olone is generated 
20, - 30, 
0 3 -+ O2 + O· 
0, - O· + O· 
Scbeme 7.1 
la) 
[b) 
lc) 
Weavers and Hofl'mann proposed that ozone first dissolves into solution and then rediffuses into the gaseous 
cavitation bubbles where it undergoes thermolytic decomposition as shown in Scheme 7.1 (Weavers and 
Hoffinann, 1998). The reduction in dissolved ozone concentration below the saturation value during 
olonation combined with sonication thus enhances the mass transfer effects and allows for more ozone to 
dissolve into solution as a function of time (Weavers and Hoffmann, 1998). The investigation of the 
degradation of 4-nitrophenol during ozonation combined with sonication by Barbier and Pettier found that 
degradation increased with increasing ultrasonic frequency (20 versus 500 kHz) because of the greater 
quantity of h}'dro:\:yl radicals escaping from the cavitation bubbles (Barbier and Petrier, 1996). 
Ozone molecules inside the cavitation bubbles and at the gasIIiquid interface also participate in the 
ultrasonic-initiated radical reactions listed in Scheme 2.1 to Scheme 2.4 (summarised in Appendix H) that 
occur in these regions. The reaction of ozone with hydroxyl radicals formed in the coUapsinC cavitation 
bubbles (reaction If] of Scheme 3.3) initiates a chain of radical reactions that leads to the decomposition of 
ozone (Scheme 3.3 in Appendix H). The reaction sequence of the chain propagation steps is reaction [1] of 
Scheme 3.3 followed by reactions le) ; fb] ; le) ; Id]; (e) and If] which continues the reaction cycle. 
Oxygen radicals generated from the decomposition of ozone (reactions [b) and (c) of Scheme 7. 1) propagale 
further free radical reactions since hydrogen peroxide is generated according to reaction la] of Scheme 7.2 
(repealed from Scheme 2.4). The hydrogen peroxide is degraded under the conditions inside the cavitation 
bubbles (discussed in Section 6 .6) to generate perhydroxyl and hydroxyl radicals according to reactions [b) 
and le] of Scheme 7.2 (repealed from Scheme 3.9). 
O' + H20 -+ H20} 
H20 2 -+ H' + HO; 
H,O, - 2HO· 
Scbeme 7.2 
[a) 
[bl 
[c) 
Cbapter 7 01.01\'£ PROCESS l"lVESTIGATION 1-5 
Ozone docomplSition due to conditions in the collapsing cavitation bubbles and due to participation in the 
ultrasonic-initiated free radical reactions occurs within the gas phase. Ultrasound, however, also initiates 
ozone decomposition in the bulk solution. Hydrogen peroxide is formed during sonication (Scheme 2.2 and 
Scheme 2.3 in Appendix H); the hydrogen peroxide either reacts further inside the cavitation bubbles or 
diffuses into the bulk solution as summarised in Figure 2.6 in Section 2.3. 1.1 . Hydrogen peroxide in the 
bulk solution dissociates (discussed in Section 6.6) to form perhydroxyl ions as demonstrated in Scheme 7.3 
(repeated from Scheme 6.2). 
H20 2 .... HOi + H" 
Scheme 7.3 
OLone molecules in the bulk solution react with the perhydroxyl ions formed from the dissociation of 
hydrogen peroxide in solution; ozone reacts with perhydroxyl ions but not directly with hydrogen peroxide 
(Staehelin and Hoigne, 1982). The reaction of ozone and perhydroxyl ions (reaction [dJ of Scheme 3.6 in 
Appendix H) initiates a chain of radical reactions that leads to the decomposition of ozone (Scheme 3.6 in 
Appendix H). Ozone decomposition initiated by the reaction of ozone with perhydroxyl ions HOi' is 
significantly greater than that initiated by reaction ",ith hydro),,'yl ions HO- since the reaction rate constant 
is over 4-fold greater. Staehelin and Hoigne calculated that ozone decomposition initiated by reaction with 
perhydroxyl ions was faster than that initiated by hydroxyl ions for aqueous solutions of hydrogen peroxide 
concentration above 0,0034 mg L-1 (Staehelin and Hoigne, 1982). 
The reduction in dissolved ozone concentration caused by sonication, as shown in Figure 7.1. is thus due to 
processes inside the cavitation bubbles reducing ozone diffusion into the bulk solution as well as due to 
enhanced decomposition in the bulk solution because of the reaction with perhydroxyl ions. 
The overshoot in dissolved ozone concentration during ozonation and when ozonation was combined with 
sonication txfore a stationary ozone concentration was reached. as shown in Figure 7.1, can tx due to the 
formation of hydrogen peroxide during ozonation (reactions [hI and [iJ of Scheme 3.3). Hydrogen peroxide 
starts to form as dissolved ozone concentration increases, however, the hydrogen peroxide dissociates to 
fonn perhydroxyl ions that reacts with the ozone. Thus, dissolved ozone concentration reaches a maximum 
once ozone decomposition (due to the reaction with pcrhydroxyl ions) becomes greater than the diffuo;:ion of 
ozone into solution, a correction in dissolved ozone concentration occurs and equilibrium is reached 
txtv.een the ozone diffusion into and ozone decomposition in solution. The maximum ozone concentration 
during ozonation combined with sonication is reached earlier than during ozonation alone (5 miD versus 
10 min) due to the enhanced formation of hydrogen peroxide. Hydrogen peroxide concentration increases 
faster because of the effect of ultrasound. Thus sufficient hydrogen peroxide to initiate ozone decomposition 
is fonned in a shorter time period causing the maximwn dissolved ozone concentration (as shown in 
Figure 7 .1) to occur at an earlier time and at a lower ozone concentration. The equilibrium concentration of 
Chapter 7 OZONE PRocESS bVESTIGAll0S 
ozone in solution during ozonation combined with sonication is thus lower (approximately SO %) than that 
during ozonation alone. 
No significant dissolved ozone was measured as shown in Figure 7.1 during ozonation in the presence of 
the added 40 rng L- l hydrogen peroxide. Similarly. no ozonation was measured during ozonation combined 
with sonication and hydrogen peroxide. The concentration of added hydrogen peroxide (40 rng L- l ) was 
significantly greater than that formed during sonication (shown in Figure 7.4 to be less than 0,5 mg L- l ) 
which caused a 50 % decrease in dissolved ozone concentration in solution. Ozone accumulation in 
solution during ozonation with added hydrogen peroxide is thus completely prevented since the ozone 
diffusing into the bulk solution reacts immediately with the perhydroxyl radicals formed from the 
dissociation of the added hydrogen peroxide. Beltran and ccrworkcrs reported that dissolved ozone 
concentration decreases as the added hydrogen peroxide concentration is increased; a 34 mg Lol hydrogen 
peroxide concentration was found to cause a 50 % decrease in the maximum dissolved ozone concentration 
(from approximately 3,4 to 1.7 mg Lol) during ozonation (at an ozone production rate of 0,22 mg 5-1) and 
that a 340 mg L-1 hydrogen peroxide concentration completely prevented any acc.umulation of ozone in 
solution (Beluan et al., 1998). 
A1though no dissolved ozone is present in solution during ozonation with added hydrogen peroxide, the 
decomposition of ozone initiates free radical reactions (as in Scheme 3.6 in Appendix H) in which highly 
reactive species are generated Hydroxyl radicals are formed during ozone decomposition ao::ording to 
reaction (g) of Scheme 3.6, the high reactivity of hydroxyl radicals is discussed in Section 1.3. The 
oxidation of organic compounds during the combination of ozone with ultrasound or hydrogen peroxide is 
thus more likel~' due to reaction y,ith free radicals (hydro",")'1 radicals) than direct reaction with ozone 
molecules. 
The addition of hydrogen peroxide during ozonation only affects ozone chemistry in the bulk solution (the 
region of h)'drogen peroxide dissociation) whereas ultrasound affects ozone chemistry both in the gas·phase 
of the ozone bubbles (because of the cavitation conditions and ultrasonic radical reactions) and in the bulk 
solution (because of the dissociation of hydrogen peroxide fonned during sonication). 
The statistical software package STATlSTlCA was used to model dissolved ozone concentration. the 
dependent variable. as a function of acoustic power and hydrogen peroxide concentration. the independent 
variables. Acoustic power values of 0 and 57 W represent the absence and presence of sonication during 
ozonation. Hydrogen peroxide concentrations of 0 and 40 mg Lol represent the absence and presence of 
hydrogen peroxide during ozonation. Surface response diagrams of dissolved ozone concentration as a 
function of the two independent variables were generated for time periods of 10 and 60 min representing a 
shon and an extended period of ozonation. The gradient profiles of dissolved ozone concentration 
Chapter 7 OWNE PROCESS iNvEsnCATION 7-7 
presented in Figure 7.2 indicate the conditions of acoustic power and hydrogen peroxide concentration 
where dissolved ozone concentration (plotted on the z·axis) is highest (dark grey) and 10\\'eSl (light grey). 
The circular symbols in Figure 7.2 indicate the experimental data points. 
= 0 
=o~ 
II!B 0.8 
_0.9 
_'2 
_1.5 
-...... 
= 0 
=02 
IIII!I 0.4 
_0.8 
_0.8 
-, 
_.bow 5 
(a) 10 miD 
15 25 35 45 ss .. 
Acoustic power (W) 
(b) 60 min 
Figure 7.2: Surface response diagrams of dissolved ozone concentration in water during 
(Y1,onation as a function of acoustic power and bydrogen peroxide concentration 
The similar gradient profiles of dissolved ozone concentration at 10 and 60 min during ozonation 
Figure 7.2(a) and Figure 7.2(b), indicates that the effects of ultrasound and bydrogen peroxide on dissolved 
ozone concentration do not change with time. Thus, the mechanism by which ozone concentration is 
rcducOO in solution by ultrasound and hydrogen peroxide is independent of time and will not change over an 
ozonation experiment. 
Chapter 7 OZON~ PRocESS iNvIsnGAl10N 7-8 
The effect of hydrogen p:roxide adctition (40 mg L- I ) on the reduction of ozone concentration is greater than 
that of sonication at an acoustic power of 57 W, the addition of hydrogen peroxide in Figure 7.2(a) nxluax1 
dissolved ozone concentration from 1,35 to 0 mg L- 1 whereas sonication only reduced ozone concentration 
from 1,35 to 0 ,65 mg L·
 '
 . The effect of sonication is shown to decrease with increasing hydrogen peroxide 
concentration since the gradient profile lines become more horizontal as peroxide concentration increases. 
The gradient profiles presented in Figure 7.2(a) and Figure 7.2(b) indicate that the reduction in dissolved 
ozone concentration obtained during sonication (without added bydrogen peroxide) is the same as that 
obtained during the ac:klition of between 15 and 20 mg L-1 hydrogen peroxide (without sonication). 
The gradient profiles presented in Figure 7.2(a) and Figure 7.2(b) also indicate the conditions where 
oxidation dne to the direct. reaction with ozone is favoured and where reaction with free radicals is favoured.. 
Oxidation of organic compounds dne to direct reaction with ozone molec:u.1es is favoured where dissolved 
ozone concentration in solution is highest, as during ozonation without sonication or added hydrogen 
peroxide. Conversely, oxidation of organic compounds due to reaction with free radicals is favoured where 
dissolved ozone concentration has been reduced by ozone decomposition, as during ozonation with 
sonication or added hydrogen peroxide. The total reduction of dissolved ozone in solution during ozonation 
with added hydrogen peroxide indicates thal oxidation of organic compounds is due only to free radical 
reactions whereas the partial reduction in dissolved ozone concentration during monation combined with 
sonication indicates that both direct reaction with ozone and free radical reactions can take place. The 
combination of ozonation with sonication can thus be applied to treat a mixture of organic compounds since 
both reaction mechanisms take place, hence compounds that preferentially react with either ozone or 
hydroxyl radicals will be oxidised.. Though, oxidation of organic compounds during ozonation is reported 
in literature to be predominantly due to free radical reactions (Belutn et al .. 1993b; Beltran et al., 1994a). 
7.2 OZONE DECOMPOSITION 
The kinetics of the decomposition of ozone was i.nvestigated by monitoring the decrease in dissolved ozone 
concentrntion of a saturated solution under different experimental conditions. An ozone-saturated solution 
was prepared by sparging the oxygen/gas gas through water in the ultrasonic cell for 20 min. The ozone 
concentration in the 2,4 mL S-I gas stream was 5,7 mg L·1 and the ozone production rate was 0.0 14 mg S-I 
(Table 5.6 in Section 5.2). Dissolved ozone concentration was measured over 40 min (after the 20 min 
saturation period) during ozonation. during sonication and when ozonation was combined with sonication. 
A oonlrOl was performed by measuring dissolved ozone concentration after the 20 min saturation period 
without ozonation or sonication. Duplicate experiments were performed for each time period. The 
dissolved ozone concentrations are presented in Figure 7.3. 
Cbapter 7 OZOl'l'l: PRocESS iNvEsnGA110N 7-9 
2 
::; 
'" 
§. 1,5 
v 
c ~ 0 · 03 ~ g • • • uls c • • • • 03/ uls g 
0 v control l u \t: • c 0 0,5 : 0 0 
0 10 2D 
'" 
.., 
Time (min) 
Figure 7,3: DilMtk'ed ozone COIIttIltratioe iD water iD the IlItruoDic ttlI after a 20 mln 
S8turatioo period mlJ'ut, ozooatioe (0.014 IIRI S' I), SOIlic:atioo (57 W) and OZOIIatioll combined 
witb sonication (If = 2 per time period) 
The maximum standard deviation in the dissolved ozone concentrations measured in the control and during 
sonication, ozonation,. and ozonation combined with sonication are 0,090; 0,033; 0,115 and 0,134 mg L' I, 
respectively. The decrease in ozone concentration of an ozone-saturated solution once ozonation had been 
Slopped, as shown by the control in Figure 7.3, indicates that ozone decomposition is relatively slow, ozone 
concentration decreased from approximately I,SO 10 1,35 mg L'I over the 40 min period (a 10 % reduction). 
The degradation rate constant was calculated (from linear regression of the data) to be 0,004 mg L'I min'l . 
The higher initial concentration of the control was due to the daily fluctuation in experimental conditions, 
the % error in ozone generation (calculated during the ozone mass balances. Section 0 .4) is 7,5 % . 
In comparison. the sonication of an ozone-saturated solution resulted in a rapid decrease in dissolved ozone 
concentration from 0.95 to 0,06 mg L'I within 20 min (a 94 % reduction). The reduction in dissolved ozone 
concentration during sonication as discussed in Section 7.1 is due to the decomposition of ozone to oxygen 
and the reaction in solution with perhydroxyl ions. The degassing nature of ultrasound would also 
contribute to reducing the dissolved ozone concentration of an ozone-saturated solution. 
The dissolved ozone concentration measured during ozonation combined with and without sonication is also 
compared in Figure 7.3. Dissolved ozone concentration remained relativel}' constant, decreasing slowl)', 
during ozonation in the absence of ultrasound whereas ozonation combined with sonication led to the rapid 
decrease in ozone concentration (within to min) from the saturation value (I ,SO mg L'I) to approximately 
0,5 mg L'I . Dissolved ozone concentration is reduced during the combination of ozonation with sonication,. 
however, radical reactions are initiated in solution. Thus the potential oxidation of organic compounds 
during ozonation is enhanced by combining ozonation with sonication. 
Chapter 7 OZONE PRoct:SS INvr.snGATION 7-10 
The dissolved. ozone concentration during ozonation combined with sonication in an ozone-saturated 
solution is shown in Figure 7.3 and in water with no pre-saturation in Figure 7.1. The dissolved ozone 
concentration in both experiments (either decreasing from saturation or increasing from zero) stabilised at 
0,5 mg L-1• This concentration is thus the equilibrium concentration of dissolved ozone during ozonation 
combined with sonication where the rate of ozone transfer into solution is equal to the rate of ozone 
decomposition initiated during sonication. 
7.3 HYDROGEN PEROXIDE FORMATION 
Hydroxyl radicals and hydrogen peroxide are formed in the free radical reactions, listed in Scheme 3.3 in 
Section 3.1.2 (summarised in Appendix 1-1), cxx::wring during ozonation. The combination of hydroxyl 
radicals, shown in Scheme 2.2 in Section 2.3.1.1 (Appendix 1-1), may also lead to the formation of hydrogen 
peroxide. Hydrogen peroxide formation was investigated during ozonation, sonication and wben ozonation 
was combined with sonication. Ozonation was performed at an ozone production rate 0[0,014 mg S-I and 
sonication at an acoustic power of 57 W. Hydrogen peroxide concentration was measured at time periods of 
0; 5; 10; 20; 40 and 60 miD. Duplicate experiments were performed for eacb time period. Hydrogen 
peroxide concentrations are presented in Figure 7.4. 
0.8 
:::; 
~ 0,6 
c 
0 
• · 03 I ~ J:: OA 
" 
c 
.6. 03/ UIS
 1 
• l 0 • uls c 0 0 N 0 0.2 N t :I: 
• 
• • 0 • I • • • • 
0 10 
'" 
:Jj .., 
'" 
00 70 
Time (min) 
Figure 7.4: Hydrogen peroxide concentration in water during ozonation (0,014 mg S-I), 
sonication (57 W) and ozonation combined with sonication (n "" 2 per time period) 
The maximum standard deviation in the hydrogen peroxide concentrations measured during ozonation, 
sonication and ozonation combined with sonication are 0 ,OO2~ 0,015 and 0,010 mg L-1, respectively. The 
rates of hydrogen peroxide formation in water during ozonation, sonication and ozonation combined with 
sonication were calculated from the regression of the data presented in Figure 7.4 using the linear 
regression model 
Cbapter 7 OZONE PRocI:SS INvI:snGATlON 7-11 
y = bx [7· 11 
where the cncfficient b, the gradient of the regression Line, represents the rate of bydrogen peroxide 
formation. The calculated rates of bydrogen peroxide formation in water during ozonation, sonication and 
ozonation combined with sonication (and the standard error of the calcula1ed values) are recorded in 
Table 7.2. 
Table 7.2: Rate of hydrogen peronde formatioe ill water ill the ultruoaic ceU duriDg ozooation 
(0,014 mg s·t),l)OGkatioa (57 W) a.ad ozooaDoa combined with lOdicatiott 
ozone 
ultrasound 
ozone/ultrasound 
Rate of Hla,; formatioD 
(mg L-t miB-t ) 
O,()()()() ± O,()()()() 
0,0034 ± 0,0001 
0,0082 ± 0,000 I 
0,000 
0,955 
0,992 
Hydrogen peroxide was DOl detected in solution (shown in Figure 7.4) during ozonation aJODe. As discussed 
in Section 7.1 and 7.2, the bydrogen peroxide formed during ozonation according to reactions (h] and li] of 
Scheme 3.3 woWd have dissociated. and reacted with the ozone in solution (reactions le) and Idl of 
Scheme 3.6). Thus preventing the accumulation of hydrogen peroxide in solution during ozonation. 
SllIehelin and Hoigne reponed thal hydrogen peroxide would not be formed during ozonation at pH values 
greater than 6 due to the reaction of ozone with the highly reactive JXrbydroxyl ions fonned from hydrogen 
peroxide dissociation (StaebeLin and Hoigne, 1982). Distilled. water with a neutral pH (around 7) was used 
in the ultrasonic cell . The pH of the solution was not measured during ozonation, aJtbough adsorption of 
carbon dioxide from the atmosphere may have made the water slightly acidic. 
Hydrogen peroxide accumulated in solution during ozonation combined with sonication since the rate of 
foonation was greater than the rate of dissociation and reaction with ozone molecules. A synergistie effect 
in the formation of hydrogen peroxide, shown in Figure 7.4 and Table 7.2, was obtained when ozonation 
was combined \\ith sonication. The rate of hydrogen peroxide formation was greater during ozonation 
combined with sonication than the sum of that formed separately during O20113tioo and sonication. Water, 
hydrogen peroxide and hydrogen gas, as shown in Scheme 2 .2 in Appendix H, are formed from the 
hydrogen and hydroxyl radicals produced in the collapsing gas cavities during sonication. Ultrasound, 
however, enhances the decoIDp.)Sitioo of ozone as shown in Figure 7.1 and discussed in Section 7.1. Thus 
more o:\:ygen radicals and molecules are produced when ozonation is combined with sonication. Additional 
hydrogen peroxide is formed by the reaction of the oxygen radicals and molecules with the hydrogen 
radicals (fonned from the sonication of water, Scheme 2. 1) that usually recombine to produce hydrogen gas 
(reaction le) of Scheme 2.2) in the absence of ozone. The concentration of'hydrogen peroxide generated in 
solution is greater than 0,0034 mg L-\ thus according to Staehelin and Hoigne (StaebeLin and Hoigne, 
Chapter 7 OZONE PROCESS INvEsnCATION 7-12 
1982) the reaction of ozone with perhydroxyl ions (reaction [dJ of Scheme 3.6) has a greater effect on the 
decomposition of ozone during ozonation combined with sonication than the reaction with hydroxyl ions 
(reaction [a] of Scheme 3.6). 
The statistical software package STATlSTlCA was used to model hydrogen peroxide concentration as a 
function of acoustic JX)wer and ozone concentration. Acoustic jXlwer values of 0 and 57 W represent the 
absence and presence of sonication. Ozone concentrntions of 0 and 5,7 mg L-1 represent the absence and 
presence of ozonation.. Surface response diagrams of hydrogen peroxide concentration as a function of the 
two independent variables were generated for time periods of 10 and 60 min, a short and extended period of 
e.xperimentation. The gradient profiles presented in Figure 7.5 indicate the conditions of sonication and 
ozonation where hydrogen peroxide concentration (plotted on the z-axis) is highest (dark grey) and lowest 
(Light grey). The cin:u1ar symbols in Figure 7.5 indicate the e.xperimental data points. 
i 
.~ 
1! 
" 
2.' ~ D o 1.' 
Cl 0.015 ~ 
EH 0.03 0 
.0.045 
.0.06 
.0.075 
.0.' _ ......
 -5 , 15 20 35 45 
" 
65 
Acoustic power foN) 
(a) 10 min 
•. ,
 ,., 
i •. ,
 .~ 3.' 
I 2S = 0 15 
CJO.1 ~ 
,",, 02 0 
_0.3 OS 
.0.4 
.0.5 
.os 
_.Ixwe ., , 
" 
20 35 
" " 
65 
Acoustic power foN) 
(b) 60 min 
Figure 7.5: Surface response diagrams of bydrogen peroxide concentration as a function of 
acoustic power and ozone concentration during ozonation of water in the ultrasonic cell 
Cbapter 7 OZONE PRocrss lNvPsnG.ATION 7-13 
TIle similar gradient profiles of ozone concentration at 10 and 60 m.in, Figure 7.5(a) and Figure 7.5(b), 
indicates that the effect of sonication and ozonation on hydrogen peroxide concentration does not change 
with time. Tbus, the mechanism by which hydrogen peroxide is formed during ozonation and sonication is 
independent of time and will not change during an experiment. 
The gradient profiles presented in Figure 7.5(a) and Figure 7.5(b) indicate that the concentration of 
hydrogen peroxide in solution is greatest during ozonation combined with sonication. Ozone concentration, 
shown in Figure 7. 1, is reduced by approximately 50 % during the combination of ozonation and sonication. 
Thus less ozone is available in solution for direct reaction with organic poUutants. however, the greater 
hydrogen peroxide concentration, shown in Figure 7.4, indicates that a greater degree of radical reactions 
occurs in solution. Thus, oxidation of organic compounds during ozonation combined with sonication is 
more likely to take place through radical reactions than the direct reaction with ozone molecules. 
7.4 MAss BALANCES 
The dissolved ozone concentration in solution as discussed in Section 7.1 is reduced when ozonation is 
combined with sonication or hydrogen peroxide. Ozone decoIDpJ:Sition was investigated by performing 
45 min mass balances 10 compare the amount of ozone generated in 45 m.in to the amount of un.reacted 
ozone in the system at the termination of the 45 min experiment. The amount of ozone decomposed or 
consumed by reaction with solutes in the water is calculated from the difference between the amount of 
ozone generated over the time span of the experiment and the unreacted ozone in the system at the 
termination of the experiment. The tota] amount of unreacted ozone at the termination of an experimenl 
consists of the dissolved ozone in solution in the ultrasonic cell, the ozone in the gas spaces of the system 
and the ozone that passed through the system and reacted with the ·potassium iodide in the gas traps. 
Mass balances were performed over a 4S min period of ozonation, ozonation combined with sonication, 
ozonation combined with hydrogen peroxide and ozonation combined with sonication and hydrogen 
peroxide. Ozonation was performed at an ozone production rate of 0,014 mg 5-1, sonication at an acoustic 
power of 57 W and hydrogen peroxide was added so as 10 prepare a 40 mg L- I solution. The experimental 
details of the mass balance: experiments are reported in Section 0 .4. Ozone and hydrogen peroxide 
concentrations were measured after the 45 min ozonation period except in solutions with added. hydrogen 
peroxide (40 mg L'I) where only ozone concentration was measured. The upper limit of the hydrogen 
peroxide concentrations that can be measured using the OMP method (described in Section 5.4. 1) is 
4,08 mg L- I . The average amounts of ozone generated in 45 min (Table 0 .4) and the amounts ofunreacted 
ozone at the end of the 45 min mass balance experiments (Table 0 .6) are summarised in Table 7.3. 
Chapter 7 OWNI. PR<x:1'.SS lNvEsnGA110N 7-14 
Table 7.3: Ozooc decomposition in water over 45 min during ozoaation. ozonation combined witb 
sonication, ozooatioo combined with bydntgeo peroxide aDd ozooatioa combined with sonication 
and bydrogeu peroxide 
Average Unreacted ozooe (DIg) % 
generated di .... ved gu-pbue ........ total (unreacted)l 
ozone (mg) ozoae ozone Kl traps (generated) 
ozone 54,6 0,64 4,18 47,09 51 ,9 1 95,1 
0,60 4,18 46,73 51 ,51 94,3 
average 0.62 4.18 46,91 51,71 94,7 
ozonelu1s 49,2 0,21 3,36 35,56 39,13 79,5 
0,23 3,36 36,25 39,84 81 ,0 
average 0,22 3,36 35,91 39,49 80,3 
ozone/H~ 49,0 0,00 3,30 35,25 38,55 78,7 
0,08 3,30 35,27 38,65 78,9 
average 0,04 3,JO 35,26 38,60 78,8 
ozonelu1s/H~ 50,8 0,04 3,07 34,37 37,48 73 ,8 
0,05 3,07 33,84 36,96 72,8 
average 0,05 3,07 34,10 37,22 73,3 
The ozonation mass balance (shown in Table 0.3) indicates that approximately 95 % of the ozone generated 
in 45 miD was detected in the system at the termination of the experiment. The 5 % difference in ozone 
mass is in part due to decomposition of ozone to oxygen (Scheme 3. 1 in Section 3.1. 1) and reaction of ozone 
with perh)'ciro)':yl ions since no hydrogen peroxide accumulation (shown in Figure 7.4) occurred in solution 
during ozonation. The 5 % difference in ozone mass is also due to the variability in the amount of ozone 
generated in 45 min. The average amount is reported in Table 7.3 whereas the actual triplicate values are 
reported in Table 0.4 in Appendix D. The error in the generated ozone in the ozonation mass balance is 
approximately 3,3 %; sample error is calculated from the standard deviation and sample mean recorded in 
Table F.26 in Appendix F. 
The tota1 mass of unreacted ozone in the different mass balances reported in Table 7.3 indicates that 
different amounts of unreacted O"Lone were measured at the termination of the 45 min experiments. The 
absolute values. however, cannot be compared as the rates of ozone generation were different. The % error 
in the rate of ozone generation throughout the mass balances was 7.5 %. Thus the amounts of unrcacted 
ozone are converted to frnctions of the ozone generated during the mass balance experiment. The unreacted 
ozone measured in the solution. gas·phase and KJ traps are reported in Table 7.4 as fractions of the ozone 
generated during the different mass balance ~riments. 
Cbapter 7 OZONE PRocr..ss IxvunGAnoN 7-15 
Table 7.4: V_reacted 0Z0De iD the solution, gu-pbue and KI traps at the tcrmiDatioa of the 
45 min cxperu.eats as a fractioD (%) 01* ~ratcd (J7,()BC duriltg the !liDS balaace c:rpcrimeats 
V_reacted OZODC (-.I.) decomposed or 
dissotvcd gu-pbuc 
_ .. reacted Ill.ODC 
..- ..- Kl traps (-/.) 
ozone 1,1 7,7 85,9 94,7 5,3 
ozone'ws 0,4 6,8 73,0 80,2 19,8 
ozone!H,o, 0,1 6,7 72,0 78,8 21 ,2 
ozone/ulsIH~ 0,1 6,1 67.1 73,3 26.7 
The amount of ozone remaining in the system at the termination of the 4S miD mass balance experiments is 
shown in Table 7.4 to decrease from approximately 95 0/_ to approximately 80 and 79 %, respectively, when 
ozonation is combined with either sonication or bydrogen peroxide. The increased ozone decomposition 
results in a lower dissolved. ozone concentration in solution as shown in Figure 7.1. The mechanisms of 
ozone decomposition are discussed in Section 7.1. The lncreased ozone decomposition in the gas bubbles 
(during cavitation) and in solution reduces the dissolved ozone in solution and hence the ozone exiting the 
ultrasonic cell. Thus, as shown in Table 7.4 the gas-pbase ozone and the ozone in the K.I traps is reduced 
when ozonation in combined with either sonication or hydrogen peroxide. 
The amount of ozone remaining in the system at the end of the ozonation combined with sonication mass 
balance (80 %) is similar, within experimental error, to that remaining at the end of the ozonation 
combined with hydrogen peroxide mass balance (79 %) although the ozone decomposition mechanisms arc 
different. As discussed in Section 7.1 ozone decomposition during ozonation combined with hydrogen 
peroxide (40 mg Lol) occurs in solution due to reaction with perhydroxyl ions formed from the hydrogen 
peroxide dissociation. Whereas ozone decomposition during ozonation combined with sonication occurs in 
the gas bubbles due to the cavitation conditions and radical reactions as weU as in solution due to reaction 
with perhydroxyl ions formed from the hydrogen peroxide generated during sonication (less than 
0.5 mg L·1) . The effect of ozonation combined with sonication on the dissolved ozone concentration in 
solution is shown in Figure 7.2 to be similar to that obtained during ozonation combined with between 
15 and 20 mg Lo1 hydrogen peroxide (without sonication). Hence the sintilar amount of ozone remaining in 
the system at the end of the mass balances in whlch ozonation is combined with either sonication or 
bydrogen peroxide. 
The lowest amount of ozone remaining in the system (approximately 13 %, Table 0.4) was measured al the 
end of the mass balance in whlch ozonation was combined with both sonication and hydrogen peroxide. 
Hydrogen peroxide is fonned during sonication (shown in Figure 6.3 in Section 6.2.1), however, the rate of 
formation decreases as the concentration of peroxide increases (discussed in Section 6.5) until the 
Cb.pier 7 OZON[ PRocrss lNvrsnGA'IlON 7-16 
equilibrium concentration is reached (discussed in Section 6.2). Hydrogen peroxide was shown 10 be 
degraded during sonication when the initial concentration of a hydrogen peroxide solution was greater than 
the equilibrium concentration (as during the nitrogen saturation experiments in Figure 6. 10 to Figure 6.12 
in Section 6.S). The equjhbrium hydrogen peroxide concentration during sonication in the presence of 
oxygen is shown in Figure 6.3(b) in Section 6.2.1 to be approximately 1 to 1,5 rng L'I . The initial rate of 
hydrogen peroxide formation during sonication is shown in Figure 7.4 to be lower than that during 
sonication combined with ozonation. The: cquihbrium concenuation of hydrogen peroxide during 
sonication combined with ozonation is thus greater, although it was not measured, than the equilibrium 
concentration during sonication without ozonation (1 to 1,5 rng L'I) . It is. however, unlikely to be as high 
as 40 rng L·' . Hydrogen peroxide is thus degraded and the concentration is reduced from 40 mg L" during 
ozonation combined with both sonication and hydrogen peroxide as it changes towards the equilibrium 
concentration. Hydroxyl radicals are formed from the degradation of hydrogen peroxide acc:ording to 
reaction la} of Scheme 6.1 . The reaction of the hydroxyl radicals with ozone (reaction rh} of Scheme 3.6) 
results in the reduced ozone measured. in the system when ozonation is oombined with both sonication and 
hydrogen peroxide (Table 7.4). 
An excess of ozone is evident in the system though less is available in solution for oxidation of organic 
compounds when ozonation is combined with either ultrasound or hydrogen peroxide. Oxidation is thus 
more likely to oocur due to reaction with free radicals when ozonation is combined with either ultrasound or 
hydrogen peroxide. 
7.S PROCESS CHEMISTRY 
Oxidation of organic solutes in water by ozone is due either to direct reaction "ith ozone molecules or 
reaction with free radicals fonned from the decomposition of ozone. The radical reactions occurring during 
ozone decomposition are listed in Scheme 3.3 in Section 3. 1.3. The combination of ozonation with 
sonication or hydrogen peroxide enhances ozone decomposition thus increasing the radical reactions 
occurring in solution and decreasing the dissolved ozone ooncentration. Hydrogen peroxiclc aclded to a 
solution or produced during sonication dissociates to form perhydroxyl ions. The reaction between ozone 
and the pcrhydroxyl ions, discussed in Section 7. 1, causes the decrease in dissolved ozone ooncentration and 
the enhanced free radical reactions. The reduction in dissolved ozone concentration during ozonation 
combined with sonication is also due to the ozone-containing bubbles acting as nucleation sites for 
cavitation. Ozone is decomposed to oxygen within the bubbles due to the high temperatures and pressures 
created during bubble collapse. Ozone decomposition due to the reaction of ozone with the ultrasonic 
initiated radical reactions occurs both within the gas bubbles and at the interface between the gas-phase and 
the bulk solution. The addition of hydrogen peroxide during ozanation on1y affects ozone chemistry in the 
bulk solution whereas ultrasound affects ozone chemisuy both in the gas-phase of the ozone bubbles as well 
as in the bulk solution. ()x.jdation of organic compounds during the combination of ozanation with either 
Chapter 7 OZONE PRocESS iNvI:snGA11ON 7-17 
ultrasou.od or hydrogen peroxide is thus more likely due to reaction with free radicals than to the direct 
reaction with ozone. A1thougb the combination of ozonation combined with sonication allows for both 
reaction mechanisms to occur. 
Ozonation is a gas-liquid reaction and is influenced by mass transfer. The presence of dissolved ozone in 
solution indicates that the free radical formation reactions are slow and take place in the bulk solution (as 
during ozonation and ozonation combined with sonication). Wbereas a negligible ozone concentration in 
solution indicates that reactions are fast and take place in the film layer SWTOUDding the ozone bubbles (as 
during ozonation combined with hydrogen peroxide and ozonation combined with both sonication and 
hydrogen peroxide). 
7.6 CONCLUDING REMARKS 
Ozone chemistry and the effects of combining ultrasound and hydrogen peroxide with ozone were 
investigated Ozone is a powerful though selective oxidant. Oxidation during ozonation takes place either 
due to direct reaction with ozone molecules or reaction with free radicals formed from the decomposition of 
ozone. 
The measurement of dissolved ozone concenttation indicated that ult:rasound reduced ozone concentration 
by approximately 50 % and added hydrogen peroxide (40 mg L-1) totally depleted the ozone in solution. 
The decrease in dissolved ozone concentration during O2onation combined with sonication is equivalent to 
that obtained with between 15 and 20 mg L·1 of added hydrogen peroxide (without sonication). The 
decrease in dissolved ozone concentration, partial or lotal, was due to reaction between ozone and 
perhydroxyl ions formed from the dissociation of bydrogen peroxide. Ozone decomposition during 
sonication also occurred within and at the interface of the cavitation bubbles due to the cavitation conditions 
and free radical reactions. The addition of hydrogen peroxide during ozonation only affects ozone 
chemistry in the bulk solution whereas ultrasound affects ozone chemistry 00th in the gas-phase of the 
ozone bubbles as well as in the bullc solution. 
The measurement of hydrogen peroxide concentration during ozonation, sonication and the combination of 
ozonation and sonication indicated that a synergistic effect occurred in the formation of hydrogen peroxide 
during the combination of ozonation and sonication. The enhanced rate of hydrogen peroxide formation is 
due to the reaction of ozone within the cavitation bubbles with hydrogen radicals that usually recombine to 
form hydrogen gas. The decreased dissolved ozone concentration in solution and the enhanced hydrogen 
peroxide formation during the combination of ozone with ultrasound indicates that the oxidation of organic 
compounds is more likely to take place due to the reaction with free radicals than the direct reaction with 
ozone. The oxidation of organic compounds is thus most effective when oionation is combined with either 
ultrasound or hydrogen peroxide so as to enhance the degree of radical reactions occurring in solution. 
Chapter 7 OZONE PRocESS iNvIsnGATION 7-18 
Process conditions are altered depending on whether ozone is combined with ultrasound or hydrogen 
peroxide. The total reduction of dissolved ozone in solution during ozonation with added hydrogen 
peroxide indicates that oxidation of organic com~ is due only to free radical reactions whereas the 
partial reduction in dissolved. ozone concentration during ozonation combined with sonication indicates that 
both reaction mechanisms, direct reaction with ozone and free radical reactions. take place. The 
combination of ozonation with sonication thus allows for the oxidation of compounds that preferentially 
react with either ozone or hydroxyl radicals. 
The equilibrium concentration of ozone in solution (0,5 mg L-1) was measured when ozonation was 
combined with sonication in the ultrasonic cell 1be implications for an ultrasonic system of sparging 
ozone through the sonicated zone is that ozone will be present to enhance the oxidation of organic 
pollutants, radical reactions will also be increased. than during sonication alone. The slow degradation of 
ozone after ozonation has been terminated indicates that in a flow loop system ozone will also be available 
to react with organic pollutants once the reaction solution has moved out of the sonicated zone. Pollutant 
oxidation can thus be achieved both in and beyond the sonicated zone with the inclusion of ozone in an 
u1trasonic process. 
8 
ATRAZINE EXPERIMENTS 
Atrazine is an effective and widely used herbicide. The stability of atrazine in the environment has lead to 
the contamination of potable water sources and is thus a potential health risk.. The potential of using 
ultrasound to degrade organic pollutants during water treatment is demonstrated in the ultrasonic cell using 
atrazine as a model compound. This chapter details the atrazine investigation in the ultrasonic cell. The 
ultrasonic degradation of atrazine is compared with that achieved during oxidation with ozone and "vith 
hydrogen peroxide. Atrazine solution chemistry during ozonation and sonication is discussed, the 
enhancing effect of hydrogen peroxide is demonstrated Potential degradation mechanisms are described to 
account for the degradation products identified using gas chromatography and mass spectroscopy. A 
summary of the experimental programme of the atrazine investigation is presented in Table 8.1 . 
Table 8.1 : Experimental programme of the atrazioe investigation iD the ultruonic cell 
Section Title Measured Variables Data Statistics 
parameters Appendix Appendix 
8.1 Atrazine chemistry latt] gas (0,) E.I F.5. ! 
8. 1.1 tntrasound 
8. 1.2 Ozone 10,] ozone E. l . 1 F.5. U 
ultrasound 
hydrogen peroxide 
8.1.3 Hydrogen peroxide [H,o,] ozone E.1.2 F.5.1.2 
ultrasound 
8.1.4 Mass balances 10,] ozone E.1.3 F.5.1.3 
[H,o,] ultrasound 
latr1 hydrogen peroxide 
8.2 Atrazine degradation 
8.2. 1 Ultrasound [atr) atrazine concentration E.2.1 F.5 .2.1 
gas (N,; 0,) 
8.2.2 Ozone [atr) ozone E.2.2 F.5.2.2 
ultrasound 
8.2 .3 Hydrogen peroxide latt] ozone E.2.J F.5.2.J 
ultrasound 
8.3 Product identification E. 3 
Cbapter 8 ATRAZI!'ro'"'£ EXPERIMEl'fI'S 8-2 
Experimental data discussed in this chapter is reported in Appendix E as indicated in Table 8.1. Statistical 
ana1ysis (mean. 95 % confidence limits, standard deviation) is reported in Appendix F. Standard deviation 
was calculated for eacb time period of an experiment though only the maximum value is reponed in this 
chapter. Each data point in experiments measuring hydrogen peroxide or dissolved ozone concentration 
was measured in a separate experiment since the uluasonic cell had to be opened during sampling 
(described in Section 5.3), thus altering the gas atmosphere in the cell . The ultrasonic cell did not have to 
be opened during atrazine sampling, atrazine concentration was thus measured continuously during an 
experiment. Atrazine was found not to interfere with the analytical methods used to measure ozone and 
hydrogen peroxide concentration (Section 5.4.4). Alrazine chemistry is discussed in Section 8. 1. The 
atrazine degradation investigation is reported in Section 8.2 and the identification of degradation products 
is detailed in Section 8.3 . 
8.1 A TRAZINE CHEMlSTRY 
Atrazine concentration in the environment is typically in the range 0 to 20 Ilg L-1. Staning concentrations 
in laboratory investigations of up to 50 mg L·\ . depending on the sensitivity and detection limit of the 
analytical method, have been reported in literature. The HPLC method used in this investigation, described 
in Section 5.3.4. measures atrazine concentrations in the I mg L-\ range. A concentration of 5 mg L' \ was 
thus used as an initial atrazine concentration. A1razine concentration in a standing 5 mg L-\ solution \\'as 
measured over 3 h as a degradation control at time periods of 0; 45; 90; 135 and ISO min. Atrazine 
concentration was also measured when oxygen was sp.uged through the solution at a flow rate of 6 mL sol. 
Experiments were performed in triplicate. The alrazine concentrations are presented in Figure 8.1. 
6 
rs 
c 
0 4 ~ 
• 
" c • 3 ~ 
0 
0 
• 2 .~ 
~ 
0 
• , 
, 
~ 
c 
c 
c 
0 
• ~ 
• 
• I o 
120 19> 
Time (min) 
I 
'''' 
210 
• cont rol 
o oxygen 
Figure 8_1: Atrazine concentration in a 5 mg L-1 atrazine solution in tbe ultrasonic cell over 
3 b in tbe absence of ullruound witbout and witb oxygen 5parging (rr = 3 per time period) 
Cbapter 8 A'TRAZIlII'E EXPERIME.I'ITS 8-3 
The maximum standard deviation in atrazine concentrations measured in the control, without and with 
oxygen sparging, are 0,34 and 0,42 mg Lol , respectively. The rates of atrazine degradation were calculated 
from the regression of the data presented in Figure 8.1 using the linear regression model 
y=a+br (S. I) 
where the coefficient b, the gradient of the regression line, represents the rate of atrazine degradation. The 
ca1culated rates of atrazine degradation in a control, without and with oxygen sparging (and the standard 
errors of the calculated values) are recorded in Table 8.2. 
Table 8.2: Rate of atraziae degradatioD in a 5 mg L-1 atrmae JOiutiOll in the ultruoDic ceU over 
3 b ia tbe abRace of ultruouad wit bout and ",itb oxygea IpargiDg 
Gas 
control 
control with oxygen 
Rate of atnziae degradatioa 
(mg L-1 min-I) 
- 0,00091 ± 0,00114 
- 0,0001" 0,OOOS2 
0,047 
0.002 
The low atrazine degradation rates reported in Table 8.2 and shown in Figure 8.1 indicate that atrazine does 
not degrade upon standing even with oxygen sparging. The statistical coefficient of variance (R2) of the 
regression models (Table 8.1) are low as the regression lines are almost horizontal. The statistical F-test at 
a 95 % confidence level indicated that there was no significant difference in the gradients of the regression 
lines for the control and with oxygen sparging. 
The chemistry of atrazine in solution during sonication is discussed in Section 8.1. I . during ozonation in 
Section 8.1.2 and in the presence of hydrogen peroxide in Section 8.1.3. Mass balances are reported in 
Section 8.1.4. 
8.1.1 Ultrasound 
Sonochcmical benefits (discussed in Chapter 2) are caused by both physical and chemical effects of 
ultrasound. Physical effects are induced by shockwave and microjet formation. and usually impact only 
polymeric and heterogeneous systems. Atrazine degradation during sonication is thus not likely to occur 
due to the ultrasonic physical effects but rather to the chemical effects, the formation of free radicals in the 
coUapsing gas cavities and their subsequent diffusion into the bulk solution. A diagram of the three 
reaction zones occurring during cavitation (gas cavity; gasIIiquid interface: bulk solution) and the processes 
occurring in each zone is presented in Figure 2.6 in Section 2.3 .1.1. 
Chapter 8 ATRAlJNE ExpERIME.NTS 8-4 
The low vapour pressure of atrazine (0,0385 mPa at 25 °C, Table 4.1 in Section 4.1.2) indicates that 
atrazine is relatively non-volatile and does not accumulate inside the collapsing gas cavities. Auazine is 
also a hydrophobic compound since it contains more hydrophobic groups (such as CHa, CH2 and CH) than 
hydrophilic groups (such as OH. COOH, CO ancS NHz). Hydrophobicity is discussed in Section 2.3. 1.1 . 
The hydrophobicity of auazine is demonstrated in that it is significantly more soluble in organic solvents 
than in water. The solubility of atrazine in ~'ater at 22 °C is 33 mg L·1 (fable 4.1) whereas the solubility in 
acetone is 3 1 ()()() mg L·1 (fomlin, 1991). Atrazine thus accumulates in the gas/liquid interfacial region 
during sonication because of Its non-volatility and hydrophobicity. Riesz and co-workers concluded that the 
gas/liquid interface was a region of low polarity with a high concentration of non-volatile hydrophobic 
compounds (Riesz et al ., 1990b; Riesz and Kondo, 1992). 
The gas/liquid interfac:e. as shown in Figure 2.6 in Section 2.3. 1.1 , is a region where both free radical 
reactions (Scheme 2.3; Scheme 2.4 ; Scheme 2.5) and pyrolysis reactions take place. The gas/liquid 
interface is a region with a high concentration of free radicals since the radicals formed within the gas 
cavities as well as the products of the reactions taking place inside the cavities (such as hydrogen peroxide) 
diffuse through the interface into the bulk solution. Atrazine thus reacts with the hydroxyl radicals present 
at the gaslliquid interface. 
The reaction of atrazine with hydroxyl radicals. as described in Chapter 4, leads to the formation of ooth 
dealkylated and hydrolysed degradation products. The mechanism by which dealkylated products are 
formed from the reaction of atrazine with hydroxyl radicals is shown in Figure 4.11 in Section 4.3. The 
formation of dealkylation products is summarised in Figure 8.2. 
Cl 
Cl 
N)..N ~ 
H3CH2CHN A N)lNHCH(CH3h 
o N)..N + H:P:z 
H CA NA N)lNHCH(CH ) 3 I 3 2 
H 2-acetamido-6-chloro-
 4-isopropy lamino-s-triazine 
atrazine 
Hp, 
deethylatrazine 
Figure 8.2 : Formation of dealkylated degradation products from tbe reaction of atrazine with 
hydroxyl radicals (Hapemau et al., 1995) 
Chapter 8 ATRA2INE EXPERIMENTS 8-5 
Dealkylation of atrazine by reaction with hydro",),1 radicals, as discussed in Section 4.3, is initiated at the 
side-chain carbon Cl. to the amine nitrogen (Hapeman et al. , 1995). The carbon is relatively electron-rich 
since the amine nitrogen provides additional electron density. Oxygen is required in the degradation 
pathway to fonn the intennediate radicals (shown in Figure 4.11) from which the acetamide and deethyl 
products are fonned. Hydrogen peroxide is produced as a byprocluct when both typeS of products are 
formed. The mechanism is sterically controlled, hence, the hydroxyl radical attack occurs preferentially on 
the ethyl side-cbain producing deethylatrazine than on the isopropyl side-cbain. 
The mechanism by which hydrolysed products are formed from the reaction of atrazine with hydroxyl 
radicals is similar to the mechanism shown in Figure 4.10 in Section 4.3 for the reaction of atrazine with 
hydroxyl ions. The formation of hydrolysed products is summarised in Figure 8.3 (from Figure 4.14). 
atrazine 
bydroxyatrazine "-... 
deethylhydroxyatrazinc 
Figure 8.3 : Formation of h)'drolysed degradation products from the reaction of atrazine with 
hydroxyl radicals (Adams and Randtke, 1992a) 
The formation of hydrolysed products by the reaction of atrazine with hydro",),1 ions. as discussed in 
Section 4.3. occurs by nuc1eophHic attack at the site of the C-C1 bond (Chan et al ., 1992). Oxygen is not 
required for the reaction to occur a nd hydrogen peroxide is not fonned as a byproduct during the reaction 
mechanism. Atrazine degradation due to reaction with hydroxyl radicals occurs more readily to produce 
dealkylated products according to the reaction mechanism presented in Figure 8.2 than to produce 
hydrolysis products according to the reaction mechanism presented in Figure 8.3 (Adams et al .. 1990). 
8.1.2 Ozone 
Atrazine chemistry was investigated during ozonation to elucidate potentia] degradation mechanisms. The 
effect of atrazine on the dissolved ozone concentration during ozonation was investigated by sparging the 
0"'1'gen/ozone gas through a 5 mg Lot atrazine solution in the ultrasonic cell and measuring the dissolved 
ozone concentration in solution. The ozone concentration in the 2.4 mL s'\ gas stream was 5,7 mg L· t and 
Chapter 8 ATRA1.lNE EXPERIMENTS 8"; 
the ozone production rate was 0,014 mg S-I (Table 5.6 in Section 5.2). Dissolved ozone concentration was 
measured in a 5 rng L- I atrazine solution during ozonation at time periods of 0; 1; 2,5; 5; 10; 20; 40 and 
60 min. Experiments were perfonned in duplicate for each time period. Dissolved ozone concentration was 
a1so measured during ozonation when the atrazine solution was sonicated at an acoustic power of 57 W, 
when 40 rng L-I hydrogen peroxide: Vt'3S added prior to ozonation and when ozonation was combined with 
both hydrogen peroxide: and sonication. Dissolved ozone concentration measured during the ozonation of 
water (Figure 7.1) is compared in Figure 8.4 to that measured in a 5 rng L- l atrazine solution. 
2 
"" 
g 1,5 
• ~ I j 0,5 
• 
• 
• 
• 
o · • • 
2 
"" 
..s 1,5 
o 
o 
• • (I 
•• 
o 
10 
• 
• 
10 
• 
• 
l 
• 
• • 
I 
• 
20 
• 
• 
• 
• 
Time (min) 
(a) water 
• 
• 
• 
Time (min) 
(b) 5 mg L-1 atrazine solution 
• 
• 
• 
• 
f 
• 
70 
70 
· 03 
.. 03/ ul s 
• 03/ H202 
"" 03/ uls/ H202 
• 03 
.. 03/ uls 
• 031H202 
-= 03/ uls/ H202 
Figure 8_4: Dissolved ozone concentration in water and a 5 mg L-I atrazine solution during 
ozonation (0,014 mg ,-1), O'LOnatiOD combined with sonication (57 W), ozonation combined witb 
hydrogen peroxide (40 mg L -I ) and ozonation combined with !J(ln.ication and bydrogen peroxide 
(n = 2 per time period) 
Cbapter 8 ATltAZII'O: EXPERlMEI\"TS 8-7 
The maximum standard deviation in the dissolved ozone concentrations measured in a 5 mg L'I atrazine 
solution during ozonation, ozonation combined with sonication, ozonation combined with hydrogen 
peroxide and ozonation combined with both sonication and hydrogen peroxide are 0,047; 0,037; 0,084 and 
0 ,030 mg L'\ respectively. 
The dissolved ozone concentration after 20 min during ozonation was reduced from approximately 
1,1 mg L'I in water, Figure 8.4(a), to 0,55 mg L,j in the atrazine solution, Figure 8.4(b). The reduction in 
dissolved ozone concentration is due to the ozone oonsumed in the oxidation of atrazine (mass transfer 
limitation). Atrazine oxidation during ozonation, as discussed in Section 4 .3. 1, occurs via two mechanisms, 
the direct reaction with ozone and the reaction with free radicals formed during ozone decomposition, Both 
dealk:ylated and hydrolysed. degradation products are formed from the reaction with free radicals whereas 
oruy deaIkylated products are formed from the direct reaction with ozone (Adams et al ., 1990; Adams and 
Randtke, 19913), The formation of dealk:ylated degradation products from the direct reaction with ozone 
(summarised from Figure 4.13) is presented in Figure 8.5 . 
Cl 
o N~N + H2~ 
Cl H C',Jlw,l)l.NHCH(CH ~ J I In 
N~N y H 2-acetamido..o--chloro-A )l 4-isopropylamino-s-triazine 
H, CH,CHN "N NHCH(CH,n 
atrazine 
deethylatrazine 
Figure 8.5: Formation of dealkylated degradation products from tbe direct reaction of 
atrazine witb ozone (Bolzaccbini et al., 1994) 
The direct reaction of ozone with atrazinc, as shown in Figure 8.5, produces both the acetamide and deethyl 
degradation products. The mechanism is sterically controlled because of the crowded transition state during 
the coordination of the ozone to the side-chain nitrogen atom. Hence, the reaction occurs preferenliall~' on 
the ethyl side-chain producing deethylatrazine than on the isopropyl side-chain, Hydrogen peroxide. as 
shown in Figure 8.5, is produced as a byproduct during the direct reaction of ozone with atrazine. 
The reaction of atrazine with hydroxyl radicals generated from the decomposition of ozone produces 
dealkylated and hydrolysed degradation products according to the reaction mechanisms presented in 
Chapter 8 A TRA2lNE EXPERIMUo'TS 
Figure 8.2 and Figure 8.3 . The radical reactions occurring in water during ozonation are listed in 
Scheme 3.3 (Appendix H) and presented graphically in Figure 3.3 in Section 3.1.3. Ozone decomposition is 
initiated by reaction with HO- ions, reaction [a1 of Scheme 3.3, generated from the dissociation of water. 
A1raz.ine reacts with the hydroxyl radicals formed in reaction le] of Scheme 3.3. The investigation by 
Beltrin and co-workers. described. in Section 3.1.2, reported that atrazine oxidation during ozonation is pH 
dependent (Beltr.in, 1995). Degradation al a pH less than 12 occurs due to both the di.rc:ct reaction with 
ozone and the reaction with free radicals whereas oxidation at a pH grealer than 12 is due only to reaction 
with free radicals. The pH of the atrazine solution during ozonation, Figure 8.4(b), was not measured. 
however, the initial neutral pH may have become slightly acidic due to catbon dioxide absorption from the 
atmosphere. The solution pH was below 12 and thus both atrazine degradation mechanisms could have 
occum:d. Atrazine degradation increases with increasing pH due to the higher concentration of hydroxyl 
ions in solution. Hydroxyl ions enhance the decomposition of ozone (according to Scheme 3.3) and thus the 
generation of hydroxyl radicals. 
OLonation is a gas-liquid reaction, the degradation of organic compounds dwing ozonation is thus 
determined by both mass transfer and chemical reaction. The kinetic regime of the absorption of ozone in 
solution according to the dimensionless Hana number (ratio of chemical reaction rate to physical absorption 
rate) is discussed in Section 3.1.2. The change in ozone concentration from within the gas bubbles, through 
the interfacial region to the bulk solution is presented in Figure 3.2 in Section 3.1.2. 
Ozone diffuses from the gas bubbles through the interfacial region into the bulk solution. As discussed in 
Section 3. 1.2 a fast reaction of a solute with ozone takes place in the proximity of the bubble interface 
resulting in a negligible amount, shown in Figure 3.2, of dissolved ozone in solution (Beltcin, 1995) . 
Conversely, dissolved ozone in solution indicates that oxidation reactions are slow and take place in the 
bulk solution. The presence of dissolved ozone in solution during the ozonation of atrazine. shown in 
Figure 8.4(b), indicates that the degradation of atrazine is due to slow reactions taking place in the bulk 
solution despite the higher concentration of atrazine at the gas-liquid interface. Atrazine is a hydrophobic 
compound (discussed in Section 8.1. 1) and thus gathers at the interface of the ozone bubbles. Beltcin and 
co-workers calculated the Hatta nwnbers to determine the kinetic regime of the direct reaction of atrazine 
with ozone and with free radicals. the Hana numbers indicated that the reactions are ver)' slow and take 
place in the bulk solution (Beltcin et al., 1993b). 
The dissolved ozone concentration after 20 min in a 5 mg L·1 atrazine solution, as shown in Figure 8.4(b), 
was reduced from 0,55 mg L·1 during ozonation alone to 0,35 mg L'l during O".lOnation combined with 
sonication. The reduction in dissolved ozone concentration caused by sonication is discussed in Section 7. 1. 
The ozone-containing bubbles act as nucleation sites for cavitation. O,lOne decomposition occu.rs because of 
the conditions within the collapsing cavitation bubbles as well as due to participation in the ultrasonic-
Cbapter 8 ATRAZlNE EXPERlMENTS 8-9 
initiated free radica1 reactions taking place within the bubbles and at the gaslIiquid interface. The enhanced 
ozone decomposition in the gas-phase reduces the transfer of ozone into the bulk solution. Ozone 
decomposition also takes place in the bulk solution because of the ultrasonic fonnation of hydrogen peroxide 
and the subsequent reaction of ozone with hydroxyl ions formed from the dissociation of hydrogen peroxide. 
Dissolved ozone concentration in the auazine solution during ozonation combined with sonication. 
Figure 8.4(b), is lower than that in water during ozonation combined. with sonication, Figure 8.4(a). Ozone 
is thus consumed in the oxidation of atrazine either via the direct reaction between ozone and atrazine or 
due to the reaction of atrazine with hydroxyl radicals generated during ozone decomposition. The presence 
of dissolved ozone in solution indicates that atrazine oxidation takes place in the bulk solution. However. 
the reduced amount of ozone in the atrazine solution during ozonation combined with sonication compared 
to during ozonation alone indicates that either atrazine oxidation is increasingly occurring at the gasIliquid 
interface (a high ozone demand at the gaslIiquid interface is shown in Figure 3.2 to reduce the dissolved 
ozone in solution, hence a change in kinetic regime from slow reactions in the bulk solution to moderately 
fast reactions at the interface) or that atrazine oxidation in the bulk solution is just significanUy enhanced. (a 
greater amount of ozone in solution is thus consumed and the dissolved ozone concentration decreases). 
Both changes in atrazine oxidation could possibly occur since ozone decomposition is enhanced both in the 
gas-phase and in the bulk solution as discussed in Section 7.1 when ozonation is combined with sonication. 
Ozone decomposition, Figure 8.4(a) and Figure 8.4(b), is enhanced by the formation of hydrogen peroxide 
during sonication. A commercial source of hydrogen peroxide can be added during ozonation in place of 
that generated during sonication to obtain a more direct and controllable effect. The reactions initiated in 
water during ozonation with hydrogen peroxide are listed in Scheme 3.6 (Appendix H), the cycle of reaction 
species is presenled graphica1ly in Figure 3.5 in Section 3.1.3. An insignificant amount of dissolved ozone 
was measured in .... 'ater during ozonation with 40 mg L-1 added hydrogen peroxide, Figure 8.4(a)_ whereas 
dissolved ozone was present in solution during the ozonation of a 5 mg L·1 alrazine solution in the presence 
of hydrogen peroxide, Figure 8.4(b). The concenlration profile of dissolved ozone in the auazine solution 
during ozonation in combination with hydrogen peroxide is similar to that measured during ozonation in 
combination with sonication. The lack of dissolved ozone in n'ater during ozonation was due to the fast 
reaction of ozone with the perhyclroxyl ions fonned from the dissociation of hydrogen peroxide, as discussed 
in Section 7.1. The presence of dissolved ozone in the atrazine solution indicates that atrazine must have 
disrupted this mechanism by taking part in the free radica1 reactions occurring in solution creating a 
preferentia1 reaction pathway involving the perhydro),:yl ions, preventing the ions from reacting with all the 
ozone in solution. Similarly, dissolved ozone in solution during ozonation combined with both hydrogen 
peroxide and sonication was prevenled from accumulating in water. Figure 8.4(a)_ but nOI in the alrazine 
solution, Figure 8.4(b). Dissolved ozone concentration in the atrazine solution increased to approximately 
0,2 mg L-1 at 5 min and then decreased steadily 10 zero over the 60 min experiment. 
Chapter 8 ATRAIDofE EXPEIUM£!\o'TS 8-10 
8.1.3 Hydrogen peroxide 
Hydrogen peroxide formation in a 5 mg L-1 atrazine solution was investigated during ozonation, sonication 
and when ozonation was combined with sonication. Ozonation was performed at an ozone production rate 
of 0,014 mg S·1 and sonication at an acoustic power of 57 W. Hydrogen peroxide concentration was 
measured at time periods of 0; 5; 10; 20; 40 and 60 min. Duplicate experiments were perfonned for each 
time period. Hydrogen peroxide concentration measured during the ozonation of water (Figure 7.4) is 
compared in Figure 8.6 to the hydrogen peroxide concentration measured in a 5 mg L-1 atrazine solution. 
2 
::; ! 1,5 -
 ~ c Jl ~ ii 
8 t 
S ' N 0,5 r-
 I 
o · • • 
o 10 
2 
::; 
~ 1,5 -
 ,g 
1 1 -
 ~ 0,5 r-
 • 
• 
• 1 
'" 
• 
• 
• 
• 
Time (min) 
(a) water 
• 
• 
• 
• 
• 
70 
• • 
• 
• o ____ ~~.L-____ ~ ____________________________ ~ 
o 10 
'" 
70 
Time (m in) 
(b) 5 mg L·t atrazine solutioD 
• 03 
• 03/ uls 
• ul s 
· 0 3 
• 03/ uls 
• uls 
Figure 8.6: Hydrogen peroxide concentration in water and a s mg L-I atrazine solutioD in the 
uJtruonic cell during ownation (0,014 mg 51), sonication (57 W) and ozonation combined witb 
sonication (n =- 2 per time period) 
Chapter 8 A TRADNE EXPERB1EI'fI'S 8-11 
The maximum standard deviation in the hydrogen peroxide concentrations measured in an atrazine solution 
during ozonation, sonication and ozonation combined with sonication are 0,029; 0,023; and 0,054 mg L-1, 
respectively. The rates of hydrogen peroxide formation in a 5 mg L-1 atrazine solution during ozonation, 
sonication and ozonation combined with sonication were calculated from the regression of the data 
presented in Figure 8.6 using the linear regression model 
y = bx [8.2J 
where the coefficient b, the gradient of the regression line, represents the rate of hydrogen peroxide 
formation. The calculated rates of hydrogen peroxide formation in a 5 mg L-1 atrazine solution during 
ozonatioD, sonication and ozonation oombined with sonication (and the standard error of the calculated 
values) are recorded in Table 8.3 . The rates of hydrogen peroxide formation in water (fable 7.2) are a1so 
included in Table 8.3 . 
Table 8.3: Rate of hydrogen peroxide formation in water and a S mg L-t atrwne Mtlution during 
ozonation (0,014 mg 5.1), Mtnication (57 W) and ownatiOD combined with toDic.ation 
water atmiDe solution 
Rate of BlOt formation Rate of BlOt formation If 
(mg L-1 min-I ) (mg L-I min-I ) 
ozone O,()()()() ~ O,()()()() 0,000 0,0285 ~ 0,0006 0,989 
ultrasound 0.0034 ± 0,0001 0,955 0,0036 ± 0,0002 0,934 
alone/ultrasound 0,0082* 0,0001 0,992 0,0283 ± 0.0010 0,965 
The rate of hydrogen peroxide formation in water during sonication, shown in Table 8.3, was the same. 
within standard deviation. as that measured in a 5 mg L·1 atrazine solution. Atrazine is not present inside 
the cavitation bubbles due to its low volatility and hence does not affect the mechanism by which hydro,,:yl 
radicals are formed. Hydrogen peroxide is produced from the reaction of two hydroxyl radicals (reaction (hI 
of Scheme 2.2) that occurs within the cavitation bubbles as well as at the gas/liquid interface (discussed in 
Section 2.3.1.1) where alrazine oollects due to its hydrophobicity (discussed in Section 8.1.1). Atrazine 
oxidation during sonication occurs by reaction with hydroxyl radicals which reduces the quantit), of radicals 
recombining lO form hydrogen peroxide. The oxidation of atrazine should thus result in a decreased rate of 
hydrogen peroxide fonnation in an atrazine solution in comparison with that in water. The similar rates of 
hydrogen peroxide fonnation in water and the atrazine solution as shown in Table 8.3 indicates that very 
few hydroxoyl radicals are consumed in the oxidation of atrazine. The rate of atrazine oxidation during 
sonication in the ultrasonic ceU is thus expected to be low. The similar rates of hydrogen peroxide 
fonnation in water and in the atrazine solution during sonication may also be a result of the hydrogen 
Cbapter 8 ATRA21NE ExrERIMENTS 8-12 
peroxide that is formed as a byproduct during atrazine oxidation. Hydrogen peroxide, as shown in 
Figure 8.2, is formed as a byproduct when dealkylated degradation products are produced from the reaction 
of atrazine with hydroxyl radicals. 
The formation of hydrogen peroxide during ozonation was significantly affected by the presence of auazine. 
Hydrogen peroxide was not formed during ozonation in water (shown in Table 8.3) whereas the rate of 
formation in the atrazine solution was 0,0285 mg L·1 min·l • The lack of hydrogen peroxide formation in 
water during ozonation is due to the fast reaction of ozone with perhydroxyl ions formed from the 
dissociation of hydrogen peroxide. The significantly greater rate of hydrogen peroxide formation in the 
atrazine solution is due to the oxidation of atrazine since hydrogen peroxide is formed as a byproduct in the 
reaction mechanisms producing the dealkylated degradation products. Hydrogen peroxide is generated 
when dealkylated attazine products are formed from both the direct reaction with ozone (Figure 8.5) and the 
reaction with hydroxyl rad1cals (Figure 8.2). Thus more hydrogen peroxide is generated from the oxidation 
of atrazine than what is depleted through the reaction of ozone with perhydroxyl ions. the dissociated form 
of hydrogen peroxide. 
The rate of hydrogen peroxide formation in an atrazine solution when ozonation was combined with 
sonication is also shown in Figure 8.6(b) to be significantly greater than that in water, Figure 8.6(a). 
Hydrogen peroxide, as described. earlier, is produced as a byproduct during the oxidation of atrazine by the 
direct reaction with ozone and the reaction with hydroxyl radicals to produce dealkylated degradation 
products. Hence the greater rate of hydrogen peroxide formation in the atrazine solution. Hydrogen 
peroxide formation in water when ozonation was combined with sonication in comparison ·with O2o03tion 
alone, as sho'wn in Figure 8.6(a). was enhanced due to the ultrasonic·initiated decomposition of ozone 
generating more rad1cal reactions and an increased rate of hydrogen peroxide formation. Similarly. the 
hydrogen peroxide formation in the atrazine solution is expected to be greater during ozonation combined 
with sonication in comparison with ozonation alone because of the ultrasonic·initiatcd decomposition of 
ozone as well as the greater atrazine oxidation producing more hydrogen peroxide as a byproauct. The rates 
of bydrogen peroxide formation during ozonation alone and during ozonation combined with sonication. 
however, are shown in Table 8.3 to be the same within standard deviation. Hydrogen peroxide fonnation is 
greater during ozonation combined with sonication. however as described. in Section 6.5, hydrogen peroxide 
decomposition via radical reactions increases with increasing peroxide concentration in solution reswting in 
similar overall rates of hydrogen peroxide formation during ozonation combined with sonication and during 
ozonation alone. 
8.1.4 Mass balances 
Atrazine chemistry during sonication. ozonation and hydrogen peroxide has been investigated through the 
measurement of dissolved ozone and hydrogen peroxide concentrations in a 5 mg L·I atraz.ine solution. as 
reponed in Section 8.1.2 and Section 8.1.3 . Atrazine oxidation during ozonation, sonication and hydrogen 
Cbapter 8 A TRA1IN£ ExpERlMEI".-rs 8-13 
peroxide addition was investigated by performing mass balances over 45 m.in to relate atrazine degradation 
to ozone decomposition. The amount of ozone decomposed or consumed in the oxidation of auazine is 
calculated from the difference between the amount of ozone generated in 45 m.in and the unreacted ozone in 
the system at the termination of the experiment. The total amount of unreacted ozone consists of the 
dissolved ozone in solution in the ultrasonic cell, the ozone in the gas spaces of the system and the ozone 
that passed through the system and reacted with the potasSium iodide in the gas traps. 
Mass balances in a 5 mg L-1 atrazine solution were performed over a 45 m.in period of ozonation, ozonation 
combined with sonication, ozonation combined with hydrogen peroxide and ozonation combined with 
sonication and hydrogen peroxide. Ozonation was performed. at an ozone production rate of 0 ,014 mg 5-' , 
sonication at an acoustic power of 57 W and hydrogen peroxide was added so as to prepare a 40 mg L-1 
solution. The experimentaJ details of the mass balance experiments are reported in Section E. !.3. Ozone 
and hydrogen peroxide concentrations were measured after the 45 min ozonation period except in solutions 
with added hydrogen peroxide (40 mg L-') where only ozone concentration was measured (detailed in 
Section 7.4). The average amounts of ozone generated in 45 min (Table 0 .4) and the amounts of unreacted 
ozone at the end of the 45 min mass balance experiments in an atrazine solution (fable E.4) are 
summarised in Table 8.4. The mass balances in water are presented in Table 7.3 in Section 7.4. 
Table 8.4 : Ozone decomposition and atrazine oxidation in as mg L-1 atrazioe solution ()\.'er 4S min 
during ozonation, ozonatioD combined with sonication, ozonation combined with hydrogen pero:lide 
and ownation combined with sonication and hydrogen peroxide 
Average Unreac:ted ozone (mg) % % 
generated dissolved gas-phase ozone in KI total (uoreacted)! atrazine 
ozone (mg) ozone OZOD. traps (generated) degraded 
ozone 54,6 0.16 3,82 43,42 47,40 86,8 26,0 
0,23 3,82 45,56 49,61 90,9 25,4 
i"erage 0,20 3,82 44,48 48,50 88,9 25,7 
ozone'uls 49,2 0. 12 2,89 30,65 33,66 68,4 36,3 
0, 1 I 2,89 30,34 33,34 67,8 30.2 
a, 'erage 0,11 2,89 30,50 33,50 68.1 33,3 
ozone/H2~ 49,0 0,14 3,18 33,61 36.93 75,4 41 .9 
0.22 3,18 34,39 37,79 77.1 46,9 
0.20 3,18 33,68 37,06 75,6 37.6 
average 0,19 3,18 33,89 37,26 76,0 42 ,1 
ozonelulsIH2~ 50,8 0,02 2,67 29,42 32, 11 63 ,2 34,5 
0.01 2,67 29,64 32.32 63 ,6 35,3 
anrage 0.02 2.67 29,53 32,22 63,4 34,9 
Chapter 8 ATRA2lNE EXPEIUMEl'roTS 8-14 
The amOUnlS of unreacted ozone presented in Table 8.4 are converted to fractions of the ozone generated 
because of the different rates of ozone generation in the mass balance experiments. The unreacted ozone 
measured in the solution, gas-phase and Kl traps are reported in Table 8.5 as fractions of the ozone 
generated during the different atrazine mass balance experiments. 
Table 8.S: Unreacted ozone in IOlution. gu-phue and KI traps at tbe termination of tbe 45 min 
experimenu as a fractioD (~o) of the generated ozoae during the atrazi.ae mass balance: experiments 
Unreacted ozone (0/0) decomposed or 
diQOlved gas-pbue ozone in total reactetl ozone 
OZOfte 07.0ftt KI traps (~o) 
ozone 0,4 7,0 81 ,5 88,9 11, 1 
ozone/uls 0,2 5,9 62,0 68,1 31 ,9 
ozone/Hz{)z 0,4 6,5 69,2 76,1 23 ,9 
ozone/uls/HzO:z 0,0 5,3 58,1 63,4 36,6 
The total unreacted amount of ozone in all experiments in an atrazine solution (Table 8.5) decreased in 
comparison \\ith that measured in water (Table 7.4) due to the ozone consumed in the oxidation of atrazine. 
Atrazine oxidation during the ozonation experiment reduced the final mass of ozone from 95 % to 89 %. 
Dissolved ozone concentration in solution., Figure 8.4(a). and total unreacted ozone, Table 8 .4. was the 
highest (89 %) during the ozonation experiment yet the lowest atrazine degradation (26 %) was measured 
The direct reaction of atrazine with ozone (mechanism presented in Figure 8.5) is thus less important in the 
oxidation of atrazine than the reaction with hydroxyl radica1s (mechanism presented in Figure 8.2 and 
Figure 8.3). A1razine degradation was greater in the ozone combination technologies (\\ith ultrasound or 
hydrogen peroxide) that promoted free radica1 reactions. The implications for an ozone process system in 
water treatment is that ozone should be used in combination with technologies that promote free radical 
reactions for enhanced oxidation of organic poUutants. 
Atrazine degradation over 45 min was the greatest during ozonation combined with hydrogen peroxide 
(approximately 42 %) although the amount ofunreacted ozone only decreased from approximately 79 % in 
water to approximately 76 % in the atrazine solution (Table 8.4). Enhanced ozone decomposition in the 
presence of hydrogen peroxide has been described in Section 7. 1 to be due to the reaction of ozone with 
perhydroxyl radicals formed from the dissociation of hydrogen peroxide. However. the small decrease in 
ozone decomposition (3 %) indicates that the bulk of the hydroxyl radicals produced that reacted with the 
atrazine originated from the hydrogen peroxide molecules (reaction [b] of Scheme 3.9). 
The oxidation of atrazine due to the direct reaction between ozone and ·atrazine is not influenced by 
ultrasound. A similar degree of oxidation according to this reaction occurs during ozonation alone and 
Chapter 8 A TRAZll"il: ExrERJME1"fI'S 8-15 
during ozonation combined with sonication. Thus the enhanced oxidation of atrazine during ozonaLion 
combined with sonication (33 % versus 26 %) is due to the alternate mechanism of atrazi.ne oxidation. the 
reaction of atrazine with hydroxyl radicals. Hydroxyl radical formation during ozonation alone is enhanced 
during ozonation combined with sonication, as described in Section 7.1, due to the greater ozone 
decomposition and the radicals formed during sonication. 
The dissolved ozone concentration in an atrazine soluLion during ozonation combined with sonicaLion \\'llS 
the same as that during ozonaLion combined with hydrogen peroxide, Figure 8.4(b), however, atrazine 
degradation was lower (33 % versus 42 %) since less hydrogen peroxide is formed in solution during 
ozonation combined with sonication. Figure 8.7(b), than what was added (40 mg L"I) during ozonation 
combined with hydrogen peroxide. The mechanism of atrazine oxidation when ozonation is combined with 
sonication or hydrogen peroxide is the same, in that it is due to the reaction with hydroxyl radicals. 
However. less atrazine degradation is observed during sonication since a lower amount of hydrogen 
peroxide is generated in solution. 
The total amount of ozone remaining in the system after the 45 min experiment was lower (shown in 
Table 8.4) during ozonation combined with sonication (72 %) than with added hydrogen peroxide (79 %). 
Ozone is degraded during sonication, described in Section 8.1.2, since the ozone bubbles act as nucleation 
sites and the ozone molecules are subjected to the high temperatures and pressures inside the collapsing 
cavities. Ozone is destroyed in the collapsing cavities, however, additional oxygen radicals, shown in 
Scheme 8. 1. are formed 
The highest decomposition of ozone (approximately 37 %) occurred when ozonation "''as combined with 
both sonication and hydrogen peroxide although atrazine degradation was not the greatest. atrazine 
degradation was similar to that acrueved when ozonation was combined with only sonication. Thus more 
ozone is completely decomposed to oxygen (reaction la] of Scheme 7.1) and does not contribute to the 
degradation of atrazinc. The investigation by Beluan and co-workers found that atrazine oxidation during 
ozonation was dependant upon hydrogen peroxide concentration (Beltnin et al.. 1998). Atrazine oxidation 
increased with increasing hydrogen peroxide concentration up until between 3,4 and 34 mg L"'. after which. 
atrazine degradation decreased with increasing hydrogen peroxide concentration. Atrazine oxidation in the 
presence of a peroxide concentration greater than 3 400 mg L'] is less than in the absence of hydrogen 
peroxide. The ncgaLive relaLionship between atrazine degradation and peroxide concentraLion is related 10 
the kinetic regime of ozone absorption which at rugh peroxide concentrations is mass transfer limited.. the 
rate of oxidation of solutes in solution becomes inversel)' proportional to the hydrogen peroxide 
concentration (Beltnin et aI., 1998). Atrazine degradation at high hydrogen peroxide concentrations occurs 
oruy due to reaction with hydroxyl radicals. the concentraLion of hydro:~:yl radicals is a function of the 
maximum physical rate of ozone absorption (Beluan et al., 1998). Similarly, the lower atrazine oxidation 
Cbapter 8 ATRA.27NE EXPERII\1E!Io7S 8-16 
in the uluasonic cell during ozonation combined with both sonication and hydrogen peroxide can be due to 
the peroxide in solution because of the interaction of the ultrasound and the added hydrogen peroxide being 
greater than the optimal peroxide concentration for atrazine degradation. 
8.2 ATRAZINE DEGRADATION 
Atrazine chemistry in solution during ozonation, sonication and with hydrogen peroxide is discussed in 
Section 8. 1 as a foundation for the investigation of the degradation of atrazine. Atrazine degradation 
during sonication is reported in Section 8.2.1, during ozonation in Section 8.2.2 and with hydrogen 
peroxide in Section 8.2.3. 
8.2.1 Ultrasound 
The ultrasonic degradation of atrazine was investigated by sonicating 5, 10 and 20 mg L· t atrazine solutions 
in the ultrasonic cell for 3 h at an acoustic power of 57 W. Attazine concentration was measured at time 
periods of 0; 45; 90; 135 and 180 min. Experiments were perfonned in triplicate. The ultrasonic 
degradation of atrazine is presented in Figure 8.7. 
25 
""20 t 
, 
, , I • 
S , 
i5 
~ IS 
-• • 5 mg/ l 
" c • 10 mg/ l • i ~ 10 • • • • .. 20 mg/ l 0 • 
•  ~ I I ~ 5 • I • 
0 
0 00 120 1,., lBD 210 
Time (min) 
Figure 8.7 : Atrazinc concentration in a nominal S, 10 and 20 mg L·t atrazine solution during 
sonication in the ultrasonic cell at an acoustic power of 57 W for J h (n z 3 per time period) 
Atrazine is degraded (as shown in Figure 8.7) during sonication in the ultrasonic cell . The maximum 
standard deviation in the measured atrazine concentrations during the sonication of nominal 5. IQ and 
20 mg L· 1 atrazine solutions are 0,44: 0,48 and 1,27 mg L'\ respectively. The rates of atrazine degradation 
in the different solutions were calculated from the regression of the data presented in Figure 8.7 using the 
linear regression mooel 
Cbapter 8 ATRAZlNI: EXPERIMENTS 8-17 
y=a+bx (8.3( 
where the coefficient b, the gradient of the regression line, represents the rate of atrazine degradation. The 
calculated rates of atrazine degradation in different atrazine solutions during sonication (and the standard 
errors of the calculated values) are recorded in Table 8.6. 
Table 8.6: Rate or atrazine degradation in a nominal S. 10 ud 20 mg Lol atrazine solution during 
!OIlication in tbe uttruonic cell at an acoustic power of 57 W over J b 
Atrazine concentration Rate of atrazine degradation 
(mg L ol) (mg L ol min-I) 
5 O,0082± 0,00 13 0,759 
10 0,0085 ± 0,0014 0,750 
20 0,0061 ± 0,0036 0,185 
The rate of atrazine degradation in solutions of initial concentration between 5 and 20 mg LO \ was not 
significantly affected by concentration. The absolute reduction in atrazine concentration in the solutions 
varied between 1,3 and 1,7 mg LO' . The slow degradation of atrazine during sonication, rep:>rted in 
Table 8.6_ is as discussed in Section 8.1.3, in which, similar rates of hydrogen peroxide formation in water, 
Figure 8.6(a), and an atrazine solution, Figure 8.6{b). indicated that low quantities of hydroxyl radicals 
were consumed during auazine oxidation. A concentration of 5 mg L'\ was used as the initial concentration 
for all funher atrazine experiments. The mean of the initial atrazine solutions rep:>rted in the atrazine 
investigation of Chapter 8 is 5,2 mg L- j , the standard deviation is 0.43 mg L- I (fable F.33 in Appendix F). 
The effect of acoustic power on alrazine degradation was not measured since the degradation rates listed in 
Table 8.6 were achieved during sonication at the maximum operating acoustic power of the ultrasonic horn 
(57 W). The rate of atrazine degradation is expected to decrease during sonication at a lower acoustic 
power since the formation of free radicals is dependent upon the intensity of cavitation. 
The effoct of a gas sparging through the solution during sonication was investigated as a means to enhance 
atrazine degradation. Atrazine is degraded by reaction with hydroX}'1 radicals and the presence of oX}'gcn 
during sonication, as discussed in Section 6.1 and Section 6.2. enhances hydrogen peroxide formation and 
free radical reactions. Conversely, the presence of nitrogen during sonication decreases hydrogen peroxide 
formation. Oxygen. as demonstrated in Figure 8.2. is required in the oxidation of atrazinc to yield 
dealkylated degradation products. Experiments were undcnaken in which oxygen and nitrogen were 
sparged at a flow rate of 6 mL sol through a 5 mg L-I atrazine solution during sonication for 3 h in the 
ultrasonic cell at an acoustic power of 57 W. Atrazine concentration was measured at time periods of 0; 45; 
Chapter 8 A'I'RA:UNE EXPERIMENTS 8-18 
90: 135 and 180 min. Experiments were performed in triplicate. The effect of oxygen and nitrogen 
sparging on atrazine degradation in the ultrasonic cell is shown in Figure 8.8. 
atrazine solution during sonication in tbe ultruonic cell at an acou5tic power of 57 W over 3 b 
(n "" 3 per time period) 
The maximum standard deviation in the atrazine concentrations measured during sonication with nitrogen 
and oxygen sparging are O,SI and 0,20 mg L", respectively. The rates of atrazine degradation during 
sonication with nitrogen and oxygen sparging were calculated from the regression of the data presented in 
Figure 8.8 using the linear regression model 
y=a+bx 18.41 
where the coefficient b. the gradient of the regression line. represents the rate of atrazine degradation. The 
calculated rates of atrazine degradation during sonication with nitrogen and oxygen sparging (and the 
standard errors of the calculated values) are recorded in Table 8.7. 
Table 8.7: Rate of atrazine degradation in as mg L'I atrazine solution during sonication witb 
nitrogen and oxygen sparging in tbe ultrasonic cell for 3 b at an acoustic power of 57 W 
Ga. 
no gas 
nitrogen 
oxygen 
Rate of atmine degradation 
(mg L'I min' l) 
0,0082 ± 0,0013 
0,00 14 ± 0,00 13 
0,0065 ± 0,0006 
0,759 
0,091 
0,913 
Cbapter 8 A TRAnN£ EXP£RIM£NTS 8-19 
The atrazine degradation rates reported in Table 8.7 indicate that oxygen is required for the degradation of 
atrazine. The mechanism for the formation of dcalkylated degradation products from the reaction of 
atrazine with hydroxyl radicals, shown in Figure 8.2, indicates the required presence of O"1'gen during 
atrazine oxidation. Atrazine degradation, shown in Figure 8.8, was significantly retarded in the absence of 
oxygen (during nitrogen sparging) in contrast to that in a sample saturated with air (no gas sparging) and 
with oxygen sparging. Atrazine degradation in the presence of oxygen, however, was not enhanced with 
increasing dissolved oxygen concentration from 10,7 mg L-1 (no gas sparging) to 39,6 rng L-' (oxygen 
sparging). The reaction mechanism of atrazine degradation thus requires the presence of oxygen whereas 
an increase in oxygen concentration is less significant. Enhanced degradation in the presence of oxygen 
confirms that atrazine degradation during sonication is due to the chemical effects of ultrasound (free 
radical reactions) and not due to the physical effects (shockwave formation) that wou1d also be evident 
during nitrogen sparging. 
8.2.2 Ozone 
Ozone is used in water treatment for disinfection, however, other benefits include the oxidation of organic 
poUutants such as atrazine. Ozone is reported in literature to significantly reduce the concentration of 
atrazine during ",-ster treatment (Adams and Randtke, 1992b; Beltnin et al., 1994a). Atrazine is degraded 
as discussed in Section 8.1 by direct reaction with ozone and by reaction with hydroxyl radicals formed 
during ozone decomposition. The relative amount of degradation due to either mechanism is determined by 
pH and reaction conditions. Atrazine degradation during ozonation in the ultrasonic celJ was investigated 
by measuring atrazine concentration over 3 h during ozonation at ozone production rates of 0,003 ; 0,014; 
0,030 and 0,047 mg 5.1. Atrazine concentration was measured at time periods ofO; 15; 30; 45; 90: 135 and 
180 min. Experiments were performed in duplicate. The degradation of atrazine during ozonation is 
presented in Figure 8.9. 
6 
! : ~-~"""-= .. ---
 • • ~ :It " 
~ 3 • 
• 
• o ---------------=~~~~· __ _" 
o 
"" 
150 leo 210 
Time (min) 
• 0 ,003 mg/ s 03 
• 0 ,014 mg/ s 03 
... 0 ,030 mg/ s 03 
:It 0 ,047 mg/ s 03 
Figure 8.9: Atrazine concentration in a 5 mg L-I atrazine llOlution during ownation over 3 h at 
owne production rates of 0,003; 0,014; 0,030 and 0,047 mg 5-1 (11 "" 2 per time period) 
Cbapter 8 ATRA1P'E EXPERIMEI'ITS 8-20 
The maximum standard deviation in the atrazine concentrations measured during ozonation at ozone 
production rates of 0,003; 0,014; 0,030 and 0,047 mg S·I are 0,18; 0,30; 0,22 and 0,25 mg L·I, respectively. 
The rates of atrazine degradation during the initial period of ozonation were calculated from the regression 
of the data presented in Figure 8.9 using the linear regression model 
y =o +bx (8.5J 
where the coefficient b, the gradient of the regression line, represents the rate of atrazine degradation. The 
calculated rates of atrazine degradation during the initial period of ozonation (and the standard errors of the 
ca1culated vaJues) are recorded in Table 8.8. 
Table 8 .8: Initial rate of atrazioe degradation in a ~ mg L -' atraziDc IOIution during ozonation at 
OZODe production rates of 0,003; 0,014; 0,030 and 0,047 mg S·l 
Ozone production rate Regression Rate of atrazine degradation 
(mg 5.1) time period (mg L·t min-I ) 
0,003 o to 180 min 0.0120 ± 0,0006 0.957 
0,014 o to 180 min 0,0143± 0,0006 0,979 
0.030 Ot045min 0,0506 ± 0,0048 0.949 
0,047 o to 30 min 0,0937 ± 0,0058 0,985 
The rate of atrazine degradation is shown in Figure 8.9 and Table 8.8 to increase with increasing ozone 
production rate. Atrazine degradation was linear over the 3 h e:\:periment (indicating a zero order process) 
during ozonation at ozone production rates of 0,003 and 0,014 mg S·I . The rates of atrazine degradation 
during ozonation at ozone production rates of 0.030 and 0.047 mg s· \ were initially linear but decreased 
with decreasing atrazine concentration (shown in Figure 8.9). The linear regression lines of the 0,030 and 
0.047 mg s'\ data in the plot of -In(ClCo) versus time presented in Figure 8.10 indicates that atrazine 
degradation during ozonation at ozone production rates of 0.030 and 0,047 mg 5'\ is a first order process. 
Atrazine degradation during ozonation thus changed from a zero order process independent of atrazine 
concentration at ozone production rates of 0,003 and 0,014 mg. S·1 to a first order process dependant on 
atrazine concentration at ozone production rates of 0.030 and 0.047 mg 5.1, An ozone production rate of 
0.014 mg S·I , as reported in Section 5.2. was used as the standard ozone production rate throughout the 
investigation. 
Chapter 8 A TRAZIl'I"I: EXPERlMJ.:NTS 
4 
3 
0-
 " <32 ~ 
Time (min) 
• 
• 
• 
• 0,003 ms/ s 
• 0,014 mg/ s 
.. 0,030 ms/ s 
*. 0,047 mg/ s 
8-21 
Figure 8.10: First order atrazine degradation in a 5 mg L '! atrazine solution during ownation 
over 3 b at owne production rates of 0,003; 0,014; 0,030 and 0,047 mg S·l 
Atrazine degradation during ozonation combined with sonication was investigated by measuring atrazine 
concentration in a 5 mg L'! atrazine solution over 3 h during ozonation at ozone production rates of 0,003: 
0.01 4; 0,030 and 0,047 mg S·l combined with sonication at an acoustic power of 57 W. Atrazine 
concentration was measured at time periods of 0; 15; 30; 45; 90; 135 and ISO min. Ex-periments were 
performed in triplicate. Atrazine degradation during ozonation combined with sonication is presented in 
Figure 8. 11 . 
o ~---------'~~~=-__ ~ __ ~ 
o 00 90 '20 
''''' 
'00 2'0 
Time (min) 
• uls;O,003 mg/ s 03 
• ul s;O,014 mg/ s 03 
.. uls;O,030 ms/ s 03 
I< uls;O,047 mg/ s 03 
Figure 8,11 : Atrazine concentration in a 5 mg L-1 atrazine solulion during ownation over 3 h 
at «LOne production rates of 0,003; 0,014; 0,030 and 0,047 mg SI combined with sonication at an 
acoustic power of 57 W (n "" 3 per time period) 
Cbapter 8 A1"RAZIl'\"£ EXPERIMENTS 8-22 
The maximum standard deviation in the atrazine concentrations measured during sonication combined with 
ozonation at ozone production rates of 0,003; 0,014; 0,030 and 0,047 mg S·I are 0,25; 0,21 ; 0, 11 and 
0,06 mg L·I . respectively. The rates of atrazine degradation during the initial period of ozonation were 
c:alcu1ated from the regression of the data presented in Figure 8.11 wing the linear regression model 
y=a+bx [8.6J 
where the coefficient b, the gradient of the regression line, represents the rate of atrazine degradation. The 
ca1cu1ated rates of attazine degradation during the initial period of ozonation combined with sonication 
(and the standard errors of the calcu1ated values) are recorded. in Table 8.9. 
Table 8.9: Initial rate of atrazine degradation in a S mg L,I atrazine solution during ozonatioD 
over 3 b at ozone production rates 0( 0,003; 0,014; 0,030 and 0,647 mg "I combined with sonication 
at an acoustic power of 57 W 
Ozone production rate Regression Rate of atrazme degradation 
(mg S,I) time period (mg L,I min,l) 
0,003 o to 45 miD 0,0287 ± 0,0045 0,871 
0,014 Ot045 min 0,0351 ± 0,0027 0,946 
0,030 o to 45 min 0,0644 ± 0,0045 0.953 
0,047 o to 30 min 0,0918 ± 0,0075 0,955 
The rate of atrazine degradation during ozonation combined with sonication is shown in Figure 8.11 and 
Table 8.9 to increase with increasing ozone production rate. Atrazine degradation is omy linear over the 
initial period and decreases, as shown in Figure 8. 11 . with decreasing atrazine concentration. The linear 
regression lines presented in Figure 8.12. a plot of -In(C/Cc) versus time. indicates that atrazine degradation 
during ozonation combined with sonication is a first order process for any ozone production rate. 
Cbapter 8 ATR.A.2'lNE ExpfJUMEl"I"TS 8-23 
6 '1 -----------------------------, 
" 
• 
• uls;O,003 mg/s 
• • uls ;O,014 mg/ s 
• uls;O,030 mg/s 
• UI5;O,047 mg/ s 
2 r 
• 
• 
o 
o ID 120 150 1ID 210 
Time (m in) 
Figure 8.12: First order atrazine degradation in a 5 mg L' l atraziDe solution during ozonation 
over 3 b at ozone production rates of 0,003; 0,014; 0,030 aod 0,047 mg 5,1 combined witb 
sonication at an acoustic power of 57 W 
The initial rate of atrazine degradation, as shown in Table 8.8 and Table 8.9, is enhanced when ozonation 
(at ozone production rates of 0,003; 0,014 and 0,030 mg s" ) was combined with sonication. The atrazine 
degradation rates during ozonation (listed in Table 8.8) and during ozonation combined with sonication 
(listed in Table 8.9) are compared in Figure 8.13. 
0,12 
,g 0,06 -
 • 1< • 
• 03 
• 03/u15 
~ 0.04 -
 ~ 
o 
o 0,1 0,2 0,3 0,4 0,5 0,6 
03 production rate (ms/ s) 
Figure 8.13: Comparison of atrazine degradation rates in a 5 mg L' t atrazine solution during 
ozonation and ozonation combined witb sonication 
The atrazine degradation rates presented in Figure 8.13 indicate the enhanced effect on atrazine degradation 
by combining ozone with ultrasound.. The effect, however, decreases with increasing rate of ozone 
production. The effect of ultrasoWld during ozonation at the different ozone production rates is presented in 
Chapter 8 ATRAZlNE EXPERIMENTS 8-24 
Figure E.I to Figure E.4 in Appendix E. The enhanced initial degradation of atrazine due 10 ultrasound 
decreases with increasing rate of ozone production until no significant enhancement takes place during 
ozonation al an ozone production rate of 0,047 rng S'1 (Figure E.4). Ultrasound enhances the degradation of 
atrazine because of the increased decomposition of ozone and, hence, the increased formation of hydroxyl 
radicals (discussed in Section 8.1.2). Hydroxyl radicals are also fonned during ozonation (in the absence of 
ultrasound). Thus the formation of hydroxyl radicals initiated. by the ultrasound becomes insignificant in 
comparison with that generated during ozone decomposition (in the absence of ultrasound) at high ozone 
production rates. The production rates of ozone applied during ozOnatiOD in water treatment are low were 
the impact of ultrasound is highest, thus, a significant improvement in the degradation of organic 
compounds can be obtained by combining ozonation with sonication. 
8.2.3 Hydrogen peroxide 
The combination of hydrogen peroxide with advanced oxidation prooes.ses such as ultrasound or ozone 
enhances the oxidation of organic pollutants due to the generation of free radicals. Hydrogen peroxide, as 
shown in Scheme 3.9 in Section 3.3.1, is decomposed by ultrasound or ozone into perhydroxyJ and hydroxyl 
radicals. The effect of hydrogen peroxide on auazine degradation in the ultrasonic cell was investigated 
over 3 h with the addition of 40 mg L'\ hydrogen peroxide to a 5 rng L' \ atrazine solution during sonication 
at an acoustic power of 57 W and ozonation at an ozone production rate ofO,014 rng S'l . The selection of 
the hydrogen peroxide concentration is detailed in Section E.2.3 in Appendix E. Atrazine concentration 
was measured at time periods of 0 ; 15; 30; 45; 90; 135 and ISO min. Experiments were perfonned in 
triplicate. The effect of hydrogen peroxide on atrazine degradation is shown in Figure 8.14. 
6 -
 • 
• 2 . ~ -
 a> 1 -
 0 
0 60 120 
Time (min) 
• 
• 
160 
• 
100 210 
• H202 
• uls/H202 
.. 0 31H202 
.. uls/ 03/ H202 
Figure 8,14: Atrazioe cODcentration in a 5 mg L' t atrazine solution with 40 mg L' \ bydrogen 
peroxide. hydrogen peroxide combined with sonication (57 W). hydrogen peroxide combined 
with ozooation (0,014 mg 5,1) and hydrogen peroxide combined with sonication and ozonation 
Cbapter 8 A TRA2'.Jl\t: EXPERIMEl'I7S 8-25 
The maximum standard deviation in atrazine concentrations measured in a hydrogen peroxide solution, 
during sonication of the peroxide solution, during ozonation of the peroxide solution and during sonication 
combined with ozonation are 0,23; 0,19; 0,21 and 0,24 mg L- t , respectively. The initial rates ofatrazine 
degradation in a hydrogen peroxide solution during ozonation, sonication and the combination of ozonation 
and sonication were calculated from the regression of the data presented in Figure 8_14 using the linear 
regression m<Xie1 
y=a+bx (8,7( 
where the coefficient b. the gradient of the regression line, represents the rate of atrazine degradation_ The 
calculated rates of atrazine degradation in a hydrogen peroxide solution during the initial period of 
olonation, sonication or ozonation CXlmbined with sonication (and the standard errors of the calculated 
values) are recorded in Table 8.10_ 
Table 8.10: Initial rate of atraziDe degradation in a 5 mg L- t atrazine solution with 40 mg L-1 
hydrogen peroxide, hydrogen peroxide combined witb IODication (57 W), hydrogen peroxide 
combined with ozoDatioD (0,14 mg S1) and bydrogen peroxide combined with sonicatioD and 
ozonation 
Regression Rate of atrazine degradation 
time period (mg L'l miD'l) 
H,o, o to 180 min 0,0021 ± 0,0007 0,422 
u1s1H,o, o to 180 min 0,0120 ± 0.0005 0.984 
o,tHz<), 0(030 min 0.0568 ± 0,0058 0.932 
ul"o,tH,o, o to 30 min 0.0370 ± 0.0056 0.860 
A(razine degradation is relatively slow, as shown in Figure 8.14. in the presence of hydrogen peroxide alone 
whereas (he degradation was significantly enhanced during the combination of hydrogen peroxide with 
ozone or ultrasound The reaction mechanisms of atrazine degradation during sonication or ozonation 
combined with hydrogen peroxide are discussed in Section 8.1. The degradation rate of atrazine during 
sonication increased from 0.0082 mg L'! min-! (fable 8.6) to 0,0120 mg L-! min-l with the addition of 
hydrogen peroxide (Table 8.10). Similarly, the degradation rate of atrazine during ozonation increased 
from 0.0 143 mg L-l min-I (Table 8.8) to 0,0568 mg L' ! min-\ with hydrogen peroxide (Table 8.10). The 
highest rate of atrazine degradation was measured when ozonation was combined with hydrogen peroxide. 
This was confinned by the mass baJancc experiments (Section 8.1A) in which 'the total atr3Zine degradation 
over 45 min was the greatest (42 %) during ozonation combined with hydrogen peroxide (Table 8.4). 
Chapter 8 A TRAZINE EXPERlME!'ro'TS 8-26 
Atrazine degradation in a 5 mg LO \ atrazine solution during sonication, ozonation and sonication combined 
with ozonation, all with and without hydrogen peroxide, are compared in Figure E06 to Figure E.8 in 
Appendix E. Hydrogen peroxide is shown to enhance atrazine degradation during sonication and during 
ozonation but not to affect degradation during the combination of sonication with ozonation. The reduced 
atrazine degradation in the presence of high bydrogen peroxide concentrations, as discussed in 
Section 8.1.4, is due to the change in kinetic regime of ozone absorption and the scavenging of hydroxyl 
radicals by the bydrogen peroxide thus hindering the reaction of atrazine and hydroxyl radicals. The 
comparisons of atrazine degradation presented in Figure E.6 to Figure E.8 indicate that ultrasound should 
be combined in water treatment with either ozone or hydrogen peroxide but not both together since the 
peroxide concentration in solution may be greater than the optimal concentration required for the oxidation 
of organic pollutants. 
8.3 PRODUCT IDENTIFICATION 
Atrazine concentration in the degradation experiments discussed in Section 8.2 was measured using the 
HPLC method described in Section 5.4.4. The HPLC chromatogram of an unreacted 5 mg L'\ atrazine 
solution is presented in Figure E.9 in Appendix E. The retention time of atrazine on the chromatograms for 
the HPLC analysis system, as shown in Figure E.9, is approximately 6,5 min, although some variation 
occurred on a daily basis. The blimp at a retention time of 8 min that was present on all chromatograms 
was an impurity in the atrazine feedstock (97 %). The same peak was also present when new atrazine 
feed.stock was obtained from Sanachem. Atrazine degradation products were detected with the HPLC 
system. the product peaks on the chromatograms increased in size as the atrazine peak decreased. Typically 
the degradation products appeared as a double peak ahead of the atrazine peak, at retention times of 
approximately 5,2 and 5,4 min. Prolonged oxidation of atrazine resulted in the product peaks increasing in 
size beyond that of the decreasing atrazine peak. 
The size of the product peaks detected during the ozonation of atrazine increased with time over the 3 h 
experiment and with increasing ozone production rate (0.003 to 0,047 mg s·\). Formation of degradation 
products occurred sooner during ozonation at higher ozone production rates. Degradation products were 
first evident at 90 min during ozonation at an ozone production rate of 0,003 mg s·\ whereas they were 
present from 15 min onwards for an ozone production rate of 0,047 mg 5.1• The decreasing atrazine peak 
and increasing product peaks on the HPLC chromatograms during the ozonation of atrazine at an ozone 
production rate of 0,047 mg s'\ are presented in Figure 8.15. 
Cbapter 8 ATRAZlNt. EXPERL\UNTS 8-27 
:1 
_. 
, j 
(a) 0 min (b) 30 min 
- 1 
-
 ,-
 " 
(c) 90 min (d) 180 nUn 
Figure 8.15: BPLC cbromatograms for tbe ozonation of a 5 mg L-t atnzine solution at an 
ozone production rate of 0,047 mg L-t 
" 
The oxidation of atrazine. as discussed in Section 8.1 and Section 8.2, was enhanced during ozonation 
(0.047 mg S·I) combined with sonication (57 W). The HPLC chromatograms over the 3 h experiments 
showed the enhanced formation of degradation products during ozonarion combined with sonication versus 
during ozonation alone. The decreasing atrazine peak and increasing product peaks on the HPLC 
chromatograms during ozonation combined with sonication are presented in Figure 8.16. 
i\ - :\ jI 
i \ - VI I \ I I--J I 
'-1 ) , ~ \~ \; '1 • 
" " 
(a) o miD (b) 30 miD 
Chapter 8 ATRAZDfE EXPERIMENTS 8-28 
-
 : 1 
- j 
: 1 
-
 :j I'i\ , rJ\ ,~ , I I '\JI '\ , ~ If - / , . 
" 
" 
(c) 90 miD (d) 180 min 
Figure 8.16: BPLC chromatograms for tbe ozonation of a 5 mg L ·1 atrazine solution at an 
ozone production rate of 0~047 mg L-1 combined with sonication at an acoustic power of 57 W 
, 
The relatively slow degradation of atrazine in the presence of hydrogen peroxide alone (Figure 8_14) was 
confirmed since no product peaks were evident on the HPLC chromatograms. The degradation of atrazine, 
discussed in Section 8. 14 and Section 8.2.3, during ozonation combined with hydrogen peroxide was 
greater than during ozonation combined with both sonication a nd hydrogen peroxide. Thus a greater 
quantity of degradation products, as shown in Figure 8. 17, were formed during ozonation combined with 
hydrogen peroxide than during ozonation combined with ooth sonication and hydrogen peroxide. 
:1 , - \ " \ IV! i \ 
' \ (\ 
-\ -I I I \J' ,0 '- I l r , . 
'-
 j 
" " 
(a) (YJ".-OneIH~O~ (b) ozone/uJs/B~02 
Figure 8.17 : HPLC cbromatograms at 180 min for a 5 mg L- I atrazine solution during 
O'LOnation (0,014 mg L-1) combined with hydrogen peroxide (40 mg L-I) and ozonation combined 
with both sonication (57 W) and hydrogen peroxide 
The double proouct peaks on the different chromatograms indicate that more than one product is being 
formed during the oxidation of atrazine. The same pattern of peaks indicates that the same products are 
being formed during ozonation, ozonation combined with sonication. ozonation combined with hydrogen 
peroxide and ozonation combined with sonication and hydrogen peroxide. The same mechanism of atrazine 
degradation is thus occurring. Minor products formed from further degradation of the primary degradation 
products or from other degradation mechanisms may be hidden on the cmomalograms under the dominant 
Chapter 8 ATRAZlNE EXPERIMl:NTS 8-29 
product peaks because of similar retention times. The overlap of the double product peaks on the 
chromatograms indicated that the products were co-eluting and could Dot be collected separately for 
identification. Gas chromatography- mass spectrornetry (GC-MS) was thus used to identify degradation 
products. Samples at the end of an experiment (at 180 miD) during ozonation (0,047 mg 5.1). sonication 
(57 W) and Q""Lonation combined with sonication of a 5 mg L·I atrazine solution were analysed. Product 
identification was the primary objective of the GC-MS analysis thus once-off samples for each experiment 
were analysed. Samples throughout the ex-periments would have to be analysed for the investigation of the 
trends in product formation and further degradation. 
GC-MS analysis was performed at the csm and Rand Afrikaans University (RAU). Analytical procedures 
are detailed in Section E.3. A 5 mg L'\ atrazine solution was analysed to identify degradation products after 
3 h of sonication, Q""l.onation and sonication combined with ozonation in the ultrasonic cell . An unreacted 
atrazine sample (control) was analysed using the same method as that used to analyse the reaction samples. 
The mass spectrum of atrazine is presented in Figure 8. 18. 
'" 
58 
215 
•• 
1Nl 
" 
173 
1ZZ 
I 6Z '" m I 115 15. 
I1 ~ " .1., '" L \ ~4 11 .;1;,1  1.,,1, 11, I I" "I'~,I I J , ,\~ ~ , , , , 
., , ' I ' , , 
" " 
.. 
" " " 
HlO 11e 12e Ile 14e 1se u;.e 171] 18e 191] 200 2 11] 221] 
Figu re 8.18: Mass spect rum of a 5 mg L-1 atrazine solution 
The molecular ion peak [M"'J at a m/e ratio of215 for atrazine is evident in the mass spectrum presented in 
Figure 8 . 18. Chlorine-containing compounds reflect a [M + 2r peak approximately a third of the intensity 
of the fMl peak because the natural abundance of the 37Cl isotope is approximately 32,5 % of the 35C1 
isotope (pavia et al ., 1979). The [M + 2f peak at m/e 217 for atrazine. a third of the size of the peak at 
215. is evident in Figure 8 . 18. Structures of some of the molecular fragments generated during the 
ionisation of atrazine. as indicaled on the mass spectrum in Figure 8 . 18. are presented in Figure 8.19. 
Chapter 8 A TRAZlNE EXPERIMENTS 8-30 
mlt! 200 
1 j 
mlt! 58 
/ mlt! 172 j 
mlt! 138 
Figure 8.19: Molecular fragments generated during the ionisation of atrazine 
The molecular fragments of atrazine presented in Figure 8. 19 indicate that the cleavage of the alkyl side 
chains occurs more readily during ionisation than that of the chlorine substituent, cleavage of the chlorine 
atom only takes place after the loss of part of the alkyl side chain groups. The peak al mle 200 (202, 3'CI) 
in Figure 8.18 indicales the loss of a methyl group ("CH)) from either the ethylamino or isopropylamino 
side chains. Cleavage of Ihe methyl group from the isopropylamino side chain produces a more stable 
radical because of the positive inductive effect of the other methyl group (Emslie, 200 I). The peak at 
mle 172 (174, HC!) is due to the cleavage, as shown in Figure 8.19, of the isopropyl group (OCH(CH)z) 
from the isopropylamino side chain. The ethyl group (OCHzCH3) in the elhyJamino side chain is nol directly 
cleaved since a significant peak is not present at mle 186 (188, 37CI) . Hapeman~Somich and co-workers 
also observed the lack of direct cleavage of the ethyl group (Hapeman-Somich et al .. 1992). The peak at 
mle 58 represents the cleavage of the complete isopropylantino side chain. The peak at mle 68 is due to the 
fragment rearrangement shown in Scheme 8. I that was proposed by Hapeman~Somich and co-workers 
(Hapeman~Somich et al., 1992). 
Scbeme 8.1 
--.. + 
• N=:C-N= C= NHl 
mI, 68 
8-31 
The GC·MS speclra presented in Figure E.lO of a 5 mg L·1 atrazine solution after sonication. ozonation and 
sonication combined with ozonation indicate that a range of atrazine degradation products are formed. 
Deethylatrazine. deisopropylatrazine and bydroxyatra2.ine are shown in Figure 4.3 in Section 4.2.2.1 to be 
the primmy metabolites of atrazine. Dcethylatrazine was identified as a degradation product in the analysis 
performed at the CSIR of the sonication and ozonation auazine samples by comparison of the [M + Hr 
peak at mle 188 with the library of compounds (Naicker, 1999). DcethylalTaZine was also identified in the 
ozonation sample anruysed at RAU, shown in Figure E. IO(b). The mass spectrum of deethylatrazine is 
presented in Figure g.20. 
'" 
'" 
Figure 8.20: Mass spectrum of deeth}'latrazine fonned during the sonication and ozonation of 
as mg L'\ atrazine solution for 3 b 
The parent peak fMr of dccthylalrazine is evident in Figure 8.20 at mle 187 (189, J1CI). The spectrum 
presented in Figure 8.20 is similar to that identified by Legube and co-workers as deethylatrazine \\ith 
peaks at mle 43, 58, 69. 83. 85, 104, 145, 172, 175. 187 and 189 (Legubc et al ., 1987). The base peak at 
mle 172 (17"', 3'CI) indicates the loss of a mcthyl group (·CH3) from the isopropylamino sidc chain. Also 
shown in Figure 8.20 is the lack or direct cleavage of chlorine from the parent compound to generate a peak 
at m/ e 152. StruCtures of the molecular fragments generated during the ionisation of deethylatrazine. as 
indlcatcd on the mass spectrum in Figure 8.20, are presented in Figure 8.21. 
Cbapter 8 ATRAZlNE ExP£RJl'tIEl'ITS 
mI~ 187 
mI.94 
mle 145 
.~ 
- _. ~ 
H 
mI~ 58 
• --~. N=C-N= C= NH2 
mI~68 
Figure 8.21 : Molecular fragments generated during tbe ionisation of deetb~' lat razine 
8·32 
Hydroxy3trazine was formed in a 5 mg L·t atrazine solution during sonication, O'"lonation and sonication 
combined with ozonation for 3 h. The mass spectrum ofhydroxyatrazine is presented in Figure 8.22 . 
l06z 
I"' 
m 
I. 
, 
'" 
1<, 
13$ 
'" '" 
265 
I 223 I 
I" 2" 
'" '" Figure 8.22: Mass spectrum of bydroryatrazine formed during tbe sonication, o~.oDatioD and 
sonication combined with ozonatioD of a 5 mg L·t atrazine solution for 3 h 
Cbapter 8 ATRAlD'iE EXPERIMENTS 8-33 
Hydroxyauazine was identified by its principle fragmentation ion at mle 149 as the {M]';' peak was not 
evident in the mass spectrum presented. in Figure 8.22. Hydroxyatrazine has two aliphatic amine side 
chains that undergo ~-cleavage as shown in Scheme 8.2 (Pavia et al., 1979). 
[R-~<J; --" R" + 
Scheme 8.2 
'\. +/ 
C=N 
/ '\. 
The base peak at mle 149 in Figure 8.22 is proposed to be due to either of the comJX>unds presented in 
Figure 8.23 that are fonned from the ~-cleavage of the hydroxyalIaZine side chains (Schwikkard. 2000). 
The OH group of bydroxyatrazine is readily lost during ionisation and the structures presented in 
Figure 8.22 are based on the prior loss of the OH group. 
0 ' 
Figure 8.23: Compounds fonned during tbe ionisation of bydroxyatrazine due to the p~leal'age of 
the alipbatic amine side chains 
Minor. unidentified products were also fonned as shown in Figure E. IO, in a 5 mg L'\ atrazine solution 
during sortication. ozonation and sonication combined with ozonation for 3 h. The mass spectrum 
presented in Figure 8.24 of the product with a retention time of approximately 12 min in Figure E.IO(b) and 
(c) is tentatively identified as deethyldeisopropylatrazine due to the characteristic fM + 2HJ2. peak at 
mle 147 (149, 37CI) although the peak at mle 73 is unexplained. The peak at mle 73 is not due to the direct 
cleavage of a fragment from the parent comJX>und but JX>ssibly the rearrangement of such a fragment. 
Nclieu and q,.workers found that dectbyldeisopropylauazine was fonned in ozonation experiments longer 
than 60 min though the concentration was always less than 2 % of the initial alIaZine (NcHieu et al .. 1996). 
The small quantities of products formed in the investigation in the ultrasonic cell did not allow for nuclear 
magnetic resonance spcctroscopy (NMR) of products to facilitate identification. 
Cbapter 8 ATRA.ZlNE EXPERJM.ENTS 
l08~ 
" 
1 
I 
J 
9? 
J 
119 133 
HIO 12D 
'" 
8-3-1 
163 179 19 1 
287 22 1 
'" 
lee 2eD 226 2ie 266 
Figure 8.14 Mass spectrum of deetbyldeisopropylatrazine formed during tbe sonication . 
ozonation and sonication combined "ft;tb ozonation of as mg L· t atrazioe solution for 3 b 
The formation of dealkyl:lled prooucts (deethylatrazine and plssibly deethyldeisopropylatrazinc) is due to 
both the reaction of atrazinc with hydroxyl radicals (Figure 8.2 in Section 8. 1.1) and the direct rea<:tion with 
ozone (Figure 8.5 in Section 8.1.2). Hydrogen peroxide is produced as a byproouct in the formation of 
deaLk)'lated degradation products. Enhanced hydrogen peroxide formation was measured in an arrazine 
solution (in comparisco with that measured in water) during 02onation and 02onation combined with 
sonication (shown in Figure 8.6 in Section 8.1.3). Hydrolysed degradation products (bydroxyalIazine) are 
only formed due to the reaction of atrazine \vim hydro~'Y1 radicals (Figure 8.3 in Section 8. 1. I). 
Both dcalkylated and hydrolysed degradation products (shown in Figure E. lO) were detected during the 
ozonation and sonication of atrazine. The degradation mechanism of atrazine during sonication is due to 
the reaction wi th hydroxyl radicals. The degradation mechanism during ozonation is due to both the direct 
reaction ,..,jlh ozone and the reaction with b)'dro~1'1 radicals. although. as concluded in Section 8. L the 
reaction with hydroxyl radicals is the dominant degradation mechanism. 
8.4 CONCLUDING R£MARKS 
Atrazi ne is nOl degraded upon standing even with O~'Ygen sparging and only slightly in the presence of 
hydrogen peroxide whereas degradation occurs during sonication and ozo03tion. Atrazinc accumulates in 
the gas/liquid interfacial region during sonication because of ilS non-volalHity and hydrophobicit)·. 
Ultrasound enhances the degradation achieved during atrazine oxidation with either ozone or hydrogen 
peroxide. The influence of uJtrasound is limited when ozon3lion is also combined with hydrogen peroxide 
or when the concenlr.:llions of ozone and hydrogen peroxide arc VCf!-' high. such concentrations though are 
Chapter 8 A TRAZlNE EXPERIMENTS S-35 
nOI appropriate for industria] application. Ultrasound initiates the decomposition of ozone and hydrogen 
peroxide to enhance the degree of free radical reactions occurring in solution. 
Atrazine degradation during ozonation can be due 10 the direct reaction with ozone or to the reaction with 
free radicals in solution. Reaction with free radicals was found to be the dominant degradation mechanism. 
The greatest atrazine degradation was measured during ozonation combined with hydrogen peroxide where 
dissolved ozone and free radicals were available in solution. Enhanced hydrogen peroxide formation during 
ozonation or sonication of an alrazine solution, in comparison with that in water, is due to the peroxide 
generated as a byproduct during the fonnation of dealkylated degradation products. 
Atrazine degradation products were identified used gas chromatography and mass spectromet.ry. The 
degradation products fonned during the ozonation. sonication and ozonation combined with sonication of 
atrazine were deethylatrazine. hydroxyatrazine and deethyldeisopropylatrazine (tentatively identified). 
Dcalkylated degradation products are fonned due to both the direct reaction with ozone and the reaction 
with free radicals whereas hydrolysed products are only formed from the reaction with free radicals. The 
dominant degradation mechanism of atrazine was via the reaction with free radicals. Other minor 
degradation products were fonned but could not be identified. 
9 
CONCLUDING REMARKS 
The deteriorating water quality in South Africa and changing legislation requiring the implementation of 
waste minimisation and pollution prevention strategies has highlighted the need for the investigation of new 
effiucnt treatment technologies such as advanced oxidation processes. This study investigated the potential 
application of ultrasound in water treatment for the oxidation of organic pollutants using atrazine as a 
model compound. The objectives of the investigation are listed below with concluding remarks addressing 
each of the objectives. 
a) Design and construct an ultrasonic cell to be used in tbe evaluation of ultrasound to degrade 
organic pollutants in water treatment 
A 500 mL batch ultrasonic laboratory reactor, with an ultrasonic horn as an energy source, was designed 
and constructed The reactor operates at atmospheric pressure and allows for the gas sparging of solutions 
during sonication. Different sample ports allow for both gas and liquid samples to be drawn, liquid samples 
can be taken continuously throughout an e>.:periment if the volume required for analysis is small. The 
ultrasonic cell is thus a versatile system that can be used in screening experiments to investigate whether 
ultrasound enhances a panicular reaction or be used to degrade a particular organic compound 
b) Use such equipment to im'estigate the process mechanism (the sequence of sub-processes whose 
o\'erall result produces tbe obsen'ed effect) occurring during sonication. 
Pollutant oxidation during sonication is due to free radical reactions occurring ",ithin the collapsing gas 
cavities. at the gas liquid interface and in the bulk solution, Hydrophobicity and volatility detennine in 
which region a compound participates in the radical reactions. Hydrogen peroxide is formed during the 
ultrasonic-initiated free radical reactions. The measurement of hydrogen peroxide concentration was used 
as a measure of the degree of radical reactions occuning during sonication and is thus a tool that can be 
used to investigate under which process conditions reactor performance is enhanced. 
The rate of hydrogen peroxide formation during sonication increases with increasing dissolved O>"1'gen 
concentration due to the additional radical reactions initiated in solution. The rate of peroxide forotalion. 
however, decreases with increasing peroxide concentration as the equilibrium between peroxide formation 
and degradation is approached, Hydrogen peroxide concentration thus decreases during sonication in a 
Chapter 9 CONCLUDING Ib:l\iARKS 9-2 
solution with an initial concentration greater than the equilibrium value. Hydrogen peroxide degradation is 
slow after sonication has been ended, thus oxidation of organic comp:>unds can occur both during and after 
sonication. A correlation (accounting for approximately 97 % of the variation in the data) was developed 
for the prediction of hydrogen peroxide concentration as a function of dissolved oxygen concentration, 
acoustic )X)wer and time. The correlation equation is 
[H 202] = --(), 5420[time ]0.699$ + 0, 5516[time 1O,69I4[ uls]-O,OO1761 [02]°.00$917 [9.1) 
where [H..o}l is measured in mg L- 1, luls) in W, [02] in mg L- 1 and (lime] in min. This correlation can be 
used to guide the choice of process conditions and mode of operation of an ultrasound process so as to 
maximise hydrogen peroxide formation and hence free radical reactions. 
c) Investigate tbe equipment operation and implications for !lCale-up. 
This study (using the measurement of hydrogen peroxide concentration as an indication of the free radical 
reactions occurring during sonication) has resulted in the foUowing guidelines being proposed for the 
scale~up of ultrasonic processes, these are: 
• The continual sparging with a gas (oxygen or air) or added hydrogen peroxide should be used 
during sonication to enhance free radical reactions. 
• Oxygen sparging and a high acoustic )X)wer input should be used in processes with a short retention 
time, conversely, air sparging and a lower acoustic energy source should be used in processes with a 
long retention time. 
• A flow loop system should be considered to maximise oxidation of organic compounds both within 
and beyond the sonicated zone, gas sparging should only occur 'within the sonicated zone 
d) Compare tbe perfonnance of ultruound, an emerging technol~' in ,uter treatment, with that of 
ozone, an already established adnnced oxidation process. 
Ozone chemistry was investigated as a foundation for the comparison of ultrasound with ozone, Oxidation 
of organic comp:>unds during ozonation takes place due to either direct reaction with ozone or reaction with 
free radicals fonned from the decomp:>sition of ozone. Ozone decom)X)sition (and hence free radical 
reactions) is enhanced with the combination of ozone with ulU'aSOund or hydrogen peroxide. A 40 mg L-1 
hydrogen peroxide concentration totally depleted the dissolved ozone concentration in solution whereas 
ultrasound only partially reduced the ozone concentration. The partial reduction obtained with sonication 
in the uluasonic cell is equivalent to that obtained with between 15 and 20 mg L+] of hydrogen peroxide. 
Enhanced ozone decomposition occurs due to the conditions (high temperatures and pressures) as well as 
the free radical reactions occurring within the collapsing cavitation bubbles and at the gas~liquid interface, 
Chapter 9 CONCLUDlNG REMARKS 9-3 
Ozone is a powerful but selective oxidant and the performance of ozonation during water treatment for the 
oxidation of organic compounds is thus enhanced with the combination with either hydrogen peroxide or 
ultrasound. Ozonation combined with hydrogen peroxide allows for oxidation primarily due to reaction 
with hydroxyl radicals whereas ozo03tion combined with sonication allows for oxidation of compounds that 
preferentially react with either ozone or hydro,""),l radicals. 
e) Use atrazine as a model organic compound and investigate degradation rates during sonication 
and ozonation. 
Atrazine was used as a model compound in the evaluation of ultrasound to degrade organic compounds in 
water treatment as it is difficult to degrade and atrazine chentisuy is well studied.. Atrazine is used 
extensively in South Africa and is highly persistent in the environment. Atrazine is also relatively safe to 
handle and is readily available. Alrazine was degraded during sonication and during ozonation, 
degradation was enhanced with the addition of hydrogen peroxide. Atrazine degradation was greatest 
during ozo03tion combined with hydrogen peroxide and decreased in the order for O2o03tion combined with 
both sonication and hydrogen peroxide > O2o03tion combined with sonication> ozOO3tion > sonication and 
hydrogen peroxide. The enhancing effect of ultrasound on atrazine degradation during ozo03lion was 
greatest at low ozone concentrations. Atrazine degradation during sonication and ozonation is via reaction 
with free radicals. Degradation during ozonation can also be due to direct reaction with ozone, though, this 
reaction is the less preferred mechanism. 
Ultrasound is a viable technology to be used to enhance the decomposition of oxidants such as ozone or 
hydrogen peroxide thus generating highly reactive free radicals. Ultrasound is most effectively applied in 
water treatment, to degrade organic compounds, as a pretreatment stage in combination with other 
technologies and not as a stand-alone process. 
f) Identify degradation products obtained during the sonication and ozonation of al razine. 
Gas chromatography and mass spectrometry were used to identify atrazine degradation products produced 
during sonication, ozonation and sonication combined with ozonation. Deethylatrazine, hydro: .. :yatrazine 
and deethyldeisopropylatrazine (tentatively identified) were all formed during sonication. ozonation and 
sonication combined with ozonation. Dealkylated degradation products are formed due to the reaction of 
atrazine with both ozone and hydroxyl radicals whereas hydrolised degradation products are only fonned 
due to the reaction of atrazine with hydro,",),1 radicals. 
Recommendations 
This study has focused on the investigation of the fundamental processes occurring during sonication and 
ozonation. Atrazine chemistry was also investigated prior to degradation experiments. The fundamental 
underslanding of the system can be used as a foundation for more application-focused investigations. The 
Cbapter 9 CONCLUDING REMARKS 9-4 
following work is recommended to be performed to continue with the investigation of advanced oxidation 
processes in water treaunent. 
• Process and kinetic modelling of the atrazine system to be performed, as a case study, to generate 
information for the modelling and design of large-scale ultrasonic reactors. 
• The ultrasottic cell to be modified to allow for continuous as well as batch operation, thus enabling 
a greater versatili ty in process operation and scale-up studies. 
• 
• 
• 
• 
The degradation of other recalcitrant organic compounds in the ultrasonic cell to be investigated to 
establish the class of pollutants that are most suited for ultrasonic degradation. 
The use of ultrasound as a pretreaunent technology for biodegradation to be investigated, especially 
for the treatment of concentrated and toxic emuents that will be generated as waste minimisation 
technologies are applied in industry. 
Different reactor configurations to be investigated to determine the most effective way of combining 
ultrasound with ozone. The investigation should not only focus on the simultaneous application of 
ultrasound and ozone but also on whether any benefit is obtained if sonication is applied before or 
after ozonation. 
The power efficiency of ultrasound should be investigated and compared with that of ozone and 
hydrogen peroxide. 
REFERENCES 
Adams. C.D. and Rand1ke. SJ. ( l 992a) OzonatioD byproducts of atrazine in 5yntbetic and natural 
waters, Environmenlal Science and Technology. 26(11 ),2218-2227. 
Adams. C.O. and Randtke, SJ. (l992b) Removal of atrazine from drinking water by ozonatioD, Journal 
o/the American Waler Works Association, 84(9), 91 -102. 
Adams, CO .. Randtke, SJ., Thurman. E.M. and Hulsey, R.A (1990) Occurrence and treatment of 
atrazine and its degradation products in drinking water, Proceedings o/the American Water 
Works Association 's Annual Conference, Cincinnati, Ohio. USA. 1. 871-894. 
Agbekodo. K.M., Legube, B. and Dard, S. (1996) Atrazine and simazine removal mechanisms by 
nanofiltration: influence of natural organic matter concentration, Water Research, 30(1 1). 
2535-2542. 
Abmad, W.Y .W. and Lamas, M. (1996) The low-temperature dyeing of polyester fabric using 
ultrasound, Journal of the Society of Dyers and Colourists. 112(9), 245·248. 
AJex, C. -C.T .. Goh, N.K. and Chia, L.S. (1995) Effects of particle size morphology on ultrasonic 
induced ca\-itationaJ mechanisms in heterogeneous systems. Journal of Ihe Chemical Society. 
Chemical Communications, (2), 20 1. 
Alliger. H. (1 975) Ultrasonic disruption. American Labormory. 7(10). 75-85. 
Anbar. M. and Pecht, I. ( l 964a) On the sonochemical formation of hydrogen peroxide in water, Journal 
of Physical Chemistry. 68 (2). 352-355. 
AnOOr. M. and Pecht. I. ( 1964b) The sonolytic decomposition of organic solutes in dilute aqueous 
solutions. L Hydrogen abstraction from sodium formate. Journal of PhYSical Chemistry. 
68(6), 1460-1462. 
AnOOr. M. and Pechl I. (1967) The sonolytic decomposition of organic solutes in dilute aqueous 
solutions. In Oxidative deamination of ethylenediamine b~' BO· radicals. Journal of 
Physical ChemiStry, 71(5), 1246-1 249. 
Ando. T. and Kimura. T. (1990) Reactivity and selectivity in organic sonochemical reactions in\'ohing 
inorganic solids, Ultrasonics. 28(5), 326-332. 
Ando, T .. Sumi . 5 .. Kawate. T .. Ichihara. J. and Hanafusa, T. ( 1984) Sonochemical switching of reaction 
pathways in solid-liquid two--pbase reaction!!. Journal of the Chemical Society. Chemical 
Communications. (7).439440. 
References R-2 
Andreozzi. R.. Caprio, V .. lnsola, A , Marona, Rand Sanchirico, R. (1989) Advanced oxidation processes 
for tbe treatment of mineral oil~ontaminated wQtewaters, A WT98 - Advanced Wastewater 
Treatment, Recycling and Reuse, Milan, Italy. 891-894. 
Anonymous (1995) Inactivating Crypto oocysts with ultrasound, Worldwater and Environmental 
Engin f!l!ring, 18(1). 42 . 
Apfel. RE. (1972) The tensile strengtb of liquids, Scientific American, 227(6), 58-71. 
Apfel. R.E. (1984) Acou!ltic cavitation inception. Ultrasonics, 22(4), 167-173. 
Arantegui, J.. Prado, J., Chamarro, E. and Esplugas, S. (1995) Kinetics of tbe UV degradation of atrazine 
in aqueous solution in the presence of hydrogen peroJide, Journal of Photochemisrry and 
Photobiology. ParI A. Chemistry, 88(1), 65-74. 
Asher. R.C. (1987) Ultrasonics in cbemical analY!lis, Ultrasonics, 25(1), 17-19. 
Ayling. G.W. and Castrantas, H.M. (1981) Waste treatment with bydrogen peroxide, Chemical 
Engineering, 88(24). 79-82. 
Bablon, G., Bellamy, W.O., Billen, G., Bourbigot, M.-M., Daniel, F.B. , Erb, F., Gomella, C., Gordon. G., 
Hanemann. P .. Joret. J.-C .. et a1. (l99Ia) Cbapter 3. Practical application of ozone: 
principles and case studies. In Ozone in Water Treatmenl: Application and Engineering. B. 
Langlais. O.A Reckhow and n .R Brink. eds .. Le""is Publishers. Chelsea. Michigan, ISBN 
0-87371-474-1. 133-3 16. 
Bablon. G., Bellamy, W.O., Bourbigot, M. -M., Daniel, F.B., Dore, M., Erb, F., Gonion, G., Langlais, B., 
Lapianche, A.. Legube, B., et a\. (l991b) Chapter 2. Fundamental aspects. In Ozone in Water 
Treatment: Application and Engineering, B. Langlais, O.A. Reckhow and D.R. Brink, eds .. 
Lewis Publishers. Chelsea, Michigan. ISBN 0-8737 1-474-1. 11-1 32. 
Badcr. H. and Hoignc. J. (198 1) Determination of ozone in water by the indigo method . Water Research, 
15(4). 449-456. 
Baga. A.N .. Johnson, O.R.A.. Nazhat. N.B. and Saadalla-Nazhat, R.A. (1988) A simple spectro 
photometric determination of b~'drogen peroxide at low concentrations in aqueous solution. 
Analytical Chimica Acta, 204. 349-353. 
Bahnemann. D. and Han, EJ. (1982) Rate constants of the reaction of the h~'drated electron and 
h~'droIyl radic::aJ ""ith ozone in aqueous solution. Journal of Physical Chemistry, 86(2). 
252-255. 
Bailey, G.W. and White. J.L. (1970) Facton influencing tbe adsorption, desorptioD and mo,'ement of 
pesticides in soil. In Residue Reviews, F.A. Gunther and J.O. Gunther, eds .. Springer-Verlag, 
New York. Library of Congress Catalogue Card Number 62-18595. 32. 30-92. 
References R-J 
Balasubramaniam, K., Abesingha. C. and Kurnar, S.B. (1 995) PotentiaJ for application of ultrasound in 
tbe wutewater treatment process, Proceedings of the 25th Mississippi Water Resources 
Conference. BJ. Danial, 00.., 28 1-287. 
Barnier, P.F. and Petrier, C. (1 996) Study at 20 kIh and 500 kHz of tbe ultrasound-<lzone advanced 
oxidation system: 4-nitropbenol degradation, Journal of Advanced Oxidation Technology, 
1(2), 154-159. 
Behki, R.M. and Khan, S.U. (1986) Degradation of atrazioe by Pselldo",QlIas: N-dealkylation and 
debalogenation of atrazine and its metabolites, Journal of Agricultural and Food Chemistry, 
34(4), 746-749. 
Bellamy, w .n ., Hickman, G.T., Mueller, P.A and Ziemba, N. (1991) Treatment of VOC-contaminatcd 
groundwater by bydrogen peroxide and ozone oxidation, Research Journal of the Water 
Pollution Control Federation, 63(2), 120-1 28. 
Beltr.in, FJ. (1995) TbeoreticaJ aspects of tbe kinetics of competitive ozone reactions in water, Ozone 
Science & Engineering, 17(2), 163-1 81. 
Beltnin, F.J. (1991) TbeoreticaJ aspects of tbe kinetics of competitive fint reactions of ozone in tbe 
O.YB1~ and OJUV oudation processes, Ozone Science & Engineering, 19(1), 13-38. 
Beltmn. F.J.. Garcia-Araya, J.F. and Acedo, B. (19943) Advanced oudation of atrazine in water. Part L 
OzonatioD, Water Research, 28{1O), 2153-2164. 
Beltr.in. FJ .. Garcia-Ara)'a, J.F. and Acedo, B. (1994b) Advanced oxidation of atrazine iD ~·ater. Part [L 
Ozonation combined witb ultraviolet radiation, Water Research, 28(10), 21 65-2174. 
Beltr.in. FJ ., Garcia-Araya, IF., Alvarez, P.M. and Rjv3S, J. ( 1998) Aqueous degradation of atrazine and 
some of its main byproducts witb O'.tOne/bydrogen peroxide. Journal of Chemical Technology 
and Biotechnology, 71 (4 ), 345-355. 
Behmn. F.J ., G6mcz-Serrano, V. and Dumn, A. (1 992) Degradation kinetics of p-nitrophenol ozonation 
in water, Water Research, 26(1).9-17. 
Beltnin. F.J ., Ovejero. G. and Acedo. B. (l993a) Olidation of atrazine in water by ultraviolet radiation 
combined witb hydrogen peroxide, Water Research, 27(6). 1013-1021. 
Bcitmn, F.J .. Rivas, F.l and Accdo, B. (1993b) Direct, radical and competith'e reactions in tbe 
ozonation of ~'ater micropollutants, Journal of Environmental Science and Health. Part A . 
Environmental Science and Engineering and Tonc and Hazardous Substance Control. A28(9), 
1947- 1976. 
Benitez. F.J. , Beltran-Heredia. J., Acero. J.L . and Gonzalcz. T. (1996) Degradation of protocatechuic acid 
by ""0 adnnced oxidation processes: ozoneIUV radiation and H20VUV radiation. Water 
Research. 30(7), 1597-1604. 
References R-4 
Serlan. J. and Mason, T.J . (1992) Sonocbemistry: from researcb laboratories to industrial plants, 
Ultrasonics, 30(4), 203-212. 
Bhatnagar, A and Cbeung, RM (1994) Sonocbemical destruction of cblorinated Cl and C2 volatile 
organic compounds in dilute aqueous solutioD, Environmental Sdence and Technology, 28(8), 
1481-1486. 
Bien, J. (1988) Ultrasonic preparation of sludges to improve dewatering, Filtration and Separation. 
25(6), '25-426. 
Bien, 1. and Wolny, L. (1997) Cbanges of some sewage sludge parameten prepared witb an ultrasonic 
field , Water Science and Technology, 36(11), 101-106. 
Blumhorst, M.R. and Weber, lB. (1994) Cbemical venus microbial degradation of cyanazine and 
atrazine in soils, Pesticide Science, 42(4), 79-84. 
Bolzacchini. E .. Brambilla, A , Orlandi, M., Polesello, S. and Rindone, B. (1994) Oxidative patbways in 
the degradation of triazine berbicides: a mecbaniltic approacb, Water Science and 
Technology, 30(7), 129-136. 
Boncz, M.A. Bruning, H., Rulk.ens. W.H., SudhOlter, E.J.R , Hannsen, G.R and BijsterOOsch, J.W. (1997) 
Kinetic and mecbanistic aspects of the oxidation of cbloropbenols by ozone. Water Science 
and Technology, 35(4), 65-72. 
Boudjouk., P. (1986) Syntbesis witb ultrasonic waves, Journal o/Chemical Education. 63(5), 427-429. 
Boudjouk., P. and HaD.. B.-H. (1981 ) Organic sonochemistry. Ultrasound-promoted coupling of 
chlorosilanes in the presence of lithium wire, Tetrahedron Letters, 22(39), 38 13-3814. 
Boudjouk.. P .. Han. B.-H. and Anderson. K.R (1982) Sonochemica1 and electrochemica1 syntbesis of 
tetramesityldisilene, Journal of the American Chemical Society, 104(18), 4992-4993. 
Boudjouk.. P .. Thompson. D.P .. Ohrbom. W.H. and Han. B.-H. (1986) Effects of ultra!J()oic " 'a\'es 00 tbe 
generation and reacth'ities of !J()me metal powders. OrganometalJics. 5(6), 1257-1260. 
Bowers. A , Nctzer. A. and Norman. J.O. (1973) Ozooation of wastewater - !J()me technica1 and economic 
aspects, Canadian Journal o/Chemical Engineering, 51(3). 332-336. 
Bradfield. E. (1997) Alrazine technical (product code 34216). Material Safety Data Sheet no. PS 057, 
Sanachem. 
Brambilla. A .. Bolzacchini, E., OrJancti, M., Polcsello, S. and Rinoone, B. (1995) Reactivit~· of organic 
micropollutants with ozone: a kinetic study. Uft Chemistry Reports. 13(1-2), 79-84. 
Braslavsky, S.E. and Houk. K.N. (1988) Glossary of terms used in photocbemistry, Pure and Applied 
Chemistry, 60(7). 1055-1106. 
Bremner. D. ( 1986) Chemical ultrasonics. Chemistry;n Britain, 22(7). 633-638. 
References R-5 
Brink, O.R, Langlais, B. and Reckhow, O.A (1991) Cbapter 1. Introduction. In Ozone in Water 
Treatment: Application and Engineering, B. Langlais, O.A Reckhow and O.R Brink. eds. , 
Lewis Publishers, Chelsea, Michigan, ISBN 0-87371-474-1 , 1-10. 
Bull, RA and Zeff, lO. (1991) Hydrogen peroxide in advanced oxidation processes for treatment of 
industrial process and contaminated groundwater, First International Symposium: Chemical 
Oxidation, Technology for the 90s, Nashville, Tennessee, W.W. Eckenfelder, AR Bowers and 
lA Roth, eels., 26-36. 
Burkha.rd. N. and Guth, J.A (1976) Pbotodegradation of atrazine, atraton and ametryne in aqueous 
solution with acetone as a pbotosen8itiser, Pesticide Science, 7(1), 65-71. 
Burkhard, N. and Guth, lA (1981) Cbemical hydrolysis of 2-<:blor0-4,6-bis(alkylamino)-1,3,5-triazine 
herbicides and their breakdown in soil under the influence of adsorption, Pesticide Science, 
12(1), 45-52. 
Busing, W.R. and Levy, H.A (1963) Crystal and molecular structure of hydrogen peroxide: a 
neutron-diffraction study, Journal o/Chemical Physics, 42(9), 3054-3059. 
Bunner, l , Gutierrez, M. and Henglein, A (1991) Sonolysis of water-metbanol mixtures, Journal of 
Physical Chemistry, 95(4), 1528-1530. 
Caprio, V. Insola, A. , Lignola, P.G. and Volpicelli, G. (1982) A new attempt for tbe evaluation of tbe 
absorption connant of ozone in water, Chemical Engineering Science, 37(1). 122-124. 
CarbeTT)', 1.B. and Benzing, T.M. (1991) Peroxide pre-oxidation of recalcitrant toxic waste to enhance 
biodegradation, Water Science and Technology, 23(1 -3). 367-376. 
Carmichael, A.J.. Mossoba, M.M., Riesz, P. and Christman, c.L. (1986) Free radical production in 
aqueous solutions exposed to simulated ultrasonic diagnostic conditions. IEEE Transacrions 
on Ultrasonics. Fe"oelectrics and Frequency Control, UFFC-33(2), 148-155. 
Central Statistical Service (2000) RSA statistics in brief, httpJIwv.'w.statssa.gov.za!stats_in_brieC2OOO. 
Chan. G.Y.S., Hudson, M.l and isaacs, N.S. (1992) Degradation of atrazine by hydrolysis and by 
bydroxyl radicals. Journal of Physical Organic ChemiSTry, 5.600-608. 
Charpentier. J.-c. (1981) Mass-transfer rates in gas-liquid absorbers and reactors. In Advances in 
Chemical Engineering, T.B. Drew. G.R Cokelet. lW. Hoopes and T. Vermeulen, eds., 
Academic Press, New York, ISBN 0-12-008511 -9. 11. 1-133. 
ChaneIjee, A . Magee. 1.L. and Dey. S.K. (1983) Tbe role of homogeneous reactions in the radiolysis of 
water, Radiation Research, 96(1), 1-1 9. 
Chendke. P.K. and Fogler, RS. (1975) Macrosonics in industf1'. 4. Chemical processing, Ultrasonics, 
13( 1),3 1-37. 
References 
Chendke, P.K. and Fogler. H.S. (1983a) Sonoluminescence and sonochemicaJ reactions of aqueous 
carbon tetrachloride solutions, Journal 0/ Physical Chemistry. 87(8), 1362-1369. 
Chendke, P.K. and Fogler, H.S. (l983b) Effect of static pressure: on tbe intensity and spectral 
distribution of tbe soDoluminescence of ' .... ter, Journal 0/ Physical Chemistry, 87(9). 
1644-1648. 
Cheng. H.H. (1990) Organic residues in soils: mechanisms of retention and extractability, International 
Journal o/Environmental A nalytical Chemistry. 39(2), 165-171 . 
Cheung, H.M., Bhatnagar, A and Jan.sen. G. (1991) Sonocbemical destruction of cblorinated 
bydrocarbons in dilute aqueous solution, Environmental Science and Technology, 25(8) , 
1510-1512. 
Cheung, H.M. and Kurup. S. (1994) Sonochemical destruction of CFC 11 and CFC 113 in dilute 
aqueous solution. Environmental Science and Technology, 28(9), 1619-1622. 
Chiu. Y . ...c .• Chang, C-N .. Line, J.-G. and Huang, S.-J. (1997) Alkaline and ultrasonic pretreatment of 
sludge before anaerobic digestion, Waler Science and Technology. 36(11). 155-162. 
Chopra. S.L., Sethi, B.K. and Baweja, AS. (1973) Residues of atrazine (2-(:blor0-4~tbylamino-
 6-isopropyl-amino-s-triazine) in soils. Journal 0/ the Indian Society of Soil Science. 21(3). 
319-322. 
Chramosta, N .. De Laat, J., Dore, M ., Suty, H. and Pouillot, M. (1993) Rate constants for reaction of 
hydrosyl radicals with .f-triazines, Environmental Technology, 14(3), 215-226. 
Chri slman. CL. , Cannichael, AJ.. Mossoba, M.M. and Riesz, P. (1987) Evidence for free radicals 
produced in aqueous solutions by diagnostic ultrasound. Ultrasonics. 25(1). 31-34. 
Chyla. A.. Lorimer. J.P .. Mason. T.J. , Smith, G. and Waltan. 0.1 .. (1989) Modifying effect of ultrasound 
upon tbe electrocbemical oxidation of cyclohuanecarbos)'late. Journal of the Chemical 
Society. Chemical Communications. (9). 603-604. 
Clarke. N. and Knowles, G. (1982) High purity water using b)'drogen peroxide and ultra\iolet 
radiation, EjJ1uem and Water Treatmenl Journal, (September), 335-341. 
Clasen. J. and Sobotta. R. (1994) lnacti\'8tion or microorgani!lms b)-' ultrasound. /Valer Supp~v, 12(3-4). 
243-251. 
Clay. S.A and Koskinen. W.C (1990) Adsorption and desorptioo of atrazine. h)-'drosyatrazioe and 
S-glutathione atrazine on two soils, Weed &ience, 38(3). 262-266. 
C\lCtzer. R. (2000) Personal communication. Sasolburg. 
Colc. CA.. Ochs. L.O. and FUlUlcll. F.C (1974) Hydrogen peroside as a supplemental osygen source. 
Journal of the Water Pollution Control Federation, 46(11). 2579·2592. 
References R-7 
Contamine. F., Faid, F., Wilhelm, AM. , Serlan, J. and Delmas, H. (1994) Chemical reactions under 
ultrasound: discrimination of cbemical and pbysical effects, Chemical Engineering Science, 
49(248). 5865-5873. 
Cook, A.M. (1987) Biodegradation of s-triazine :lCnobiotics, FEMS Microbiology Reviews, 46(2), 93-116. 
Cook.. AM., Beilstein, P., Grossenbacher, H. and Hilner, R (1985) Ring cleavage and degradative 
pathway of cyanuric acid in bacteria, Biochemical Journal. 231(1), 25-30. 
Cost, M .. Mills. G .. Glisson.. P. and Lakin, 1. (1993) Sonocbemical degradation of p-nitropbenol in the 
presence of cbemical components of natural waters, Chemosphere, 27(9), 1737-1743. 
Crawford AE. (1963) A practical introduction to ultrasonic cleaning, Ultrasonics. 1(2),65-69. 
Crum. LA, Roy. RA, Oinno, MA and Church, C.C. (1992) Acoustic: cavitation produced by 
microtecond pulses of ultrasound: a discussion of some selected results. Journal of the 
Acoustical Society of America, 91(2), 1113-1119. 
Crwn. L.S. and Reynolds, G.T. (1985) SonoluminetCence produced by stable cavitation. Journal of the 
Acoustical Society of America. 78(1 ), 137-139. 
Currell. D.L. and Zecbmeister, L. (1958) On tbe ultrasonic cleavage of some aromatic and beterocyclic 
rings, Journal of the A merican Chemical Society. 80(1), 205-208. 
Cyr. P.J.. Zhang. Q. and Suri, RP.S. (1998) Sonochemical destruction of tricbloroethylene in water, 
A U7'98 - Advanced Waste-water Treatment. Milan, Italy, 173- 179. 
Dam. E. (1976) Physicochemical aspects of disinfection of water by means of ultrasound and ozone, 
WarerResearch. 10, 677~84 . 
Danckwerts. P.v. (1970) Chapter S. Absorption into agitated liquids. In Gas-liqUid reactions. 
McGraw-Hill. New York.. Library of Congress Catalogue Card Number 75-91678. 96-1 51 . 
Dao. T.H., Lavy. T.L. and Sorenscn. RC. (1979) Atrazine degradation and residue distribution in soil, 
Soil Science Society o/America Journal. 43(6), 1129-11 34. 
Davidson. R.S .. Safdar. A.. Spencer. 1.D. and Robinson. B. (1981) Applications of ultrasound to organic 
chemistl1'. Ultrasonics, 25( 1).35-39. 
De Laat. J.. Bcrger, P., T Poinot. Kar]X1 vel Leitner. N. and Dort. M. (1997) Modelling the oxidation of 
atrazine b~' H102lUV. Estimation of kinetic parameten. Ozone Science & J£ngineering. 
19(5). 395-408. 
De Laat. 1.. Chramosta, N .. Dore!. M .. Suty, H. and Pouillot, M. (1994) Rate constants for reaction of 
hydroxyl radicals with some degradation byproducts of atrazine by 0 3 or O.J lhOl. 
Environmental Technology. 15(5), 41 9-428. 
References 
De Laat, J., Doli, M. and Suty, H. (1995) Oxidation of s-t ria:zine5 by advllDced oxidation processes. 
Byproducts and kinetic rate constllDts, Revue des Sciences de L'Eau, 8(1), 23-42. 
de Souza-Barooza, J.C., petrier, C. and Luche, J.-L. (1988) Ultrasound in organic syntbesis. 13. Some 
fundamental IlSpectS of tbe sonocbemical Barbier reaction, Journal of Organic Chemistry, 
53(6), 1212-1218. 
de Visscher, A , van Eenoo, P., Drij"ers, D. and van Langenhove, H. (1996) Kinetic model for tbe 
sonocbemical degradation of mODocydic aromatic compounds in aqueous solution, Journal of 
Physical Chemistry, 100(28), 11636-11643. 
Degrois. M. and Baldo. P. (1974) A new electrical bypotbesis e:r.plaining sonoluminncence. cbemical 
actions and otber effects produced in gaseous cavitation, Ultrasonics, 12(1), 25-28. 
Del Duca. M .. Yeager, E .. Davies. M.O. and Hovorka. F. (1958) IJotopic tecbniques in tbe ItUd~' of tbe 
soDocbemical formation of bydrogen peroxide, Journal of the Acoustical Sociery of America, 
30(4), 30 1-307. 
Delchar, T.A. and Melvin, R J. (1994) A calorimeter for ultrasoond total power measurements. 
Measurement Science and Technology, 5(12). 1533- 1537. 
Didenko. Y.T., Nastich, D.N., Pugach, S.P., Polovinka, Y.A. and Kvocbka, V.I . (1 994) Tbe effect of bulk 
solution temperature OD the intensity and spectra of water sonolumincscence. Ultrasonics. 
32(1), 71-76. 
Donaldson, DJ ., Farrington, M.D. and }(rous, P. (1 979) Cavitation-induced polymerisation of 
nitrobeozcne, Journal of Physical Chemistry, 83(24). 3130-3135. 
Dries. 0 .. de Ca rte. B .. Liesscns, J.. Steurbaut, W., Dejonckheere, W. and Verstraetc. W. (1 987) 
Recalcitrance of atr azinc at low le\'el§ to aerobic and hydrogenot rophic mic roorganisms. 
Biotechnology Letters. 9(11 ), 8 11 -8 16. 
Duguet. J.P .. Bemazeau. F. and MaUeviaUe. J. (1990) Remon J of atrazine by ozone and ozone-b~'drogen 
peroxide combinations iD surface water, Ozone Science & Engineering, 12(2), 195-1 97. 
Duguet. JP .. Brodard E., Dussert, B. and Mallevialle, J (1 985) Improvement in the effecth'eness of 
ozoDation of drinking water through the use of hydrogen peroxide. Ozone Science & 
Engineering, 7. 21-258. 
Ounn. F. (1991) Ultrasound, IEEE Transactions on Education, 34(3). 266-268. 
DW AF (1997) National water policy for Soutb Africa. White Paper, Department of Water Affairs and 
Forestry, hnp:llwww-dwaf.pw".gov.za/idwaflDocumentsl. 
DW AF (2<X>O) Wate r con!J(n'ation strategy for tbe industry. mining and power generation user sector, 
Depanment of Water Affairs and Forestry, htlp:llwww~waf.pwv . go\'.za/idwaflDocumentsl. 
References R-9 
Eaton, AD., Clesceri, L.S. and Greenberg., AE. , eds. (1995a) SedioD 4500-CI B. Iodometric method I. 
In Standard Methods for the Examination of Water and Wastewater, 19th edition, American 
Public Health Association, Washington DC, ISBN 0~87553~223~3, 4/38-4/39. 
Eaton, AD., Clesceri, L.S. and Greenberg, AE. , eds. (1995b) Section 4500(0) B. Indigo colorimetric 
method. In Standard Methods for the Examination of Water and Wastew.'ater, 19th edition, 
American Public Health Association, Washington DC, ISBN 0-87553~223~3 , 41104-4/ 106. 
Ebert, E. and Dumford, S. W. (1976) Effects of triazine herbicides on tbe pbysiology of plants. In Residue 
Reviews, F.A Gunther and J.D. Gunther, eds. , Springer~Verlag. New York, ISBN 0387902228, 
65, 1-97. 
EHC 22 (1982) Environmental health criteria 22: ultrasound, World Health Organisation., 
ISBN 92-4-154082-6. 
Ellgehausen, H., Guth, lA and Esser, H.D. (1980) Fadors determining the bioaccumulation potential of 
pesticides in tbe individual compartments of aquatic food chains, Ecotoxicology and 
Environmental Safety, 4, I34~157. 
El'piner, I.E . (1959) On the chemical adion of ultrasonic waves, Soviet Physics ~ AcoustiCS, 5(2 ), 
135-146. 
EI'piner, I.E. (1960) Ultrasonic luminescence: a lJUn'ey, Soviet Physics - Acoustics, 6(1). 1-12. 
Emslie. N. (2001) Personal communication, Sasolburg. 
Erickson, L.E. and Lee, K.H. (1989) Degradation of atrazine and related s-triazines, Critical Reviews in 
Environmental Control, 19(1), 1~14 . 
Esser, H.o.. Dupuis. G .. Ebert. E., Marco, G. and VogeJ, C. (1975) Chapter 2, s~Triazines . In Herbicides: 
chemistry, degradation and mode of action, P.c. Kearney and 0 .0 . Kaufman. eds., Marcel 
Dekker. Inc .. New York, ISBN 0-8247-6175~8, 1, 129~208 . 
Eul, W .. Scherer. G. and Helmling.. 0. (199 1) Practical applications of h~'drogen peroxide for 
wastewater treatment. First International Symposium: Chemical Oxidation , Technology for the 
90s, Nashville. Tennessee. W.W. Eckenfelder. AR. Bowers and J.A. Roth, eds .. 68~77 . 
Fau!. C.FJ. (1996a) Ultrasound for the treatment of water and industrial effluent, ESKOM repan 
no. TRRlDS96IETO 16. 
Fau!. C.F.J. (19%b) Ultrasound for the treatment of water and wastewater, Part 2. ESKOM repan 
no. TRRIDS961ET028. 
Fischer. C.~H., Hart, E.J. and Henglein, A (1986a) BID isotope eIchange in the D2-820 system under 
the innuence of ultrasound, Journal of Physical Chemistry, 90(2). 222-224. 
References R-IO 
Fischer, c.-H., Hart, EJ. and Henglein, A (1986b) Ultrasonic irradiation of water in tbe presence of 
18,1802: isotope excbange and isotopic distribution of H202, Journal of Physical Chemistry, 
90(9), 1954-1956. 
Flint, E.B. and Suslick, K.S. (1989) Sonolumi.nescence from non-aqueous liquids: emission from small 
molecules, Journal of the American Chemical Society, 111(18), 6987-6992. 
Flint, E.B. and Suslick, K.S. (1991) The temperature of cavitation, Science, 253(5026), 1397-1399. 
Flisak, F. and Perna, A (1977) The influence of ultrasonics on liquid-liquid extraction, Ultrasonics, 
15(1), 27-29. 
Floros, 1.0. and Liang, H. (1 994) Acoustically assisted diffusion tbrougb membranes and biomaterials, 
Food Technology, 48(12), 79-84. 
Flosdorf, E .W., Chambers, L.A and Malisofi", W.M. (1936) Sonic activation in chemical systems: 
oxidations at audible frequencies, Journal of the American Chemical Society, 58(4), 
1069-1076. 
F1ynn, H.G. (1982) Generation of transient cavities i.n liquids by microsecond pulses of ultrasound, 
Journal of the Acoustical Society of America, 72(6), 1926-1 932. 
F1ynn, H.G. and Church, C.C. (1984) A mechanism for tbe generation of cavitation maxima by pulsed 
ultrasound, Journal of the ACousfi cal Society of America , 76(2), 505-512. 
Foster. D.M., Rachwal, AJ . and White, S.L. (1992) Advanced treatment for tbe removal of atrazine and 
otber pesticides, Water Supply, 10(2), 133-146. 
Franck, E.U. (1987) Auids at high pressures and temperatures, Journal of Chemical Thermodynamics, 
19(3). 225-242. 
Frizzell. L.A. (1988) Chapter 7. Biological effects of acoustic cavitation. In Ultrasound: Its Chemical, 
Physical and Biological Effects, KS. Suslick. ed .. VCH Publishers, Inc., New York. 
ISBN 0-89573-328-5, 287-303. 
Galassi. S .. Provini, A and Halfon, E. (19%) Risk assessment for pesticides and tbeir metabolites in 
water. International Journal of Environmental Ana~ytical Chemistry, 65(1-4), 33 1-344. 
Gan, J.. Seeker, RL. , Koskinen. w.e. and BuWer, 0 .0 . (1996) Degradation of atrazine in h\'0 soils as a 
function of concentration. Journal of Environmental Quality, 25(5). 1064-1 072. 
Gautheron. B. , Tainturier, G. and Degrand, C. (1985) Ultrasound-induced electrocbemical syntbesis of 
tbe anions Sez2" Se2" Tez2. and Te2" Journal of the American Chemical Society. 107( 19), 
5579-5581. 
Gaynor, 1.0., MacTavisll, D.C. and Find1ay, W.!. (1992) Surface and subsurface transport of atrazine 
and alacblor from a Brookston clay loam under continuous corn production, Archives of 
Environmental Contamination and Toxicology, 23(2), 240-245. 
References R-ll 
Gaynor, J.O .. MacTavish, o.e. and Findlay, W.L ( 1995) Atrazine and metolacblor loss in surface and 
subsurface nmoff from tbree tillage treatments in corn. Journal of Environmental Quality, 
24(2). 246·256. 
GelS (1999) Water affairs and forestry , South Africa Yearbook, http://www.gov.zalyearbooklwatcr. 
Gchringer, P., Szinovatz. W .• Eschweiler, H. and Haberl, R (1997) Oxidath'e treatment of a wastewater 
stream from a mola.net processing plant using ozone and advanced o~dation technologies, 
Ozone Science & Engineering, 19(2), 157-168. 
Geller, A. ( 1980) Studies on the degradation of atrazine by bacterial communities enriched from 
,'arious biotopes. Archives of ElWironmental Contamination and Toncology, 9(3). 289-305. 
Gericke, G. (1999) Advanced o~dation processes and the current ESKOM position with respect to the 
tecbnology. ESKOM repon no. TRRIC99/ 102. 
Giardi, M.T. , Giarc:lina, M.C. and Filacchioni, G. (1985) Cbemical and biological degradation of primary 
metabolites of atrazine by a Nocardia straiD, Agricultural and Biological Chemistry, 49(6) . 
1551·1558. 
Giguere. P.A and Srinivasan. T.K.K. (1977) The groond-state geometry of tbe BlOt and IhOt 
molecules, Journal of Molecular Spectroscopy, 66( 1), 168-170. 
Glaze. W.H. (1987) Drinking-water treatment with ozone, Environmental Science and Technology, 
21 (3). 224·230. 
Glaze. W.H .. Beltran, F., Tubkanen, T. and lean&, J.-W. (1992) Chemical models of advanced oxidation 
processes. Wafer Pollution Research Journal o/Canada, 27(1). 23-42. 
Glaze. W.H. and Kang. J.·W. (1989a) Advanced oxidation processes. Description of a kinetic model for 
the oxidation of hazardous materials in aqueous medi a. with ozone and hydrogen pero~de in 
a scmibatcb reactor. Industrial and Engineering Chemistry Research. 2S(l I}. 1573-1580. 
Glaze. W.H. and Kang, J.-W. (l989b) Advanced oxidation processes. Test of a kinetic model for tbe 
oxidation of organic compounds with ozone and b~'d rogen peroxide in a .semibatcb reactor, 
Industrial and Engineering Chemistry Research, 28(11 ). 1580-1 589. 
Glaze. W.H .. Kang. J .-W. and Chapin. D.H. (1 987) The cbemistry of water treatment processes 
in" ohing 07.ooe, bydrogen peroxide and ultra\'iolet radiation, Ozone Science & Engineering. 
9(4).335-352. 
Glaze. W.H .. Peyton. G.R . Lin. S .. Huang, R Y. and Burleson, J.L. ( 1982) Destruction of pollutants in 
water with ozone in combination with ultraviolet radiation. Part 2. Natural tribalomethane 
precursors. Environmental Science and Technology, 16(8), 454-458. 
References R-12 
Glaze. W.H. , Schep, R., Chauncey, W., Ruth. E.C., Zamoch, J.1 .• Aieta, E.M, Tate, C.H. and McGulre, 
M. J. (1990) Evaluating oxidants for tbe remo\'aI of model taste and odour compounds from a 
municipal water supply, Journal of the American Water Works Associatian, 82(5), 79-84. 
Gogale. P.G. and Pandit. AB. (2000) Engineering design method fOr cavitational reacton: L 
Sonocbemical reactors, AIChE Journal, 46(2),372-379. 
Gomella. C. (1972) Ozone practices in France. Journal of the American Waler Works Association, 64(1). 
3945. 
Gondrexon, N., Renaudinn, v. , Bemis. A, Gonthier. Y. and Boldo, P. (1993) Ultrasonic degradation 
kinetic study of chlorophenol aqueous solution, Environmental Technology, 14(6). 587-593. 
Gonzalez. MC. and MAnire, 0 .0 . (1997) Kinetics of 0 - and 0 3- in alkaline aqueous solutions, Water 
Science and Technology. 35(4), 49-55. 
Goodwin. T.J. (1990) Cbapter 5. Equipment. In Critical Reports on Applied Chemistry: Chemistry wi/h 
Ultrasound. T.J. Mason, ed. , Elsev:ier Applied Science, New Yor1c, ISBN 1-851 66422-X. 28, 
159-187. 
Goswami. S .• Gbosh. P.N. and Basumallick, S. (1988) Measurement of ultrasonic power in liquids - a 
review. Journal of the Institution of Engineers (India). 
Telecommunications, 69(1). 12-27. 
Part ET: Electronics and 
Greenwood, N.N. and Eamshaw, A. (1984a) Section 14.1.4. Atomic and pbysical properties of oxygen. 
In Chemislryofthe Elements, Pergamon Press, Oxford, ISBN 0"()8"()22057-6. 704-707. 
Greenwood, N.N. and Eamshaw, A (l984b) Section 14.1.5. Other forms of oxygen: ozone. In Chemistry 
oflhe Elements. Pergamon Press, Oxford ISBN 0..()8..()22057-6. 707-712. 
Greenwood. N.N. and Eamshaw. A (1984c) Section 14.2,3, Hydrogen peroxide. In Chemistry of the 
Elemems, Pergamon Press, Oxford, ISBN 0..()8~22057-6. 742-748. 
Grigg. B.C .. Assar, N. A and Turco, R.F. (1997) Remo\'al of atrazine contamination in soil and liquid 
systems using bioaugmeotation. Pesticide Science, 50(3). 211-220. 
Grinthal. W. and Ond.rey. G. (1992) Ultrasound seen but not heard, Chemical Engineering, 99(10 ). 
37-41. 
Guilh~n . M.D .. Blanco. J.. Canga. J.S. and Blanco, C.G. (1991) Study of the efl'ecti\'eness of 27 organic 
solvents in the extraction of coal tar pitcbes, Energy & Fuels, 5(1), 188· 192 . 
Gutierrez. M .. Henglein, A. and Dohrmann, J.K. (1987) H atom reactions in the sonolysis of aqueous 
solutions. Journal of Physical Chemislry, 91(27), 6687-6690. 
Gutierrez.. M., Henglein. A. and Ibanez. F. (1991) Radical scavenging in the sonolysis of aqueous 
solutions of r , Br', and N3-, Journal of Physical Chemistry. 95(15), 6044~7. 
References R-13 
Haag, W.R. and Yao, C.C.D. (1992) Rate constants for reaction of hydroxyl radicals with se"'eraJ 
drinking water contaminants, Environmental Science and Technology, 26(5), 1005-1013 . 
Haglund, P. (1997) Advanced ozone technology improves water quality in Nuho,.. No,,'gorod. Russia, 
Water and Wastewater International, 12( 1), 14-17. 
Han, B.-H. and Boudjouk, P. ( 1981) Organic sonochemistry. Ultruound-promoted coupling of organic 
balides in tbe presence of lithium wire, Tetrahedron Letters, 22(29), 2757-2758. 
Hall, B.-H. and Boudjouk, P. (1982a) Organic sonocbemistry. U1traJOUnd-promoted reaction of zinc 
with a,a '-dibromo-o-xylene. Evidence for facile generation of Q-xylylene, Journal a/Organic 
Chemistry. 47(4), 751-752. 
Han; B. -H. and Boudjou~ P. ~ J982b) Or;:anic sonochemidl}'. Sonic acceleration of the Refnnnaukr 
reaetion, Journal ojOrganic Chemistry, 47(25), 5030-5032. 
Han, B.-H. and Boudjouk, P. (l982c) Organic sonocbemidry. Ultrasonic acceleration of the reduction 
of simple and deactivated aryl balides using lithium aluminum bydride, Tetrahedron Lerters, 
23(16), 1643-1646. 
Hapeman, C.J. (1994) Cbapter 14. Oxidation of s-triazine pesticides. In American Chemical Society 
Symposium Series: Emerging Technologies in Hazardous Waste Management I V. D. W. Tedder 
and F.G. Pohland, eds., American Chemical Society, Washington, DC, ISBN 0-8412-2857-4, 
SS4, 223-233 . 
Hapeman, CJ. , Karns, 1.S. and Shelton, D.R. (1995) Total mineralisation of aqueous atrazine in tbe 
presence of ammonium nitrate using ozone and Klebsiella terrageno (strain DRS-I): 
mecbanistic considerations for pilot scale disposal. Journal 0/ Agricultural and Food 
Chemistry. 43(5). 1383-1391. 
Hapeman-Somich, C.L Gui-Ming, Z. , Lusby, W.R, Muldoon. M.T. and Waters, R (1992) Aqueous 
ozonation of atrazine. Product identification and description of the degradation pathway, 
Journal of Agricultural and Food Chemistry, 40(11), 2294-2298. 
Harcourt. R.D. and Roso, W. (1978) Valence-bond studies of 4-electron 3-centre bonding units. 1. The 
x-electrons of OJ. NOz' and CH1N1, Canadian Journal a/Chemistry. 56(8), 1093-1 101 . 
Harcourt. RD .. Skrezenek. F.L. . Wilson. RM. and Flegg.. RH. (1986) Ab initio \'3lence bond calculations 
for the ground datc of ozone, Journal o/the Chemical Society. Faraday Transactions 1/, 82(4). 
495-509. 
Hargis, L.a. (1988) Chapter 11. Oxidation-reduction titrations. In Ana(vtical Chemistry: Principles and 
Techniques, Prentice-HaIL Inc. , Englewood Cliffs. New Jersey. USA ISBN 0- 13-033507-X. 
260-299. 
References R-14 
Harris, C.I . (1967) Fate of 2-ch1oro-s-triazine herbicides in soil. Journal 0/ Agricultural and Food 
Chemistry, ]5(1 ), 157-162. 
Hart, EJ., Fischcr. C.-H. and Henglein, A (1986) lsotopic exchange in the sonolysis of aqueous solutions 
containing jot,14N1 and 15.ISN1, Journal o/Physical Chemistry, 90(22), 5989-5991. 
Hart, E.1. , Fischer, C. -H. and Henglein, A (l990a) Pyrolysis of acetylene in sonolytic cavitation bubbles 
in aqueous solution, Journal a/Physical Chemistry, 94(1 ). 284-290. 
Hart, E.l , Fischer, C.-H. and Henglein, A (l99Ob) Sonolysis of hydrocarbons in aqueous solution, 
Radiation Physics and Chemistry, 36(4), 511-516. 
Hart, E.J. and Henglein, A (1985) Free radical and free atom reactioDs in tbe soDolysis of aqueous 
iodide and formate solutions, Journal 0/ Physical ChemiStry, 89(20), 4342-4347. 
Hart, E.1. and Henglein, A (1986) Sonolytic decomposition of nitrous oxide in aqueous solution, Journal 
0/ PhYSical Chemistry. 90(22), 5992-5995. 
Han, EJ. and Henglein. A (1987) Sonochemistry of aqueous solutions: 82-~ combustion in cavitation 
bubbles. Journal of Physical Chemistry. 91(13), 3654-3656. 
Harvey, E.N. (1939) Sonoluminescence and sonic Chemiluminescence, Journal o/the American Chemical 
Society, 61(9), 2392-2398 . 
Haverhoek. S .. Kroning, l ., van Staveren, N.A and de Ruyter, M.A (1997) Advanced treatment of 
effluents from an agrocbemicaJ formulation plant, Warer Science and Technology, 35(10), 
155-1 63. 
Henglein. A. (1 985) Sonolysis of carbon dioxide, nitrous oxide and methane in aqueous solution. 
Zeitschriftflir Natur/orshung.. 408(1). 102-107. 
Henglein. A. (1987) Sonochemistr)': historical de,,·elopmenu and modern aspects, Ultrasonics, 25(1) . 
6-16. 
Henglein. A. and Gutierrez, M. (1988) Sonolysis of polymers in aqueous solution. New observations on 
pyrol~' sis and mecbanical degradation. Journal 0/ Physical Chemistry. 92(1 3). 3706-3707. 
Henglein. A. and Gutierrez. M. (1993) Sonochemistry and sonoluminescence: effe<:ts of external 
pressure. Journal o/Physical Chemistry, 97( 1). 158-162. 
Henglein. A. and Konnann, C. (1985) Scavenging of 80· radicals produced in the sonolysis of water, 
Internafionol Journal o/Radiation Biologv, 48(2). 251-258. 
Henglein. A . Ulrica R. and Lilie, J. (1989) Luminescence and chemical action by pulsed ultrasound. 
Journal afthe American Chemical Society, 1 J 1(6). 1974-1 979. 
References R-15 
Hogrefe, W. , Grossenbacher, H., Cook, A.M. and Hiitter, R (1985) Biological treatment specific for an 
industrial wastewater containing s-triazines, Biotechnology and Bioengineering, 27(9) , 
1291-1296. 
Hogrefe, W., Grossenbacher, H. , Cook, AM. and Hiiller, R (1986) Biotreatment of s-triazine-containing 
wastewater in a fluidized bed reactor, Biotechnology and Bioengineering, 28( 10), 1577-1581 . 
Hoigne, J. (1975) Comparison of tbe cbemical effects of ozone and of irradiation on organic impurities 
in water, Proceedings of the lAEA Symposium, Munich, 297-305. 
Hoigne, 1. (1997) lnter~alibration of HO· radical sources and water quality parameters, Water 
Science and Technology, 35(4), 1-8. 
Hoigne, 1. and Bader, H. (1976) The role of bydroxyl radical reactions in ozonation processes in 
aqueous solutions, Water Research, 10(5), 377-386. 
Hoigne, 1. and Bader, H (1983) Rate constants of reactions of ozone witb organic and inorganic 
compounds in water - Part I, Water Research. 17(2), 173-183. 
Howkins. 5.0 . (1969) Diffusion rates and tbe effect of ultrasound, Ultrasonics, 8(April), 129-130. 
Hoyle, 8.S. and Luke. S.P. (1994) Ultrasound in the process industries, Journal of Engineering Science 
and Education, 3(3), 119-1 22. 
Hua. I. , H6chemer, RH. and Hoffmann, M.R (l995a) Sonocbemical degradation of p-nitropbenol in a 
parallel-plate near-field acoustical processor, Environmental Science and Technology, 29(11 ). 
2790-2796. 
Hua, I., H6chemer, RH. and Hoffmann, M.R. (1995b) Sonolytic bydrolysis of p-nitropbenyl acetate: tbe 
role of supercritical water, Journal of Physical Chemistry, 99(8). 2335-2342. 
Huang, CM. and Banks, M.K. (1996) Effect of ozonation on tbe biodegradability of atrazine in GAC 
columns, Journal of Environmental Science and Health. Part B. Peslicides. Food Contaminants 
and Agricultural Wastes, B31(6), 1253-1266. 
Huang, CP .. Dong.. C. and Tang, Z. (1993) Advanced chemica1 oxidation: its present role and potentia1 
future in hazardous waste treatment. Waste Management. 13(5-7), 361-377. 
Huang. R.L.. Gog, S.H. and Ong. S.H. (1974) Cbapter 3. Cbaracteristics of radical reactions. In The 
Chemistry of Free Radicals, Edward Amold (Publishers) Ltd , London, ISBN 0 7131 2418 O. 
19-44. 
Huber. W. (1993) Ecotoncological rele\'ance of atrazine in aquatic systems, Environmental Toxicology 
and Chemistry, 12(10). 1865-1881 . 
Hunicke, R.L. (1990) Industrial applications of high power ultrasound for cbemical reactions, 
Ultrasonics, 28(5). 291-294. 
References R-16 
HUSlen, K., Moza, P.N. and Pouyet, B. (1991) Pbotocatalytic degradation of s-triazine berbicides. 
Toxicological and Environmental Chemistry, 31-32, 97-102. 
International Critical Tables ( 1928) Vol. Ill, McGraw-Hill, New York. 
Jennings, B.H. and Townsend, S.N. (1961) The sonochemical reactiolls of carbon tetrachloride and 
chloroform in aqueous suspension in an inert atmosphere, Journal of Physical Chemistry. 
65(9), 1574-1579. 
Jessee. lA, Benoit, RE., Hendricks, AC. , Alien, G.c. and Nea1, J.L. ( 1983) Aaaerobic degradation of 
cyanuric acid, cysteine, and atrazine by a facultative anaerobic bacterium, Applied and 
Environmental Microbiology, 45(1), 97-102. 
Jones. K.H. , Sanderson. D.M. and Noakes, D.N. ( 1968) Acute toxicity data for pesticides, World Review 
of Pest Control. 7(3), 135-143. 
Jutzi, K.. Cook, AM. and Hfitter, R (1982) The degradative path"'ay of tbe s-triazine melamine, 
Biochemical Journal, 208(3). 679~84. 
Kang, I -W., Park, H.-S., Wang, R-Y., Koga, M., Kadokami, K., Kim, H.-Y., Lee, E .-T. and Oh, S.-M. 
(1997) Effect of ozonation for treatment of micropollutants present in drinking water 
source, Water &ience and Technology, 36(12), 299-307. 
Kasrils, R. (2000) Address by the Minister of Water Affairs and Forestry at the National Water Week 
Celebrations. 
Kaufman, 0 .0 . and Keamey, P.C. ( 1970) Microbial degradation of s-triazine herbicides. In Residue 
Reviews, F.A. Gunther and J.O. Gunther, eds., Springer-Verlag, New York. Library of Congress 
Catalogue Card Number 62-18595, 32, 235-265. 
Keam~'. P.c. . Mu1doon. M.T. and Somich. CJ. ( 1987) UV-(lZO~atioD of ele\'eD major pesticides as a 
waste disposal pretreatment. Chemosphere. 16(10·12), 2321-2330. 
Keamey. P.c. , Mu1doon. M.T. . Somich, C.L Ruth, J.M. and Voaden. DJ. (1988) Biodegradation of 
ozonated atrazine as a l't·astewater disposal system. Journal of Agricultural and Food 
Chemistry, 36(6). 1301-1306. 
Khan. S.U. and Schnitzer, M. ( 1978) UV irradiation of atrazine iD aqueous fuMe acid solution. Journal 
of Environmental Science and Health. Part B. Pesticides, Food Contaminants and Agricultural 
Wastes. B13(3), 299-310. 
Kidd.. M. (1997a) Chapter 3. The en\-ironmeut and the c.onstitution. In Environmental Law. A South 
African Guide. Juta & Co. Kenwyn. South Africa, ISBN 0 7021 4204 2. 34-46. 
Kidd.. M. (I997b) Chapter 6. Water law and the environment. In ElfVironmental Law. A South African 
Guide. Juta & Co, Kenwyn, South Africa. ISBN 0 7021 4204 2. 83-.101. 
References R-17 
Kiipatrick.. M.L.. Henick, e.C. and Kilpatrick. M. (1956) Tbe decomposition of ozone in aqueous 
solution. Journal o/the American Chemical Society, 78(9), 1784-1789. 
Kim. S.1 .. Oh. K.H., Lee. S.H., Choi. S.S. and Lee, K.C. (1997) Stud)' on JeCondary reaction and fate of 
bazardous chemicals by oxidants, Water Science and Technology, 36(12), 325-331. 
King. R.O. and Forster, e.F. (1990) Effects of sonication on aeth'ated sludge. Enzyme and Microbial 
Technology, 12(2), 109-115. 
Kirchener, J.R. (1981) Hydrogen peroxide. In Kirk-Othmer Encyclopedia 0/ Chemical Technology. 
M. Grayson and D. Eckroth, eels.. 3rd edition. John Wiley & Sons, New York, 
ISBN 0-471-02066-4, 13, 12-38. 
Koga, M., Kadokami, K and Shinohara, R (1992) Laboratory-scale ozonation of water contaminated 
""lib trlU!e peitieidC!J. Wotu,. St!iullc(! OIlD TecJJ1!OltJgv. 2~9_ ll ) . 2257-2260. 
Koike. T. (1992) Sonolysis studies of alcobols in aqueous solutions by gaseous products analysis, 
Bul/elin a/the Chemical Society 0/ Japan, 6S(ll). 3215-3217. 
Kondo. T .. Kirschenbaum, L.J ., K.im, H. and Riesz. P. (1993) Snnolysis of dimetbyl sulforide-water 
mixtures: a spin-trapping study. Journal o/Physical Chemistry, 97(2). 522-527. 
Koskinen. W.e.. Chen&, H.H., Jarvis, L.J. and Sorenson, B.A. (1995) Cbaracterization of mechanisms of 
pesticide retention in llOils using tbe IUpercritical fluid extraction tecbnique. International 
Journal o/Environmental Analytical Chemistry, 58(1 -4), 379-385. 
Koskinen, W.C. and Clay. S.A. (1997) Factors affecting at razine fate in Nortb Central US soils. Reviews 
0/ Environmental Contamination and Toxic%gy, ISI , 117-1 65 . 
Koskincn. W.e. and Rochcttc. E.A. (1996) Atrazine sorption-desorption in field-moi~1 soils. 
Internarional Journal of Environmental Analytical Chemistry, 65(1-4), 223-230. 
Koilincn. W.C .. Sellung. KE .. Baker. lM .. Barber. B.L. and Dowdy. R.H. (1994) Ultntsonic 
decomposition of atrazine and alacblor in water. Journal 0/ Environmental Science and 
Health. Part B. Pesticides. Food Contaminants and Agricultural Wastes, 829(3). 581-590. 
Kotronarou. A .. Mills. G. and HofTmann. M.R (1991) Ultrasonic irradiation of p-nHropbenol in aqueous 
solution. Journal 0/ Physical Chemistry, 95(9). 3630-3638. 
KOlronarou. A .. Mills, G. and Hoffmann. M.R (1992a) Decomposition of parathion in aqueous solution 
by ultrasonic irradiation. Environmental Science and Technology. 26(7). 1460-1462. 
Kotronarou. A. . Mills. G. and HofImann, M.R. (1992b) Oxidation of bydrogen sulfide in aqueous solution 
b~' ultrasonic irradiation . Environmental Science and Technology. 26(12). 2420-2428. 
Koubck, E .. Hagget, M.L. . Battaglia. CJ., Ibne-Rasa, KM .. Pyun. H.Y. and Edwards, J.O. (1963) Kinetics 
and mecbanism of tbe spontaneous decompositions of some peroxoacids, hydrogen peroxide 
and t-but~' 1 bydroperoxide. Journal o/the American Chemical Society. 8S(l5). 2263-2268. 
References R-18 
Krishna, C.M., Kondo, T. and Riesz, P. (1 989) Sonocbemistry of alcobol-water mixtures: spin-trapping 
evidence for tbermal decomposition and itOtope-exchange reactions, Journal of Physical 
Chemistry, 93(13), R66-5172. 
Krishna, C.M., Lion, Y., Kondo, T. and Riesz, P. (1987) Thermal decomposition of metbanol in the 
sonolysis of metbanol-" 'ater mixtures. Spin-trapping evidence for isotope excbange 
reactions, Journal of Physical Chemistry, 91(23), 5847-5850. 
Kri$loL D.S., KlOI2, H. and Parker, R C. (1981) The effect of ultrasound on tbe alkaline hydrolysis of 
nitropbenyl esters, Tetrahedron Letters, 22(10), 907-908. 
Kruger. E.L., Rice, PJ., Anhalt, l e ., Anderson, T.A and CoalS, J.R (1997) Comparative fates of 
atrazine and deethylatrazine in sterile and non-sterile soils, Journal of Environmental 
Quality, 26(1), 95-101. 
Kruger, E.L., Zhu. B. and CoalS, J.R. (1996) Relative mobilities of atrazine. five atruine degradates. 
metolacblor and simazine in soils of Iowa. Environmental Toxicology and Chemistry. 15(5). 
69 1';;95. 
Kruus, P. , Lawrie. J.AG. and O'Neill, M.L. (1988) Polymerisation and depolymerilation by ultrasound. 
Ultrasonics, 26(6). 352-355. 
Krzesinska, M. and Pajak., J. (1988) Application of ultrasonic measurements to the determination of 
concentration of coal extracts, Fuel Processing Technology, 19(3), 229-233. 
Kuo, J.-F. and Moll, L. (1997) Disinfection and antimicrobial processes, Water Environment Research, 
69(4). 526-534. 
Kyle. B. (1 995) Successful lnduslrial Experimentation, VCH Publishers. New York.. ISBN 1-5608 1-050-5. 
Labum. R J. (l995) Water suppl~' in South Africa. Journal of Water Supp(v Research and Technology -
 Aqua, 44(4), 161-165. 
Lambert. S.O., Graham. NJ.D. and Croll, R T. (1996) Degradation of selected herbicides in a lowland 
surface water b~' ozone and ozone-bydrogen peroxide, Ozone Science & Engineering. 18(3). 
25 1·269. 
Lamoureux. G.L., Stafford.. L E. and Shimabukuro. R.H. (1972) Conjugation of 2-chlor0-4.6-
 bis(alk~'lamino}-s-triazine! in higher planls. Journal of Agricultural and Food Chemistry, 
20(5), 1004-1010. 
La,,)'. T.L .. Roeth. F. W. and Fenster. C.R (1973) Degradation of 2,4-D and a.razine at tbree soil deptbs 
in the field. Journal o/Environmental Quality, 2(1). 132·1 37. 
Leclercq, P.A and PacAkova, V. (1 979) Gas chromatrograpby and mu s spectrometry of 
bis(alkylamino)-s-triazines. Journal of Chromatography, 178(1). 193-207. 
References R-19 
Leeman, S. and Yaughan, P.W. (1992) Cavitation pbenomena. In Cu~nt Trends in Sonochemistry, G.J . 
Price. ed. , RoyaJ Society of Chemistry, Cambridge, ISBN 0-85186-365-5,. 
Leeson, A , Hapeman, CJ. and Shelton. D.R. (1993) Biomineralisation of atmine ozonation products. 
Application to tbe development of a pesticide waste disposal system, Journal of Agricultural 
and Food Chemistry. 41(6), 983-987. 
Legube, 8., Guyon, S. and Dore, M. (1987) Ownation of aqueous Mllutions of nitrogen heterocyclic 
compounds: benzotriazoles. atrazine and amitrole, Ozone Science & Engineering, 9(3) , 
233-246. 
Lenart. 1. and AusUlnder, D. (1980) The effect of ultrasound on diffusion tbrougb membranes, 
Ultrasonics, 18(5), 216-218. 
Liakou. S .. PavIOll; S. and Lyberatos, G. (1997) OzouatioD of HO dyes, Water Science and Technology, 
35(4).279-286. 
Lii. R. -R. , Gorse, RA, Sauer, M.C. and Gordon, S. (1980) Rate constant for the reaction of HO· with 
BOO·, Journal of Physical Chemistry. 84(8), 819-821 . 
Lillard. H.S. (1994) Decontamination of poultry skin by sonication, Food Technology. 48(12), 72-73. 
Lin. J.-G .. Chang. C.-N. , Wu, J.-R and Ma. Y.-S. (1996) Enhancement of decomposition of 
2-<:blorophenol with ultruoundIH10z process, Water Science and Technology, 34(9), 41-48. 
Lindley. 1.. Lorimer. J.P. and Mason, T.J. (1986) Enhancement of an UUmann cou pling reaction induced 
by ultrasound, Ultrasonics. 24(5), 292-293 . 
Lindley. 1.. Lorimer. lP. and Mason. TJ. (1987) Sonocbemically enhanced UlImann reactions. 
UltrasoniCS. 25(1), 45-48. 
Liu. S.-C. and Wu, H. (1934) Mecbanism of oxidation promoted b~' ultrasonic radiation, Journal of the 
American Chemical Sccie~y. 56(5). looS -1 007. 
Loosli. R. (l99S) Epidemiology of atrazine, Reviews of Environmental Contamination and Toxicology. 
143. 47-S7. 
Lorimcr. J.P. (1990) Chapter 4. Polymers. In Critical Reporls on Applied Chemistry: Chemislry 'K-Uh 
Ultrasound. TJ. Mason. ed. . EIscvier Applied Science. Ne\'!' York. ISBN 1-851664 22-X 28. 
lI5-I S7. 
Lorimer. JP .. Mason, T.J. and Fiddy, K. (1991 ) Enhancement of cbemical reacth'it)· b)' power 
ultrasound: an aJternath'e interpretation of tbe bot spot. Ultrasonics. 29(4). 338-343. 
Luche. J.L. (1992) Developments of the new 'experimental tbeory' of sonocbemistry initiated in 
Grenoble. Ultrasonics. 30(3). 156-1 62 . 
References R-20 
Luche, lL. and Damiano, l-<:. (1980) Ultrasound in organjc syntbesis. 1. Effect of tbe formation of 
litbium organometallic reagents, Journal of the American Chemical Society. 102(27) . 
7926-7927. 
Luche, lL., Einhom, C., Einhom, l , de Souza-Barboza, l e., ?etrier, C., Dupuy, e., Detair, P. , Allavena, 
e. and Tuschl, T. (1990) Ultrasooic waves as promoters of radical processes in cbemistry: tbe 
case of organometallic reactions, Ultrasonics. 28(5), 316-321 . 
Lykins. B.W. and Koffskey, W. (1986) Producu identified at an alternative disinfection pilot plant, 
Envi,.onmental Health Perspectives, 69, 119-127. 
Lykins, B.W., Koifsk.ey, W.E. and Miller, RG. ( 1986) Cbemical products and toxicologic effects of 
disinfection, Journal of the American Water Works Association. 78(1 1), 66-75. 
Makino. K., Mossoba, M.M. and Riesz, P. (1982) Cbemical effects of ultrasound on aqueous solutions. 
Evidence for HO· and H· by spin trapping, Journal of the American Chemical Society, 
104(12), 3537-3539. 
Makino, K .. Mossoba, M.M. and Riesz, P. ( 1983) Cbemical effects of ultrasound on aqueous solutions. 
Formation of bydroxyl radicals and bydrogen atoms. Journal of Physical Chemistry. 87(8). 
1369-1377. 
Manly, T.O. (1962) Hydrogen peroxide and other peroxides in cbemical manufacture, Chemistry & 
Industry, (I). 12-18. 
Marco. A , Esplugas, S. and Sawn, G. (1997) Bow and wby combine cbemical and biological processes 
for wastewater treatment. Waler Science and Technology, 35(4). 32 1-327. 
Margulis. M.A. (1 %9) Sonotuminescence and ullrasonic chemical reactions, Soviet PhYsics - Acoustics. 
15(2). 135-151. 
Margulis, M .A ( 1985) Sonoluminescence and sonocbemical reactions in ca\>ltation field s. A re\>le",', 
Ult,.asonics. 23(4). 157-169. 
Margulis. M .A (1992) Fundamental aspects of sonocbemi!j:lry. Ultrasonics, 30(3), 152-255. 
Martin. N. and Galey. C. (1994) Use of stalic mixer for oxidation and disinfection by ozone. Ozone 
Science & Engineering. 16(6). 455-473. 
Martin, P.D. (1992) Sonochemistry: a cbemical engineer's \·ie\\". In Cur,.ent T,.ends in Sonochemistry. 
G.J. Price. ed.. RO)'al Society of Cbemistry, Cambridge, ISBN 0-85186-365-5, . 
Martin. P.D. (1993) Sonocbemistry in industry: progress and prospects. ChemiStry & industry. (7). 
233-236. 
Martin. P.D. and Ward, L.D. (1992) Reactor design for sonocbemical engineering. Transactions of the 
institute o/Chemical Engineering, 70(A). 296-303 . 
References R-21 
Mason, Tl (1990) Chapter 1. Introduction. In Critical Reports on Applied Chemistry: Chemistry with 
Ultrasound. TJ. Mason, eel , Elsevier Applied Science, New York, ISBN 1-85166-422-X, 28, 
1-25. 
Mason, T.J. (1991) Chapter X. In Practical Sonochemistry: User's Guide to Applications in Chemistry and 
Chemical Engineering, ElIis Horwood, New York, ISBN 0-13-682642-3. 
Mason. TJ. (1992) Industrial sonochemistry: potential and practicality, Ultrasonics, 30(3), 192-1%. 
Mason, TJ. (1993) Sonochemistry: a technology for tomorrow, Chemistry & Industry, (2). 47-50. 
Mason, TJ. and Berlan, J. (1992) Ultrasoond in industrial processes. The problems of scale-up. In 
Current Trends in Sonochemistry, G.J. Price. ed , Royal Society of Chemistry, Cambridge. 
ISBN 0-85186-365-5, 148-157. 
Mason, T J. and Cordemans de Meulenaer, E. (1998) Chapter 8. Practical considerations for process 
optimization. In Synthetic Organic Sonochemistry, l-L. Luche, ed. Plenwn Publishers, New 
Yod<, , 301 -329. 
Mason, T,], and Cordemans, E.D. (1996) Ultrasonic intensification of cbemical processing and related 
operations: a review, Transactions of the Institute of Chemical Engineering, 74( A), 511 -516. 
Mason, Tl, Lorimer, J.P. and Bates, D.M. (1992) Quantifying sonochemistry: casting some light on a 
'black art', UltrasoniCS. 30(1 ), 40-42. 
Mason. TJ.. Lorimer, J.P. and Mistry, B.P. (1985) The effect of uitrasoond on the solvolysis of 
2-chloro-2-methylpropane in aqueoos ethanol, Tetrahedron, 41(22), 5201-5204. 
Mason, T.J., Lorimer, lP. and Walton, DJ. (1990) Sonoelectrochemistry, Ultrasonics, 28(5), 333-337. 
Mason, T.J., Newman. A P. and Phull, S.S. (1993) Sonochemistl1' in water treatment, Proceedings of fhe 
2nd infernational Conference on Advances in Water and Effluent Treatment, M.J.D. White, ed .. 
243-250. 
Mason, T.L Newman. A,P., Phull, S.S., Pollet, B. and Hun, K R (1994) The use of ultrasound to 
enhance water treatment, Proceedings of H.rdrotop '94, Colloque Mieu.x Gere L'Eau. 2. 
508-515. 
Mason. W.P. (1976) Sonics and ultrasonics: earty history and applications, iEEE Transactions on Sonics 
and Ultrasonics. SU-23(4), 224-232. 
Masten. SJ. and Davies. S.H.R. (1994) The use of ozonation to degrade organic contaminants in 
wastewaten. Environmental Science and Technology, 28(4), 180-185. 
Masten. S.l , Galbraith, MJ. and Davies, S.H.R. (1997) Oxidation of 1,3,.5-trichlorobenzene using 
adnnced oxidation processes, Ozone Science & Engineering. 18(6), 535-547. 
Matturro. M.G., Liotta, R. and Reynolds, R.P. (1 990) Ultrasonically enhanced coal extraction: use of 
base-solvent ~'ollen coal slurries. Energy & Fuels. 4(4). 346-35 1. 
ReferenCe! R-22 
Mead. E.L .. Sutherland, RG. and Verrall, RE. (1975) Tbe ultrasonic degradation of tbymioe, Canadian 
JournalolChemistry. 53(16), 2394-2399. 
Mead, E.L. , Sutherland, RG. and Venall, RE. (1976) The effect of ultrasound on water in tbe presence 
of dissolved gases, Canadian Journal ofChemisrry, 54, 1114- 1120. 
Meakins, N.C., Bubb, lM. and Lester, I N. (1994) Tbe bebaviour of the s-triazine berbicides, atrazine 
and simazine, during primary and secondary biological waste water treatment, 
Chemosphere. 28(9), 16 11-1622. 
Meijers, R.T., Oderwald-Muller, EJ., Nuhn, P.A N.M. and Kruithof, l e. (1995) Degradation of 
pesticides b)' ownation and advanced oxidation, Ozone Science & Engineering, 17(6) , 
613-086. 
Meijers, R.T. . van der Veer, A l and Kruithof, le. (1 993) Degradation of pesticide! by ozonation and 
advanced oridation, Water Supply, 11(304), 309-320. 
Miller, lL.. Wollum, A.G. and Weber, J.B. (1 997) Degradation of carbon-14-atrazine and 
carbon-14-metolacblor in soil from four deptbs. Journal of Environmental Quality, 26(3). 
633-638. 
Miller, N. (1 950) CbemicaJ action of sound waves on aqueous solutions. Transactions of the Faraday 
Society, 45. 546-550. 
Mirgain, 1., Green. O.A and MORteil, H. (1993) Degradation of atrazme in laboratory microcosms: 
isolation and identification of tbe biodegrading bacteria. Environmental Toncologv ond 
Chemistry. 12(9), 1627-1634. 
Mirgain. I. , Green. G.A. and Monteil. H. (1 995) Biodegradation of the herbicide atrazine in 
groundwater under laboratorJ conditions. Environmental Technologv. 16(10). 967-976. 
Misik. V .. Miyoshi. N. and Riesz, P. (1995) EPR spin-trapping study of the sonolysis of 820ID10 
mixtures: probing the temperatures of cavitatioo regions. Journal of Physical Chemi.ury. 
99( J I ). 3605-36 11 . 
Montiel. A and Welte, B. (1992) Alternative options for atrazine: incidence on water treatment de\i ce, 
Water Science and Technology. 25(11 ). 103-1 10. 
Morooka, S .. lkcmizu, K. and Kato, Y. (1979) The decomposition of ozone in aqueous solution. 
Internalional Chemical Engineering. 19(4}. 650-654. 
Mwler. M. (2000) Inter-basin '-\'ater sbaring to acbieve ll'ater security - a Soutb African perspecth'e, 
Proceedings of the World Water Forum. 
Munter. R. . Preis, S .. Kamenev, S. and Siirde, E. (I993) Methodol~' of ozone introduction into water 
and wastewater treatment. Ozone Science & Engineering. 15(2). 149-1 65. 
References R-23 
Nadez.hclin. AD. (1988) Mecbanism of ozone decomposition in water. Tbe role of termination, 
Industrial and Engineering Chemistry Research, 24(4), 548-550. 
Naicker, S. (1999) The extraction and quantification of triazine berbicides using bigb performance 
liquid cbromatograpby-ultra violet diode array detector .. B.Tech dissertation, ML Sultan 
Technikon. Durban. 
NCRP 74 (1983) Biological effects of ultruound: mechanisms and clinical implications, National 
Council on Racliation Protection and Measurements, ISBN 0-913392-64-2. 
NebeL C. (1981) Ozone. In Kirk-Othmer Encyclopedia of Chemical Technology, M. Grayson and 
D. Eckroth, eds., 3rd edition. John Wiley & Sons, New York. ISBN 0-471..02069-9, 16, 683-71 3. 
Neis, U. (2000) Ultrasound in water, wastewater and sludge treatment, Water2] , (April), 36-39. 
Nelieu, S., Kerhoas, L. and Einhorn, J. (1996) Atrazine degradation by ozonation in tbe presence of 
methanol as scavenger, international Journal of Environmental Analytical Chemistry, 65(1-4 ), 
297-311. 
Neppiras, E.A (1980) Acoustic cavitation, Physics Reports, 61 (3). 159-251 . 
Neppiras. E.A. (1984) Acoustic cavitation: IlD introduction, Ultrasonics. 22(1).25-28. 
Neppiras, E.A and Fill, E.E. (1969) A cyclic cavitation process, Journal of the Acoustical Society of 
America. 46(5). 1264-1 271. 
Nick. K .. SchOler, H.F. , Marlc, G., SOylemez, T. , Akhiaq, M.S. , Schuch.mann, H.-P. and von Sonntag. C. 
(1992) Degradation of some triazine herbicides by UV radiation sucb as used in tbe UV 
disinfection of drinking water, Journal of Water Supply Research and Technology - Aqua, 
41 (2). 82-87. 
Niemczewski . B. (1980) A comparison of ultrasonic ca,itation intensity in liquids. Ultrasonics. 18(3). 
107-110. 
Noltingk. B.E. and Neppiras, E.A. (1950) Cavitation produced by ultrasonics. Proceedings of the 
Physical Society of London. Section B, 63, 674~85. 
Nsabimana. E .. Bohatier. L Belan. A , Pepin, D. and Charles, L. (1996) Effects of the herbicide atrazine 
on the actinted sludge process: microbiology and functional ,,;el\'s. Chemosphere. 33(3). 
479-494. 
Numachi. F. (1965) An experimental study of accelerated cavitation induced by ultrasonics, Journal of 
Basic Engineering. (December). 967-976. 
O'Keeffe. J.H .. Uys. M. and Bruton, M.N. (1992) Cbapter 13. Freshwater systems. In Environmental 
Management in South Africa. R.F. Fuggle and M.A. Rabie. eds .. Juta and Company. Cape Town. 
ISBN 0-7021 -2847-3. 277-315. 
References R-14 
Olson, R G. and Hammitt, F.G. (1969) Higb-speed photograpbic studies of ultrasonically induced 
cavitation, Journal of/he Acoustical Sociely of America. 46(5), 1272-1283. 
Olson, T.M. and Bam ier, P.F. ( 1994) Oxidation kinetics or natural organic matter by sonolYJis and 
ozone, Wa/er Research, 28(6), 1383-1394. 
Otter, E., Baser, I. and Acar, K. (1995) Use or ultrasonic energy in reactive dyeing or cellulosic fabrics, 
Journal of the Sociely of Dyers and Colourists, 111(9), 279-281. 
Orlandini, E., Gebereselassie, TG., Kruithof, J.C, and Schippers, J.e. (1997) Effect or ozonation on 
preloading or background organic matter in granular activated carbon filten , Wafer Science 
and Technology, 35(7), 295-302. 
Orlandini. K , Siebel, M.A., Graveland, A. and $chippers, lC. (1996) Pesticide removal by combined 
IY.wnation and granular activated carbon filtration, Wo/er Supply, 14(2).99-108. 
Orzechowska, G.E., Poziomek, E.]., Hodge, V.F. and Engelmann, W.H. (1995) Use of sonocbemistry in 
monitoring cblorinated bydrocarbons in water, Environmental Science and Technology, 29(5), 
1373-1379. 
Paillard.. H., l..egube, B., Allemane, E. and [)ore, M. (1991) O xidation or atrazine and Jimazine by 
combined ozone/hydrogen peroxide process, Proceedings of the Water Qualiry Technology 
Conference, Orlando, Florida, 1, 609-622. 
Pakiari. A.R and Linnen, J.W. (1975) Applications of a simple molecular ll'anrunction. Pa rt 9. 
Floating spberical GauSJian orbital calculation for hydrogen peroxide, Journal of the 
Chemical Sociely. Faraday Transactions 1I. 71 (9), 1590-1 594. 
Pandit. AB. and Moholkar. V.S. (1996) Harness cavitation to improve processing, Chemical Engineering 
Progress, 92(7). 57 .. 69. 
Pape. B.E. and labile. MJ. (1970) Pbotochemistry of bioactive compounds. Pbotochemistry or selected 
2-<hlon>- and 2 .. methyUbi0-4.6-di(alk~'lamino)-s-triazine berbicides. Journal of Agricultural 
and Food Chemisrry. 18(2), 202-207. 
Paris. D.E and Lewis. D.L. (1973) Chemical and microbial degradation of ten selected pesticides in 
aquatic systems. Residue Reviews. 45.95-124. 
Parke. A v'M. and Taylor. D. (1956) Tbe cbemical action of ultrasonic wa' ·es. Jaurnal of the Chemical 
Sociery. Part IV, (855), 4442-4450. 
Pavia. D.L., Lampman, G.M. and Kriz, G.S. (1979) Chapter 6. Mass spect rometry. In Introduction to 
Specrroscopy, Saunders College Publishing. Philadelphia. ISBN 0 .. 7216-7 119-5, 225 .. 299. 
Payne. P.A ( 1994) Ultrasonic t ransducers: design, constructioD and applications, International Journal 
of Materials and Product Technology, 9(4-6).403-427. 
Peleg. M. (1976) The chemistry or ozone in tbe treatment or water, Water Research, 10(5), 36 1-365. 
References R-25 
Pelizzeni, E.. Carlin. V., Maurino. V ., Minero, C., Dolci, M. and Marchesini, A. (1990) Degradation of 
atrazine in soil through induced photocatalytic processes, Soil Science, 150(2), 523-526. 
Pelizzeni, E ., Minero. C., Pramauro, E., Barbeni, M., Maurino, V. and Tosato, M. (1987) Pbotocatalytic 
degradation of atrazine at ppb levels under solar light and in the presence of Ti<h particles, 
Chimica e /'industria, 69(10). 88-89. 
Perkins, lP. (1990) Chapter 4. Power ultraJOUnd . In Sonochemistry: The Uses of Ultrasound in 
Chemistry. TJ. Mason, eel, Royal Society of Chemistry, Cambridge, ISBN 0-85 186-293-4. 
PeOOns, lP. (I 997a) Personal communication: fax dated 13 May 1997. 
Perkins, J.P. (1997b) Personal communication: fax dated 30 JWle 1997. 
Perkowski, J. , Kos, L. and Ledakowicz, S. (1996) Application of ozone in tertile wutewater treatment, 
Ozone Science & Eng;neering, 18( 1), 73-85 . 
Pesunan. J.M., Engberts, lB.F.N. and de Jong, F. (1994) Sonochemistry: theory and applicationl, 
Journal of the Royal Netherlands Society. 113(12),533-542 . 
petrier. C , David, B. and Laguian. S. (1996) Ultrasonic degradation at 20 and 500 kHz of atrazine and 
pentachlorophenol in aqueous solution: preliminary results. Chemosphere, 32(9). 1709-1718. 
Pettier. C. and Francony, A. (1997) lncideoce of wave-frequency on the reaction rates during ultrasooic 
wastewater treatment, Water Science and Technology, 35(4), 175-180. 
petrier. C., Jeunet, A., Luche, J.-L. and Reverdy, G. (1992) Unexpected frequency effects on the rate of 
oridative processes induced by ultrasound, Journal of the American Chemical Society, t 14(8). 
3 \48-3\50. 
PCtricr. C .. Lamy, M.-F .. Francony. A. . Benahcene, A. and David, B. (1994) Sonochemical degradation of 
phenol in dilute aqueous solutions: comparison of reaction rates at 20 and 487 kHz. Journal 
of Physical Chemistry, 98(41 ). 10514·10520. 
Petrier. C .. Micolle. M .. Merlin. G .. Luche. J .• L. and Reverdy. G. (1992) Characteristics of pentachloro 
phenate degradation in aqueous solution by means of ultrasound, Environmental Science and 
Technologv. 26(6). 1639-1642. 
Peyton. G.R. ( 1990) Chapter 14. Oxidative treatment methods for removal of organic compounds from 
drinking water supplies. In Significance and Treatment of Vo latile Organic Compounds in 
Water Supplies, N.M. Ram, R.F. Christman and K.P. Cantor, eds .. Lewis Publishers lnc .. 
Chelsea. Michigan. ISBN 0 ·87371 -123-8, 313-362. 
Pcyton. G.R. . Huang, F. Y .. Burlcson , J.L. and Glaze. W.H. (1982) Destruction of pollutants in water with 
ozone in combination with ultraviolet radiation. Part 1. General principles and oxidation of 
tetrachloroethylene. Environmental Science and Technology, 16(8), 448-45 3. 
References R-26 
Podolske, J.R and Johnston, H.S. (1983) Rate oftbe resonant energy-transfer reaction between <heaa) 
and BOO-, Journal of Physical Chemistry, 87(4), 628-634. 
Povey, M.J.W. (1989) Ultrasonics in Food Engineering. Part n. Applications, Journal of Food 
Engineering, 9(1), 1-20. 
Prados, M., Paillard, H. and Roche, P. (1995) BydroIyl radical oIidation processes for tbe removal of 
triazine from natural water, Ozone Science & Engineering, 17(2), 183-194. 
Prasad Naidu, DV , Rajan, R. , KUmM, R., Gandhi, K.S., Arakeri, V.H. and Chandrasekaran, S. (1994) 
Modelling of a batcb sonocbemical reactor, Chemical Engineering Science, 49(6), 877·888. 
Price, G.J. (1993) Applications of bigb intensity ultrasound in polymer cbemistry, Chemistry & Industry, 
(3), 75-78. 
Price, GJ. and Lenz, EJ. (1993) The use of dosimeters to measure radical production in aqueous 
sonocbemical systems, Ultrasonics, 31(6), 451-456. 
Price. G.J. , Manhias, P. and Lenz, E.J. (1994) The use of high power ultruound for the destruction of 
aromatic compounds in aqueous solution, Process Safety and Environmental Protection. 
Transactions of the Institu tion of Chemical Engineers, 72(BI ), 27-31. 
Price, G.J. and Smith, P.F. (1993) Ultrasonic degradation of polymer solutions. 2. The effect of 
temperature, ultrasound intensity and dissolved gases on polystyrene in toluene, Ultrasonics, 
34(9), 4111-4117. 
Price, G.J .. Smith, P.F. and West, PJ. (1991) Ultrasonically initiated polymerisation of methyl 
methacrylate. Ultrasonics, 29(3), 166-170. 
Pryor. M.1. and Freese. 5.0. (201X) The treatment of eutrophic water using pre- and intermediate 
IYLOnation, perOIone and pica carbon. WRC report no. 694/ 1/00. ISBN I 86845 588 2. 
Pugin, B. (1987) Qualitative cbaracterization of ultrasound reactors for heterogeneous sonochemistry, 
Ultrasonics. 25(1). 49-55. 
Qiao. X., Ma, L. and Hurnmel. H.E. (19%) Persistence of atrazine and occurrence of its primary 
metabolites in tbree soils, Journal of Agricultural and Food Chemistry, 44(9).2846-2848. 
Quibell. G .. van Vliet. H. and van der Merwe, W. (1997) Characterising cause-and-ef'fect relationsbips in 
support of catchment water quality management. Water £4, 23(3). 193-199. 
Rabie, M.A and Day. J.A (1992) Chapter 25. Rivers. In Environmental Management in South Africa, 
R.F. Fuggle and M.A Rabie. eds .. Juta and Company, Cape To·wn. ISBN 0-7021-2847·3. 
647-668. 
Rachwal, Ai.. Foster, D.M. and Holmes. M. (1992) Combining ozone/advanced oxidation and biological 
fiUration processes for organics remO\-'al from ",'ater, Proceedings of the Water Quality 
Technology Conference. Toronto, Ontario. 1, 542·560. 
References R-27 
Rassokhin. D.N .. Gokzbaev, M.B,. Bugaenko, L.T. and Kovalev. G.V, (1994) Sooolysis of aqueous 
solutions under argon: dependence of the rate of hydrogen peroxide formation on hydro:s:yl 
radical scnenger concentration. MendeJeev Communications, 25-27. 
Rassokhin, D.N. , Kovalev. G.V. and Bugaenko, L.T. (1995) Temperawre effect on the sonolysis of 
methanoVwater mixtures, Journal of the American Chemical Society, 117{l), 344-347. 
Ratoarinoro. N .. Contamine, F., Wilhelm, AM., Berlan, J. and Delmas, H. (l995) Activation of 
solid-liquid chemical reaction by ultruound, Chemical Engineering Science, 50(3). 554-558. 
Ratoarinoro. N. , Wilhclm, AM., Berlan., J. and Delmas, H. (1992) Effects of ultrasound emitter type and 
power on a beterogeneous reaction, Chemical Engineering Journal. 50(1 ). 27·31. 
Raveton. M., Ravanel, P., Kaouadji, M., Bastide, 1. and Tissut, M. (1997a) The cbemical transformation 
of atrazine in corn seedlings. Pesticide Biochemistry and Physiology. SI(3), 199-208. 
Raveton. M., Ravanel, P., Serre, A.M., Nurit, F. and Tissut, M. (1997b) Kinetics of uptake and 
metabolism of atrnine in model plant syfrtems. Pesticide Science, 49(2). 157-1 63. 
Reimers. R.S .. deKemion., P.S. and Leftwich, D.B. (1979) Sonies and electrostatics - an innovath'e 
approacb to water and waste treatment, Proceedings of the Water Reuse Symposium, 2. 
1390-1 416. 
Reisse. J., Francois, H., Vandercammen, J. , Fabre, 0., Kirsch-de Mesmaeker, A . Maerschalk., C. and 
Delplancke, J.-L. (l994) Sonoelectrocbemistry in aqueous electrolyte: a new type of 
sonoelectroreactor. EJecrrochimica Acta, 39(1), 37·39. 
Rice. R. G. (1997) Applications of ozone for industrial waste,,'ater treatment: a review. Ozone Science & 
Engineering, 18(6), 477-515. 
Richards. W.T. and Loomis. A L. (1927) Tbe chemical effect~ of bigb frequenc)' sound waves, L A 
preliminary sun·e)'. Journal of the American Chemical Society, 49(12). 3086-3100. 
Rieke. R.D. and Rhyne, L.D. (1979) Preparation of highly reactive metal powders. Actinted copper 
and uranium. The UlImann coupling and preparation of organometallic sptties. Journal of 
Organic Chemislry. 44(1 9). 3445-3446. 
Riesz. P. and Kondo. T. (1992) Free radical formation induced by ultrasound and its biological 
implication~ . Free Radical Biology and Medicine, U P), 247-270. 
Riesz, P., Kondo. T. and Krishna. C.M. (1990a) Free radical formation by ultrasound in aqueous 
solutions. A spin trapping study, Free Radical Research Communications, 10(1 -2),27-35. 
Riesz. p .. Kondo. T. and Krishna. C.M. (l 990b) Sonochemistry of ,·olatile and non-,'olatile solutes in 
aqueous solutions: e.p.r. and spin trapping studies, Ultrasonics, 28(5), 295-303. 
Robens. R.T. (1993) High intensity ultrasonics in food processing. Chemistry & Industry. (4). 119-1 2 L 
References R-28 
Robinson. S.F. and Monsen, R.M. (199 1) Hydrogen peroxide a.n.d environmental immediate response, 
First International Symposium: Chemical Oxidation, Technology for the 90s, Nashville. 
Tennessee, W.W. Eckenfelder. AR Bowers and l.A Roth, eds .. 51-67. 
Rochc. P. and Prados, M. (1995) Removal of pesticides by use of ozone or bydrogen peroxide/ozone, 
Ozone Science & Engineering, 17(6), 657-672. 
Roche, P., Yolk, C., Carbonnier, F. and Paillard, H. (1994) Water oxidation by ozone or ozone/bydrogen 
peroxide using tbe "ozotest" or "peroxotest" methods, Ozone Science & Engineering. 16(2). 
135-1 55. 
Rocke, AJ. ( 1994) ScMnbein's cbemical goddess, Chemistry & Industry, (2 1), 871 . 
Rodriguez, C.J . and Harldn, J.M. (1991) Degradation of atrazine in subsoil •• and groundwater mixed 
witb aquifer sediments, Bulletin of Environmental Contamination and Toxicology, 59(5) , 
728-735. 
Roi. N.A (1957) The initiation and de"elopment of ultrasonic cavitation, Soviet Physics - Acoustics. 
3(1), 1-16. 
Rooney. J.A (1988) Chapter 2. Otber non-linear acoostic pbenomena. In Ultrasound: liS Chemical. 
Physical and Biological E./ftCIS, K.S. Suslick., ed.. , VCH Publishers, Inc., New York. 
ISBN 0-89573-328-5, 65-96. 
Rosen. H.M. ( 1973) Use of ozone and oxygen in advanced wastewater treatment, Journal oflhe Water 
Pollution Control Federation, 45(12}, 2521 -2536. 
RSA (19%) The Constitution of tbe Republic of Soutb Africa, ISBN 0-620-20214-9. 
Ruppen. G .. B::mcr. R. and Heisler, G. (1994) UV-O.l. UV-H10z. UV-TiOz and tbe pbolo-Feulon 
reaction: comparison of advanced oxidation processes for wastewatcr treatment. 
Chemosphere. 28(8). 1447-1454. 
Sadeghi. K.M .. Lin. J.-R. and Yen, T.F. (1994) Sooocbemical treatment of fossil fuels . Energv Sources, 
16(3). 439-449. 
Sakai. Y .. Sadaoka. Y. and Takamaru, Y. (1977) Decomposition of chloral hydrate in aqueous solution 
b~' tbe action of ultrasound. Journal of Physical Chemistry, 81(6). 509·511 . 
Sakscna. T.K. and Nyborg, W.L. (1970) Sonoluminescence from stable cavitation, Journal of Chemical 
Physics, 53(5).1722-1734 . 
Schmitt, F.O., Johnson. c.H. and Olsoo, AR. ( 1929) Oxidations promoted by ultrasonic radiation . 
Journal of the American Chemical Society. 51(2), 370-375. 
Schulz. C.R. and Prcndiville, P.W. (1993) Designing bigb concentration 07.one contactors for drinking 
,,'ater treatment planu. Ozone Science & Engineering. 15(3). 245·266. 
References R-29 
Schwikkard. G.W. (1995) An investigation of tbe sonocbemical degradation of bydantoin compounds, 
School of Chemical Engineering. MScEng thesis, University of Natal, Durban. 
Schwikkard, S.L. (2000) Personal communication, Sasolburg. 
Sehgal, c., Steer, RP., Sutherland, RG. and Verrall, RE. (1979) Sonoluminescence of argon saturated 
alkali metal salt solutioDS as a probe of acoustic ca"itatiOD, Journal of Chemical Physics. 
70(5). 2242-2248. 
Sehgal, c., Sutherland, RG. and Verrall, RE. (1980a) O ptical spectra of sonolumincscence from 
transient and nable cavitation in water saturated witb various glUeS, Journal of Physical 
Chemistry. 84(4), 388-395. 
Sehgal, c., Sutherland, RG. and Verrall, RE. ( J980b) Sonolumincscence of NO- and N(h-!laturated 
water as a probe of acoustic cavitation, Journal of Physical Chemistry. 84(4). 396-401. 
Sehgal. c., Sutherland, RG. and Verrall, RE. (198Oc) Sonoluminescence intensity as a function of bulk 
solution temperature, Journal of Physical Chemistry. 84(5), 525-528. 
Schgal, C., Sutherland, RG. and Verrall, RE. (198Od) Selective quencbing of species that produce 
sonoluminescence, Journal of Physical Chemistry, 84(5). 529-531. 
Sehgal, C., Vu, TJ., Sutherland, R.G. and Verrall, RE. (1 982) Use of 2~-dipbenyl-l-picrylbydrazyl to 
investigate tbe cbemical bebaviour of free radicals induced by ultruonic cavitation, Journal 
of Physical Chemistry. 86(15), 2982-2986. 
Sehgal. C.H. and Verrall, RE. (1982) A revi"' of tbe electrical bypothesis of !IOnolumincscence, 
UltrasoniCS. 20(1 ), 37-38. 
Sehgal. C.M. and Wang, S.Y. (1981 ) Tbreshold intensities and kinetics of SQnoreaction of thymine in 
aqueous solutions at 10"" ultrasonic intensifies. Journal of the American Chemical Society. 
103(22).6606-66 11. 
Seiler. A . Brenneisen. P. and Green. D.H. (1992) Benefits and risks of plant protection products -
 possibilities of protecting drinking 't\'ater: case at razine. Water Supp(v. 10(2). 3 1-42. 
Sen Gupta. S.K .. Peori. RG. and Wickware, S.L. (1995) Destruction of organic contaminants in 
industrial wastewater. Industria/ Water Treatment. (September/October). 17-24. 35-37. 
Serpone. N .. Terzian. R.. Hidaka, H. and Pelizzetti. E. (1994) Ultrasonic induced dehaJogenation and 
oxidation of 2-, 3- and 4-<:hlorophenol in air~quilibrated aqueous media. Similarities with 
irradiated semiconductor particulates, Journal of PhYSical Chemistry. 98(10). 2634-2640. 
Seybolcl C. A and Mersie. W. (1996) Adsorption and desorption of atrazine, deetbylatrazine. 
deisopropylatrazine, hydroxyatrazine, and metolachlor in two soils from Virginia.. Journal of 
Environmental Quality. 25(6). 1179-11 85. 
References R-30 
Shah, Y.T., Pandit, AB. and Moholkar, v.S. (l999a) Chapter 1. Sources and types of cavitation. In 
Cavitation Reaction Engineering, Kluwer AcadenticIPlenum Publishers, New York, 
ISBN 0-306-46141-2, 1-14. 
Shah, Y.T., Pandit, AB. and Moholkar, V.S. (1999b) Cbapter 2. Cavitation bubble dynamics. In 
Cavitation Reaction Engineering. Kluwer AcademicIPlenum Publishers, New York, 
ISBN 0-306-46141-2, 15-54. 
Shah, Y.T. , Pandit, A.B. and Moho1kar, V.S. (1999c) Chapter 4. Gas·ljquid cavitation chemistry. In 
Cavitation Reaction Engineering, Kluwer AcademicIPJenum Publishers, New York. 
ISBN 0-306-46141-2, 85-153. 
Shah, Y.T. , Pandit, A.B . and Moho1kar, V.S. (1999d) Cbapter 6. Cavitation reactors. In Cavitation 
Reaction Engineering. Kluwer Acadentic/Plenum Publishers, New York, ISBN 0-30646141 -2, 
193-245. 
Shah, Y,T. , Pandit, AB. and Moholkar, V.S. (199ge) Chapter 7. Models for cavitation reactors. In 
Cavitation Reaction Engineering. Kluwer AcademicIPlenum Publishers, New York. 
ISBN 0-30646141·2, 247-275. 
Shah. Y.T. , Pandit, A.B. and Moholkar, V.S. (l999f) Chapter 9. CAV-OX® process. In Cavitation 
Reaction Engineering, Kluwer Academic/Plenurn Publishers. New York. ISBN 0-306-46141 -2, 
313-332. 
Shapir, N. and Mandelbaum, RT. (1997) Atrazine degradation in subsurface soil by indigenous and 
introduced microorganisms, Journal of Agricultural and Food Chemistry, 45(1 1). 4481-4486. 
Shimabukuro. R.H. (1%7) Atrazine metabolism and herbicidal selectivity. Plan! Physiology. 42(9 ). 
1269-1276. 
Shimabukuro. RH .. Frear, O.S .. Swansen, H.R and Walsh, W.e. (1 97 1) Glutathione conjugation: an 
enz~'matic basis for atrazine resistance in corn. Plant Physiologv. 47(1). 10·14. 
Shimabukuro, RH .. Swanson, H.R. and Walsh, W.e. (1970) Glutathione conjugation: atrazine 
detoxication mechanism in corn. Plant Physiology. 46(1 ). 103· 107. 
Shimizu. Y .. Yamamoto, Rand Shlmizu. H. (1989) Effects of ultrasound on dyeing of n~'lon 6, Textile 
Research Journal, 59(11). 684~87. 
Shukla. S.R and Mathur. M.R (1995) Low-temperature ultrasonic dyeing of silk. Journal of the Society 
of D,yers and Colourists, 111(11). 342·345. 
Siddiqui, M.S . (1996) Cblorine-ozone interactions: fonnation of cblorate. Water Research. 30(9), 
2 160-2170. 
Simmon. V.F., Kauhanen. K. and Tardiff. RG. (1979) Mutagenic acti\'it)-' of cbemicals identified in 
drinking water. Developments in TOxicology and Environmental Science. 2, 249-258. 
References R-31 
Singer. P.e. and Zilli. W.R (l97S) Ozonation of ammonia in wastewater. Water Research. 9(2). 127· 134. 
Sirotyuk.. M.G. (1963) Ultruonic cavitation, Soviet Physics-Acoustics, 8(3). 20] -213. 
Sirotyuk, M.G. (1966) Cavitation strength of water and iu distribution of cavitation nuclei, Soviet 
Physics - Acoustics. 11 (3), 318-322. 
Sittig, M., ed.. (1980) Atrazine. In Pesticide Manufacturing and Toxic Materials Control, Noyes Data 
Corporation. Park Ridge. New Jersey. ISBN 0-8ISS-0814-X. 62~7. 
Skipper, H.D., Gilmour, C.M. and Furtick. W.R (1967) Microbial versus chemical degradation of 
atrazine in soils, Proceedings of the Soil Science Society of America, 31 (S), 653-6S6. 
Skipper. H.D. and Volk. V.V. (1972) Biological and chemic&l. degradation of atrazine in three Oregon 
soils, Weed Science. 20(4), 344-347. 
Smith. D.W., Mobammed, A and Finch. G.R (1992) A bench-scale, two-phase reactor of ozone 
treatment: feasibility studies for high strength ~'astes, Ozone Science & Engineering. 14(S). 
381-389. 
Smith, J.M. (1970) Chapter 12. The global rate and laboratory reacton. In Chemical Engineering 
Kinetics, 2nd edition, McGraw-HiIl. New York. Library of Congress Catalogue Card Number 
74-99204, 466-492. 
Solomon. K.R. Baker, D.B., Richards, RP .. Dixon, K.R , Klaine, 5.1., La Point, T. W., Kendall, R.J. , 
Weisskopf. c.P. , Giddings, J.M, Gies)" J.P., et al. (1996) Ecological risk assessment of 
atrazine in North American surface waten. Environmental Toxicology and ChemiStry, 15(1 ), 
31-76. 
Somich. c.l. MuJdoon. M.T. and Keamcy, P.c. (1990) On-site treatment of pesticide waste and rinsale 
using ozone and biologically acth'e soil. Environme{Jtal Science and Technology. 24(5 ). 
745-749. 
Sonic Systems (1994) High intensity ultrasonic liquid process systems instruction manual. 
Sorbios (1993) Ozone generator GSG instruction manual . 
Sorenson. B.A. Koskinen. W.C .. Buhler. 0 .0 .. Wyse, D.L.. Lueschen. W.E. and Jorgenson.. M.D. (1994) 
Formation and mo\-'ement of I"C_atrazine degradation products in a day loam soil in the 
field , Weed Science. 42(2). 6 18-624. 
Sorenson. B.A. , Koskinen. W.C .. Buhler. 0 .0., Wyse, D.L., Lueschen. W.E. and Jorgenson. M.D. (1995) 
Fate of 14C_atrazine in a silt loam soil. Internalional Journal of Environmental Ana~ytical 
Chemistry. 61(1). 1-10. 
Sorenson. RA .. Wyse. D.L. , Koskinen. W.C .. Buhler. 0 .0 .. Lueschen, W.E. and Jorgcnson, M.D. (1993) 
Formation and movement of 14C_atrazine degradation products in a sandy loam soil under 
field conditions. Weed Science. 41 (2). 239-245. 
References R-32 
Solelo. J.L., BeIlT1n, FJ., Benilez, F.J. and Beltr.\n-Heredia, J. (1987) Ozone decomposition in water: 
kinetic study, Industrial and Engineering Chemistry Research, 26(1), 39-43. 
Spinks, J.W.T. and Woods, RJ. (1976) Cbapter 7. Water and aqueous solutions. In An Introduction to 
Radiation Chemistry. 2nd edition, John WHey & Sons, NC\'!' York, ISBN 0-471-81670-1 , 
247-359. 
Srinivasan, R , Shirgaonkar, I.Z. and Pandit, A.B. (1995) Effect of !JOnieation on crystal properties. 
Separation Science and Technology. 30(1 0). 2239-2243. 
S13ehelin. J., Btihler, R.E. and Hoigne, J. (1984) Ozone decomposition in water studied by pulse 
radiolysis. 2. 80· and ·80 .. as cbain intermediates, Journal of Phys;cal Chemistry, 88(24), 
5999-6004. 
Staehelin. J. and Hoigne, J. (1982) Decomposition of ozone in water: rate of initiation by hydroxide ions 
and bydrogen peros.ide, Environmental Science and Technology. 16(10), 676-681. 
S13ehelin. J. and Hoigne, J. (1985) Decomposition of ozone in water in tbe presence of organic solutes 
acting as promoters and inbibiton of radical cbain reactions. Environmenral Science and 
Technology, 19(12), 1206-1213. 
S13lsoft, Inc. (2000) STAT/ST/CA, hnp:\\www.statsoft.com. 
Stojanovic. BJ .. Kennedy, M.V. and Shuman, F.L. (1972) Edapbic aspects of the disposal of unused 
peJticides. pesticide wastes, and pesticide containen, Journal of Environmental Quality. 1( 1). 
54-62. 
Stratton. G.W. (1984) Effects of the herbicide atrazine and its degradation products. alone and in 
combination, on phototropbic microorganisms. Archives of Environmenral Conramination and 
Toxicology, 13(1).35-42 . 
Streng. A.G. (196 1) Tables of ozone properties. Journal of Chemical and Engineering Data, 6(3). 
431-436. 
Stucki. G .. Yu. C.W .. Baumgartner, T. and Gonzalez-Valero. IF. (1995) Microbial atrazine 
mineralisation under carbon limited and denitrifying conditions. Water Research, 29(1 ). 
291 -296. 
Suslick. K.S . (1989) The chemical effects of ultrasound, Scientific American. 260(2), 62-68. 
Suslick. K.S. (1990) Sol'lochemistry. Science. 247(4949). 1439-1445. 
Suslick. KS. and Casadonte. D.J. (1987) Heterogeneous sonocataJysis with nickel po\o\'de r, Journal of the 
American Chemical Society, 109(11).3459-3461. 
Suslick. KS .. Casadonte, DJ. and Doktycz , S.J. (1989) The effects of ultrasound on nickel and copper 
po\o\'den. Solid State Ionics. 32-33(1). 444-452. 
References R-33 
Suslick, K.S. , Doktycz, S.J. and Flint, E.B. (1990) On tbe origin of sonoluminescence and sonocbemistry. 
Ultrasonics, 28(5). 280-290. 
Suslick. K.S. , Gawienowski, J.J ., Schubert, PT and Wang, H.H. (1983) Alkane sonochemistry, Journal of 
Physical Chemistry, 87(1 3), 2299-2301. 
Suslick. K.S., Gawienowsld, 1.1., Schubert, P.F. and Wang, HH (1984) Soooc::hemistry in non-aqueous 
liquids, Ultrasonics, 22(1 ), 33-36. 
Suslick, K.S., Goodale, J.W., Schubert, P.F. and Wang, H.H (1983) Sonochemisb:y and sonocataJysis of 
metal carbonyls, Journal of the American Chemical Society, 105(18), 5781-5785. 
Suslick, K.S. and Hammerton, D.A (1986) The site of sonochemical reactions, IEEE Transactions on 
Ultrasonics, Ferroelectrics and Frequency Control, UFFC-33(2), 143-147. 
Suslick. K.S. , Hammerton, D.A and eline, RE. (1986) The sonochemicaJ hot spot, Journal of the 
American Chemical Society, 108(18),5641-5642. 
Suslick, K.S. and Johnson, RE. (1984) Sonochemical activation of transition metals, Journal of the 
American Chemical Society, 106(22). 6856~858. 
Suslick. K.S. and Kemper, K.A. (1993) The effect of nuorocarboo gases on sonoluminescence: a failure 
of the electrical hypothesis, Ultrasonics, 31(6),463-465. 
Suslick. K.S. , Schubert, P.F. and Goodale, J.W. (1981) Sonochemistry and sonocatalysis of iron 
carbonyl!, Journal of the American Chemical Society, 103(24). 7342-7344. 
Taube. H. and Bray, W.e. (1940) Cbain reactions in aqueous solutions containing ozone, hydrogen 
peroxide and acid, 62. 3357-3373. 
ler Haar. G.T. (1988) Cbapter 8. Biological effects of ultrasound in clinical applications. In Ultrasound: 
Its Chemical. Physical and Biological Effects, K.S. Suslick. ed .. VCH Publishers. New York.. 
ISBN 0-89573-328-5, 305-320. 
Thakore. K.A. (l990a) Pbysico-cbemical study on applying ultrasonics in textile dyeing. Part I. 
American DyestujJReporrer, 79(5), 45-47. 
Thakore, K.A. (l990b) Physico-cbemical study on applying ultrasonics in tenile dyeing. Part 11. 
American Dyestuff Reporter, 79(6), 38-43. 
Thoma, G .. Gleason. M. and Popo\', V. (1998) Sonochemical treatment of benzeoeitQluene cootaminated 
wastewater, Environmental Progress, 17(3), 154-160. 
Thomas, lR and de Vrics, L. (1959) Sonically induced beterolytic cleavage of polymetbylsiloxane, 
Journal of Physical Chemistry, 63(2). 254·256. 
Thompson, L.H. and Doraiswamy. L.K. (1999) Sonochemistry: science and engineering, Industrial and 
Engineering Chemistry Research, 38(4). 1215-1249. 
References R-34 
Thurman, E. M., Goolsby, D.A. , Meyer, M.T. and Kolpin, D.W. (1991) Herbicides in surface waters of 
tbe midwesteru United States: tbe effect of spring flush, Environmental Science and 
Technology, 25(10), 1794-1796. 
Thunnan, E.M., Goolsby, D.A. , Meyer, M.T. , Mills, M.S. , Pomes, M.L. and Kolpin, D.W. (1992) A 
reconnaissance study of berbicides and their metabolites in surface water of tbe mid",·estern 
United States using immunoassay and gas cbromatograpby/mass spectrometry, 
Environmental Science and Technology, 26(12), 2440-2447. 
Thwman, E.M., Meyec, M.T., MiUs, M.S., ZiIrunennan, L.R., Pe"),, CA and Goolsby, DA (1994) 
Formation and transport of deetbylatrazine and deisopropylatrazine in surface water, 
Environmental Science and Technology, 28(13), 2267-2277. 
Tiebm, A., Nickel, K. and Neis, U. (1991) The use of ultrasound to accelerate tbe anaerobic digestion of 
sewage sludge, Water Science and Technology, 36(11), 121-1 28. 
Tkacbuk,. N.G., Kravets, V.V., Nikitin, G.A. and Semenova, E.l. (1 990) Intensification of tbe activity of 
activated sludge microorganisms with uJtrasound, Sovier Journal of Waler Chemislry and 
Technology. 11(6), 82-85. 
Todd.. J.H. (1970) Measurement of cbemica1 activity of ultrasonic cavitation in aqueous solutions, 
Ultrasonics, 8(4), 234-238. 
Tomlin. C.D.S., ed (1 997) Atranne. In The Pesticide Manual, 11th edition, British Crop Protection 
Council, Famham, Surrey. ISBN I 901396 11 8, 55-57. 
Torrents. A., Anderson., B.G., BilbouJian. S., Johnson, W.E. and Hapeman, c. J. (1 997) Atrazine 
pbotolysis: mecbanistic investigations of direct and nitrate-mediated bydroxyl radica1 
processes aod tbe innuence of dissoh'ed organic carbon from tbe Cbesapeake 8a)o'. 
Environmental Science and Technology, 31(5), 1476-1482. 
Trapido, M. , Hirvoncn, A. . Yeressinina. Y. , Hcntunen, J. and Munler, R. (1997) Ozonation. ozone/UV and 
UVIB10 1 degradation of cblorophenols, Ozone Science & Engineering, 19(1),75-96. 
Upham. S .L.. BOOdy, 8., Xing. X .. Trosko. lE. and Masten, S.J. (1997) Non-genotoxic effects of selected 
pesticides and tbeir disinfection byproducts on gap juoctionaJ intercellular communication, 
Ozone Science & Engineering, 19(4), 35 1-369. 
van der Mcrwc. S.W. (1 995) An o\-'eniew of water supply management in Soutb Africa, Journal of 
Water Supply Research and Technology -Aqua. 44(4), 151-160. 
van Leeuwen. HJ . (1987) Preliminary investigation into the impro\-'ement of tbe biodegradability of 
organic substances in surface waters and effluents tbrough ownarioD. Water Science and 
Technology, 19. 93 1-937. 
References R-35 
van Rensburg, J.F.J., Hassetl, A , Theron, S. and Wiechers, S.O. (1981) The fate of organic 
micropollutants througb an integrated wastewater treatment/water reclamation system, 
Water Science and Technology, 13(1), 537-552. 
Vanderheyden, V., Debongnie, P. and Pussemier, L. ( 1997) Accelerated degradation and mineralisation 
of atrazine in surface and subsurface lOil materiaJs, Pesticide Science, 49(3), 237-242 . 
Vanysek, P. (1998) Electrocbemical series. In CRC Handbook a/Chemistry and Physics, D.R. Lide, 00. , 
79th edition. CRC Press. ISBN 0-84934179-2.8121-8/34. 
Vaughan, P.W. and Lecman, S. (1989) Acoustic cavitation revisited, Acustica, 69(3), 109-119. 
Verrall. R.E. and Sehgal. C.M. (1988) Cbapter 6. Sonolumincscence. In Ultrasound: ils Chemical, 
Physical and BioJog;caJ Effects, K.S. Suslick, eel., VCH PublishCIS, New York, ISBN 
0-89573-328-5, 227-286. 
Vining, G. G. (1998) Statistical Methods /01' Engineers, Brooks/Cole Publishing Company, Pacific Grove. 
Florida. USA. ISBN 0-534-23706-1. 
Vogel. A.I. (196 1) A Text-book a/Quantitative inorganic AnalysiS, 3rd edition, Longman Group Limited, 
London. ISBN 0-582-44247-8. 
Yolk. c., Roche, P .. Jorel, J.-C . and Paillard, H. (1997) Compa rilOn of the effect of ozone, 
ozone-hydrogen peroxide syl1em and catalytic ozone OD the biodegradable organic matter of 
a fulvic acid solution, WaleI' Research, 31(3), 650-656. 
Volskay. V.T. and Grady, C.P.L. (1988) Toxicity of selected RCRA compounds to activated sludge 
microorganisms, Journal o/the Water Pollution Control Federation , 60(10), 1850-1 856. 
Volskay. VT .. Grady. C.P.L. and Tabak. H.H. ( 1990) Effect of selected RCRA compounds on acth'ated 
sludge activit)'. Research Journal o/the Water Pollution Control Federation, 62(5),654-664. 
von Sonnlag. C. ( 1996) Degradation of a romatics by adunced oxidalion processes in wale r 
remediation: some basic considerations, Journal 0/ WaleI' Supply Research and Technology -
 Aqua. 45(2). 84-91. 
von Sonntag, C . Dowidcit. P., Fang , X., Mertens, R. , Pan, x.. Schuchmann. M.N. and Schuchmann. H. -P. 
(1997) The fale of peroI~' 1 radicals in aqueous solution. WaleI' Science and Technology, 35(4). 
9-15. 
Vorstcr. I. (200 1) Persona] communi cation: email dated 2 February 2001. 
Wakeman. R.J. and Tarleton. E.S. (1991) An eIperimeotaJ stud)' of electroacoustic crossflow 
microfiltration. Transactions o/the Institute o/Chemical Engineering, 69(A). 386-397. 
Walms lcy, U . (1995) Market forces and tbe management of water for the environment. WaleI' SA, 
2\ (1).43-50. 
References R-36 
Walton.. DJ .. Burke. L.D. and Murphy. M.M. (1996) Soooelectrocbemistry: cblorioe, bydrogeo and 
osygeo evolution at platinised platinum, £Iecrrochimica Acla, 41(17), 2747-2751. 
Weavers. L.K. and Hoffmann. M.R (1998) Sonolytic decomposition of ozone in aqueous solutioo! mass 
transfer effects. Environmental Science and Technology, 32(24), 3941-3947. 
Weber, J.B. (1970) Mechanism! of adsorption of s-triazines by clay colloids and factors affecting plant 
availability. In Residue Reviews, F.A. Gunther and lD. Gunther, eds., Springer-Verlag, New 
York. Libr3I)' of Congress CataJogue Card Number 62-1 8595, 32, 93-130. 
Webster. E. (1963) Cavitation, Ultrasonics, 1(1), 39-48. 
Weissler. A. (1959) Formation of bydrogeo peroxide by ultrasonic waves: free radicals, Journal 0/ the 
American Chemical Society, 81(5}, 1077-1081. 
Weissler. A., Cooper, H. W. and Snyder, S. (1950) Chemical effect of ultrasonic ",'nes: oxidation of 
potassium iodide solution by carbon tetracbloride, Journal o/the American Chemical Society, 
74(4). 1769-1 775. 
WeLhouse. GJ. and B1eam, W.F. (1993a) Atrazine bydrogen-bonding potentials, Environmental Science 
and Technology, 27(3). 494-500. 
Welhouse, G.J. and SIeam, W.F. (l993b) Cooperative bydrogen bonding of atrazine, Environmental 
Science and Technology, 27(3). 500-505. 
Welty. J.R. Wicks, C.E. and WilsoD, RE. (1984) Fundamentals o/Momentum, Heat and Mass Trans/er, 
3rd edition. John Wiley and Sons, New York, LibraI)' of Congress CataJogue Card Number 
83-17065, 
Weng. C.-N., Hoven. D.L. and Schwartz. BJ. (1986) Ozonation: an economic choice for water 
treatmen •. Journal o/ the American Water Works Association, 78(11). 83-89. 
West. R . Fink. MJ. and Michl. 1. (1981) Tetramesityldisilene. a stable compound containing a 
silicon-silicon double bond, Science, 214(4527). 1343-1344. 
Westerhoff. P .. Song, R., Amy, G. and Minear. R (1997) Applications of O'.tone decomposition models. 
Ozone Sc,ence & Engineering, 19(1).55-73. 
Wilber. G.G. and Parkin. G.F. (1995) Kinetics of alachlor and atrazine biotransformation under nrious 
electron acceptor conditions. Environmental Toxicology and Chemistry. 14(2), 237-244. 
Winkelmann. D.A. and Klaine. SJ . (1991a) Degradation and bound residue formation of atrazine in a 
western Tennessee soil. Environmental Toxicology and Chemistry. 10(3), 335-345. 
Winkelmann. D.A and Kiaine. S.J. (l991b) Degradation and bound residue formation of four atrazine 
metabolites. deethylatrazine, deisopropylatrazinc, dealk)'latrazine and bydrosyatrazinc, in 
a western Tennessee soil. Environmental Toxicology and Chemistry, 10(3). 347-354. 
References R-37 
Winshlp, S. (1999) Evaluation of different methods to produce free radicals for the oxidation of 
organic molecules in industrial efl1uents and potable water with special reference to 
CAV-OX®. WRC report no. 388/1 /99, ISBN 1 86845484 3. 
Wolf. D.e. and Martin, JP. (1975) Microbial decomposition of ring_14C atrazine, cyanuric acid, and 
2-<:.hlor0-4,5-diamino-s-triazine, Journal of Environmental Quality, 4(1), 134-1 39. 
Xiong, F. and Graham. NJ .D. (1992) Rate constants for berbicide degradation by ozone, Ozone Science 
& Engineering, 14(4), 283-301. 
Young, F.R (1976) Sonoluminescence from water containing dissolved gases, Journal of the Acoustical 
Society oJ America, 60(1), 100-104. 
Young, F.R (1989) Chapter 1. Introduction . In Cavitation, McGraw-Hill, London, ISBN 0-07-707094-1 , 
1-7. 
Yue, P.L. (1992) Degradation of organic pollutants by advanced oxidation, Transactions of the Institute 
o/Chemical Engineering. 70(B), 145-148. 
Vue, P.L. (1997) Oxidation reactors for water and wastewater treatment, Water Science and 
Technology. 3S(4}, 189-1 96. 
Zaidi. S.A.H. (1993) Ultrasonically enhanced coal desulphurisation, Fuel Processing Technology. 33(2). 
95-1 00. 
ZechmeiSler. L. and Wallcave, L. (1955) On tbe cleavage of benzene. thiophene and (uran rings by 
means of ultrlUOnic waves, Journal of the American Chemical Society, 77(10), 2853-2855. 
Zwiener, C , Weil. L. and Niessner. R (1995) Atrazine and parathion-metbyl removal by UV and 
UV/Ol in drinking water treatment, International Journal oJ Environmental Ana~vtical 
Chemistry, 58(1-4). 247-264. 
APPENDICES 
A 
EQUIPMENT CHARACTERISATION 
The experiments performed for equipment characterisation are described in this Appendix. Ultrasonic 
experiments are described in Section A 1 and ozone experiments in Section A2. 
A.I ULTRASOUND EQUIPMENT 
Experiments relating to the operation of the u1trasonic born, the evaluation of machining the eroded surface 
of a horn tip and the measurement of acoustic power, are described, respectively, in Section A.l.l and 
Section A 1.2. The investigation of the effect of volume on hydrogen peroxide formation with the low 
intensity ultrasonic horn is presented in Section A 1.3. Acoustic power reduction due to the original design 
of the ultrasonic cell lid is described in Section A 1.4. Diagrams illustrating the mixing in the ulttasonic 
cell are presented in Section A 1.5 and the calibration of the rotameter for gas experiments is presented in 
Section A 1.6. 
A.l.1 Evaluation of machining the surface of 8 horn tip 
The effect of machining the surface of a horn tip on the performance of the ultrasonic system was evaluated 
by measuring the formation of hydrogen peroxide generated by a new tip and comparing it to that generated 
after machining. Hydrogen peroxide concentration was measured after a 20 min sonication period al 
transducer displacement amplitudes of 2; 4; 6; 8; 10; 11 and 12}llll. Experiments were pcrfonned in 
triplicate. The comparison of hydrogen peroxide fonnation generated using a new and 3 machined horn tip 
is presented in Figure A.I . 
Appendb: A EQUIPMENT CHARACTERISATION A-2 
0 ,16 
::; l E 0 ,12 
c ~ 0 
" l! 0,00 L • new tip 
• I g " machined tip 
0 t 0 
N 2 0,04 r-
 :z: 
> I 
0 
0 2 4 • • 10 12 14 
Transducer displacement amplitude (IJm) 
Figure A.t : Hydrogen peroxide formation witb a new and a macbined born tip (n = 3) 
The data presented in Figure A 1 is recorded in Table AI . 
Table A.t: Hydrogen peroxide formation with a new and a macbined horn tip 
Transducer Hydrogen peroxide concentration 
vibration (mg L-1) 
u.m) n~' homtip machined horn tip 
0 0,000 0,000 0,000 0,000 0,000 0,000 
2 0,018 0.024 0,021 0,015 0,026 0,018 
4 0.041 0,032 0,050 0,041 0,050 0,044 
6 0.053 0.071 0,062 0.073 0.065 0.065 
8 0,073 0.082 0,082 0,079 0.082 0.076 
10 0.094 0.099 0.094 0.097 0,094 0,097 
11 0, 106 0,097 0.100 0.103 0.106 0.112 
12 0.106 0.106 0,112 0.115 0.109 0.106 
Statistical analysis of the data presented in Figure A 1 is reported in Table F. l in Appendix F, The 
maximum standard deviation in the hydrogen peroxide concentrations measured during sonication with the 
new and machined horn tips are 0.009 and 0.006 mg L-1, respectively. The rate of hydrogen peroxide 
fonnation during sonication with a new and a machined horn tip over 20 min were calculated from the 
regression of the data presented in Figure A.I using the linear regression model 
y = bx lA. 11 
where the coefficient b, the gradient of the regression line. represents the rate of hydrogen peroxide 
formation. The calculated rates of hydrogen peroxide formation for the new and machined horn tips (and 
the standard errors ofthe calculated "alues) are recorded in Table A.2 . 
Appendix A EQUIJ'~U:!'I" CIlARACTERlSATIOX A-3 
Table A.2: Rate of bydrogen peroxide formation during sonication ""itb a new and a macbined 
born tip over 20 min 
new horn tip 
machined horn tip 
Rate of B1C),: formation 
(mg L ol min·l ) 
0,0094 ± 0,0002 
0,0097 ± 0,0002 
0,978 
0,977 
The statistical F-test at a 95 % confidence level indicated that the rates of hydrogen peroxide formation 
during sonication with a new and a machined horn tip were not significantly different. 
A.l.2 Acoustic power measurement 
The efficiency of the energy transformation of the ultrasonic horn was determined by comparison of acoustic 
power with electrical power. Acoustic power is the energy delivered into a reaction solution. electrical 
power is the energy drawn at the wall. The electrical power supplied to the transducer of the ultrasonic 
horn was read from the meter on the face of the ultrasonic generator when the transducer was loaded. The 
acoustic power was calculated, as recommended by the manufacturer, from the difference in electrical power 
supplied to the transducer when loaded and wilhoutload (Sonic S)'stems. 1994). This method was also used 
to calculate acoustic power in the investigation of ultrasonic chemical and physical effects by Ratoarinoro 
and co-workers (Contamine et aI. , 1994; Ratoarinoro et al .• 1995). 
The calculated acoustic powers during sonication at transducer displacement amplitudes of 5, 8 and 11 fJ.m 
are recorded in Table A.3 . The electrical power drawn by the transducer when unloaded was 4. 7 and 
12 W. respectively. for sonication at transducer displacement amplitudes of 5. 8 and 11 fJ.m. 
Table A.3 : Calculated acoustic power during the sonication of SOO mL of water at transducer 
displacement amplitudes of 5,8 and 11 fJ.m 
Time Acoustic power (W) 
(min) 5"m 8"m 11 }J.m 
0 24.5 24.3 25 ,0 40,0 40,0 39.5 59.5 57,0 56,5 
I 24.5 24,2 24.5 39.5 39.5 39.2 59.3 57,0 56.5 
2 24.5 24.2 24.5 39.5 39,5 39.0 59,0 57.0 56.3 
3 24.3 24,0 24.5 39.5 39,3 39.0 59.0 57,0 56,0 
4 24.3 24.0 24,5 39.5 39.2 39.0 58.8 56.5 56.0 
5 24,3 24.0 24.5 39,5 39,2 39,0 58.5 56,5 55.5 
6 24.2 24,0 24.3 39.5 39.0 39.0 58.3 56,2 55.0 
7 24,2 24.0 24.3 39,5 39,0 39,0 58,2 56,0 54,S 
8 24.2 24.0 24.2 39.3 39.0 39,0 58,0 56,0 54.5 
9 24.0 23 .8 24,1 39.5 39.0 39.0 57,8 55.0 54,0 
10 24.0 23.6 24.0 39.3 39,0 38,8 57,5 55.0 54.0 
Appendix A EQ1.ilPM1:!\I'T CBARACT£R1SAnO~ A-4 
Statistical analysis of the data recorded in Table A3 is reported in Table F.2 in Appendix F. The maximum 
standard deviation in the measured acoustic powers at displacement amplitudes of 5, 8 and 11 IJ.m are 0,36; 
0,29 and 1,97 W, respectively. 
The average acoustic power measured during sonication at each transducer displacement amplitude is 
compared in Table A4 with the average electrical power. The energy conversion in the ultrasonic hom is 
approximately 85 % (fable 8.4). 
Table A.4: Con,·ersion of electrical power to acoustic power 
Transducer A,·erage acoustic Average electrical Energy 
displacement power power con,'crsioD 
u.m) (W) (W) (%) 
5 24,2 28,2 86 
8 39,3 46,3 85 
II 56,7 68.7 83 
Calorimetry was used to confirm the calculated acoustic p:>wers. Calorimetry is based on the assumption 
that all the acoustic energy entering a system is converted into heat energy, the resulting temperature 
increase of the system being directly proportional to the amount of energy entering the system (Delchar and 
Melvin, 1994; Gosv.'anU et al., 1988). The acoustic power delivered to a reaction solution was calculated 
using equation A2. 
Q = m Cpc;{, (A.2( 
where Q is the acoustic power, m the mass of the water sonicated. Cp the heat capacity of the water and 1fi 
the rate of temperature increase of the water at time zero (Welty et aL 1984). The use of calorimetrJ to 
determine acoustic power has been demonstrated by Mason and co-workers in a review of parameters that 
affect sonochemistf)· and by Lorimer and co-workers in a investigation of the ultrasonic decomposition of 
potassium persulphate (Lorimer et al.. 1991 : Mason et al .. 1992). CalorimelI)' is frequently used in 
ultrasonic investigations to calculate acoustic power (Section 2.5.5). 
Temperature profiles were measured in 500 mL of water in an insulated plastic beaker during sonication 
with the ultrasonic horn at transducer displacement amplitudes of 5. 8 and 11 Jlm. A schematic diagram of 
the insulated vessel is shown in Figure A2. 
Appendix A EQUlPME!'ro'T CIlARACTERlSATION 
ultrasonic horn 
polysterene pieces 
water 
thennometer 
plastic beaker 
wrapped with 
insulation 
Figure A.2: Scbematic diagram of the insulated vessel used in calorimetric experiments 
A-5 
Temperature was measured during sonication with a spirit-filled 76 mm immersion thennometer every 
minute from 0 to 10 min, every 2 min from 10 to 20 min and every 5 min from 20 10 40 min. Temperature 
profiles recorded during sonication at transducer displacement amplitudes of 5, 8 and 11 ~ are shown in 
Figure A.3. Experiments were perfonned in triplicate. 
100 
BO -
 G' • 
'-- • 
S w - • • • • • • • 
- • • • • 0- 40 • • • • • E • • • ~ .... tti .' • • >- • • • •• 
20 
... 5 IJm 
• BlJm 
• 11 IJm 
o ------__________________________ ~ 
o 10 3J 50 
Time (min) 
Figure A.3: Temperature measurements in SOO mL of ,nter in an insulated ' "essel during 
sonication over 40 min at transducer displacement amplitudes of 5, 8 and 11 ).lm (n = 3 time 
period) 
The data presented in Figure A.3 is recorded in Table A.5. 
Appendix A EQUIPMENT CHARACTERISATION A-<i 
Table A.S : Temperature meaJU.remenU in SOO mL of water in an insulated vessel during 
sonication m'er 40 min at transducer displacement amplitudes of S, 8 and 11 I-Lm 
Time Temperature (0C) 
(min) 5~m 8~m 1) I-Lm 
0 23,7 23,7 23,8 23,8 24,0 23,8 23,0 23,4 24,0 
24,3 24,2 24,0 24,6 24,9 24,3 24,3 24,7 25,2 
2 24,9 24,8 24,7 25,7 25,9 25,2 25,6 26,0 26,7 
3 255 25,5 25,3 26,7 26,9 26,2 27,0 27,4 28,0 
4 26,0 25,9 26,0 27,7 27,9 27,1 28,4 28,8 29.5 
5 26,6 26,6 26,6 28,6 2S,S 2S,I 29,9 30,2 30,S 
6 27,2 27,1 27,1 29,5 29,S 29,0 3 1,3 31 ,7 32,1 
7 27,S 27,7 27,8 30,4 30,8 30,0 32,7 33,0 33,5 
S 28.4 2S,2 2S,3 31,3 31,7 31,0 34,1 34,4 34,S 
9 29,0 2S,9 28,9 32,3 32,6 32,0 35,5 35,S 36,2 
10 29,6 29,4 29,3 33,2 33,5 32,S 36,S 37,2 37,6 
12 30,6 30,3 30,1 35,2 35,4 34,7 39,7 39,9 40,2 
14 31.6 31,2 31 ,0 37,0 37,2 36,5 42,3 42.5 42,9 
16 32,4 32,0 3 1,S 3S,9 39,0 3S,3 44,9 45,1 45.4 
18 33,1 32,7 32,6 40,7 40,9 40,2 47,4 47,7 47,8 
20 33,8 33,3 33,3 42,5 42,7 42,0 50,0 50,1 50,2 
25 36, 1 35,2 35,3 46,S 46,S 46.2 55,8 56,0 56,0 
30 38,1 37, I 37,2 50,S 50,9 50,3 61 ,3 61.5 6 1,3 
35 39,9 38,9 39, 1 54,7 54,8 54.1 66,3 66,5 66,3 
40 41.6 40,5 40,8 58,3 58,3 5S,2 70,8 70,8 70,8 
Statistical analysis of the data presented in Figure A3 is reponed in Table F.3 in Appendix F. The 
maximum standard deviation in the temperatures measured during sonication at transducer displacement 
amplitudes of 5. 8 and 11 )J.ffi are 0,57; 0,41 and 0,56 °C, respectively. 
Acoustic power was calculated using equation A.2 for sonication at transducer displacement amplitudes of 
5. 8 and 11 j.J.m. The rate of temperature increase ('[f; ) was dctcnnined from regression over the 0 to 
10 min temperature were the temperature increase was linear. Density and heat capaci~' were calculated at 
the average temperature of each experiment. The calculated acoustic power using calorimetry is compared 
in Table A6 y,'ith the values calculated from the difference in electrical power supplied to the transducer 
when loaded and unloaded. 
Appendix A EQUlPMEJIo'T CHARACTERISATION A-7 
Table A.6: Comparison of acoustic power calculated from calorimetr), with acoustic po1\'er 
calculated from tbe power drawn by the traosducer wben loaded and unloaded 
Transducer Acoustic power Acoustic power from Difference 
displacement from calorimetry electrical method (%) 
(><m) (W) (W) 
5 20,2 24.2 \7 
8 32,8 39,3 17 
11 47,9 56,7 16 
Acoustic power calculated. from calorimetry was similar though lower (approximately 17 %) than the 
acoustic power calculated from the power drawn by the ultrasonic transducer when loaded and un10aded 
The lower value calculated using calorimetry can be due to not all acoustic energy being convened into heat 
(some energy is lost due to sound) and that not all the energy convened to heat was measured. The energy 
used in heating the plastic beaker is not taken into account in equation A2, also, energy may have been lost 
to the atmosphere despite the insulation cladding and polystyrene pieces. 
Bhatnagar and Cheung noted during the investigation of the ultrasonic degradation of chlorinated volatile 
organic compounds that although the W-385 sonic:ator was capable of delivering 475 W, when set at the 
maximwn output it delivered only 40 % (200 W) of the rated acoustic power. the ultrasonic energy was not 
all convened to heat energy and may also have been convened to sound (Bhatnagar and Cheung, 1994). 
Petrier and co-workers noted during the investigation of the ultrasonic degradation of phenol that the 
acoustic power estimated by the calorimetric method was approximately 50 % of the electrical power 
supplied to the transducer (Petrier et al .. 1994). 
A modification of equation A.2 to account for the heat absorbed by the reaction vessel has been reponed in 
literature (Thompson and Ooraiswamy. 1999). The modified equation is 
Q=mCp ar dl IA31 
where A .. , is the area of the wetted surface of the vessel. x ... the thickness of the inner wall, CJ)\"aK1 the heat 
capacity of the vessel and T,. the temperature of the inner vessel wall (Thompson and Doraiswamy. 1999). 
Acoustic power reponed in this investigation is calculated from the power drawn by the transducer when 
loaded and unloaded as there is a greater uncertainty in the value calculated using calorimetry. 
Appendix A EQUIPMENT CHARACTERISATION A-ll 
A.1.3 Volume experiments with the low intensity ultrasonic horn 
The low intensity ultrasonic horn was used to investigate the effect of sample volume on the formation of 
hydrogen peroxide during sonication. Hydrogen peroxide formation was measured in water volumes of 100: 
200: 300: 400; 500; 600; 700; 800; 900 and 1 000 mL at time periods of 0; 2; 4; 6; 8; 10: 15 and 20 min. 
The transducer displacement amplitude was set at 11 f.UTl and the experiments were performed in duplicate. 
The formation of hydrogen peroxide in different volumes of water is shown in Figure A4. 
0,. 
l 
'il, o,s 
• • 100 mL 
E • • 200 mL 
~ 0,4 - .300 mL 
0 • 0 400 mL 
" 
t t 
 0,3 - It. 500 ml c • • t ~ • lJ. 600 ml 
0 • • • o 700 ml 0 0,2 e N • • e • Q 800 mL 0 ~ N • ~ : i • 900 mL '" 0,1 i • f • I , ! ~ c o 1000 mL ~ -0 • 
0 • • 12 1. 2. 
Time (min) 
Figure AA: Hydrogen peroxide concentration in water volumes of 100 to ) 000 mL during 
sonication witb tbe low intensity ultrasonic born (n = 2 per time period) 
Hydrogen peroxide concentration. shown in Figure A.4, increased with time and decreased with increasing 
water volume. Thc data presented in Figure A,4 is recorded in Table A. 7. 
Table A.7 : Hydrogen peroxide concentration ;n water , 'olumes of 100 to I 000 mL during 
sonicat ion with t.he low intensily ultrasonic born 
T ime Hydrogen peroxide concentration (mg L' I) 
(min) 100 mL 200 mL 300 mL 400mL SOOmL 
0 0,00 0,00 0,00 0,00 0,00 0.00 0,00 0,00 0,00 0,00 
2 0. 11 0. 10 0,77 0,06 0,04 0,04 0,04 0,05 0,02 0.02 
4 0, 19 0,19 0, 11 0, 12 0,07 0.08 0,07 0,07 0,03 0.04 
6 0.28 0,26 0 ,15 0,15 0, 11 0,09 0.08 0,10 0.Q7 0.06 
8 0.32 0.34 0, 19 0,19 0.13 0.14 0.11 0,11 0.08 0.07 
10 0,37 0,37 0,23 0,23 0, 15 0,14 0,13 0.13 0,10 0,10 
15 0.45 0.47 0.28 0,26 0.21 0,19 0.17 0. 18 0.14 0,14 
20 0,55 0.56 0 ,34 0,32 0,23 0.23 0,21 0,20 0,17 0,17 
AppendiI A EQUIP"ttl'lo'T CHARACTERISATION 
Table A. 7 conL 
Time Hydrogen peroxide concentration (mg L"I) 
(min) 
o 
2 
4 
6 
8 
10 
IS 
20 
600mL 
0,00 0,00 
0,02 0.D2 
0,04 
0,05 
0,06 
0,07 
0, 12 
0, 15 
0,04 
0,05 
0,06 
0,09 
0, 11 
0,14 
700mL 
0,00 0,00 
O,oI 0,02 
0,04 
0,05 
0,06 
0,07 
0,09 
0,13 
0,oJ 
0,04 
0,05 
0,07 
0,10 
0,11 
800mL 
0,00 0,00 
O,oI 0,0 1 
0,02 
0,03 
0,06 
0,07 
0, 10 
0,12 
0,03 
0,04 
0,05 
0,06 
0, 10 
0, 13 
900mL 
0,00 0,00 
0,02 0.D2 
0,03 
0,05 
0,05 
0,06 
0,09 
0,10 
0,03 
0,04 
0,05 
0,06 
0,09 
0,11 
The mass of hydrogen peroxide generated in each water volume is reported in Table A.8. 
A-9 
1000 mL 
0,00 0,00 
0,00 0,0 1 
0,02 
0,03 
0,04 
0,05 
0,06 
0,09 
0,02 
0,04 
0,04 
0,06 
0,08 
0,09 
Table A.8 : Mus of hydrogen peroxide generated iD water volumes of ]00 to 1 000 mL during 
sonication with the low intensity ultrasonie horn 
Time 
(min) 
o 
2 
4 
6 
8 
10 
15 
20 
o 
2 
4 
6 
8 
10 
15 
20 
100 mL 
0,000 0,000 
0,0 11 0,0 10 
0,0 19 0,019 
0,029 0,026 
0,032 0,034 
0,037 0,037 
0,045 0,047 
0,055 0,056 
600mL 
0,000 
0,0 14 
0,027 
0,032 
0,04 1 
0,048 
0,070 
0,092 
0,000 
0,0 17 
0,027 
0,030 
0,04 1 
0,055 
0,067 
0,084 
Hydrogen peroude mass (mg) 
200mL 
0,000 0,000 
0,0 15 0,012 
0,022 0,024 
0,030 
0,037 
0,045 
0,056 
0,068 
0,030 
0,039 
0,046 
0,053 
0,064 
700mL 
0,000 
0,0 11 
0,028 
0,037 
0,042 
0.055 
0,068 
0,089 
0,000 
0,0 19 
0,026 
0,033 
0,039 
0,052 
0,070 
0,079 
300mL 
0,000 0,000 
0,013 0,0 12 
0,023 0,025 
0,032 
0,_ 
0,046 
0,063 
0,070 
0,028 
0,040 
0,044 
0,056 
0,070 
800mL 
0,000 
0,0 11 
0,021 
0,030 
0,05 1 
0,061 
0,080 
0,099 
0,000 
0.0 15 
0,025 
0,036 
0,040 
0,053 
0,080 
0,106 
'OOmL 
0,000 0,000 
0,0 19 0,020 
0,029 0,029 
0,036 
0,043 
0,052 
0,069 
0,083 
0,040 
0,044 
0.052 
0,07 1 
0,080 
900 mL 
0,000 
0,026 
0,029 
0,048 
0,045 
0,062 
0,086 
0,093 
0,000 
0.02 1 
0,033 
0,043 
0,052 
0,057 
0,086 
0,097 
SOOmL 
0,000 0,000 
0,0 13 0,013 
0,0 18 0,02 1 
0,036 
0.04 1 
0,050 
0,073 
0,083 
0,030 
0.040 
0,050 
0,067 
0,086 
1000 mL 
0,000 
0,008 
0,026 
0,034 
0,048 
0,050 
0,069 
0,098 
0,000 
0,0 11 
0,026 
0,040 
0,040 
0,063 
0,090 
0,098 
Statistical analysis of the data presented in Figure A 4 is reported in Table FA in Appendix F. The 
maximum standard deviation in the hydrogen peroxide concentrations measured in water volumes of 100: 
Appendis A EQUIJ'MEI'ro1 CRARACT£RISA TION' A-IO 
200; 300; 400; 500; 600; 700; 800; 900 and I 000 mL are 0,013; 0.013: 0,0 17; 0,007; 0,007; 0,009; 0.009: 
0,009; 0,006 and 0,014 mg L,l , respectively_ The initial rate of hydrogen peroxide formation was calculated 
for each water volume from the regression of the data presented in Figure A.4 over 0 to 8 min using the 
linear regression model 
y = bx IA41 
where the coefficient b, the gradient of the regression line, represents the rate of hydrogen pcroxicle 
formation. The calculated rates of hydrogen peroxide formation in the different water volumes (and the 
standard errors of the calculated values) are recorded in Table A9. 
Table A.9: Rate of hydrogen peroxide formation in water volumes of 100 to 1 000 mL during 
sonication with tbe low intensity ultrasonic born 
Water volume Rate of Bl<h formation 
(mL) (mg L" min·1) 
100 0,044 ± 0,0011 0,980 
200 O,02S± 0,0008 0,966 
300 0,017 ± 0,0005 0,974 
400 0,0 I S± 0,0008 0 ,914 
500 0.0 IO ± 0,0003 0.978 
600 0,009 ± 0,0004 0,937 
700 0,008 ± 0,0004 0.933 
800 0,007 ± 0,0003 0,%1 
900 0,007 ± 0,0005 0.874 
1000 0,0()6 ± 0,0002 0.981 
The temperature profiles in the different water volumes during sonication was recorded by measuring 
temperature every 2 min for a IO min sonication period.. The temperature profiles were performed in 
triplicate, the temperature measurements are recorded in Table A.IO. 
Appendix A EQIJIPMDI"T CRAllACTERlSATION A-ll 
Table A.IO : Temperature profiles over 10 min in lnter volumes of 100 to I 000 mL during 
soaication with the low intensity ultruonk horn 
Time Temperature (IIC) 
(miD) lOOmL 200mL 300mL 400mL 
0 25,3 24,5 24,7 24,4 23,7 24,5 23,S 24,0 24,3 25,2 26,0 25,7 
2 37,1 36,4 36,2 31,0 29,6 31 ,0 27,5 2S,I 2S,4 2S,O 2S.9 2S,3 
4 45,4 44,S 44,S 35,S 34,2 35,9 30,7 31 ,5 32,1 30,9 31 ,5 31 ,0 
6 53,0 51 ,9 51 ,3 40,4 3S,9 40,1 33,9 35,0 35,2 33,5 34,0 )),6 
S 59,2 57,6 57,0 44,9 43,7 44,9 37,5 3S, I 38,7 35,S 36,3 35,9 
to 65,9 63,1 63,5 49,1 47,7 49,0 40,0 41 ,1 41 ,9 38,0 3S,6 3S,O 
SOOmL 600mL 700mL 800mL 
0 25,9 26,2 26,5 26,2 26,8 26,0 26,3 26,7 26,5 24,2 24,5 24,8 
2 28,0 29,0 29,0 2S,S 2S,S 2S,O 2S,O 28,4 2S,3 25,6 25,S 25,S 
4 30,5 31,2 31,2 30,5 30,5 30,0 29,S 29,0 29,7 27,2 27,5 27.1 
6 32,5 33,2 33,1 32,4 32,5 31,S 31 ,0 31 ,1 31 ,3 2S,6 2S,6 2S,S 
8 34.8 35,4 35,3 34,0 34,1 33,2 32,S 32,4 32,S 29,S 30,0 30.0 
to 36.S 37,1 37,5 36,0 36,0 35,1 34,1 34,0 34, I 31,0 31,3 31 ,1 
900mL 1000 mL 
0 24,0 24,4 24,7 26,0 25,9 25,S 
2 25.3 25,8 25,9 27,0 27, 1 26,9 
4 26,4 26,1 27,0 2S,4 2S, I 2S,O 
6 27,8 2S,2 28,5 29,5 29,2 29,0 
8 28.8 29.6 29,8 30,6 30,2 30,0 
to 30.0 30,7 30,S 31,6 31 ,2 31 ,0 
Smtistical analysis of the data recorded in Table AIO is reported in Table F.5 in Appendix F. The 
maximum standard deviation in the temperatures measured during sonication in water \'olumes of 100: 200: 
300: 400: 500; 600; 700; 800; 900 and I 000 mL are 1,51 ; 0,95; 0 ,95; 0,46; 0,58: 0,49; 0,44: 0,30: 0,53 and 
0.25 IIC. respectively. 
Acoustic power was calculated fo r each water volume using equation A.2 and ultrasonic intensity was 
calculated by dividing the acoustic power by the water volume. The rate of temperature increase (1:) for 
equation A2 was determined from regression over the 0 to 4 min temperature data. Density (for mass 
calculation) and heat capacity were calculated at the average temperature of each experiment. The 
calculated values of acoustic power and ultrasonic intensity for each of the water volumes are recorded in 
TablcAll . 
Appendix A EQUIPME.!"IT CHARACTERISATION' A-12 
Table A l l : Acoustic power ud ultrasonic intensity in water volumes of 100 to I 000 mL during 
lODicatioo with the low intensity ultrasonic horn 
Water volume 
(mL) 
lOO 
200 
300 
400 
500 
600 
700 
800 
900 
1 ()()() 
Acoustic power 
(W) 
34,9 
38,5 
38,5 
38,1 
41 ,3 
41 ,6 
36,4 
38,_ 
33,3 
39,3 
mtruonic iotensit), 
(W cm.)) 
0,349 
0,192 
0,128 
0,095 
0,083 
0,069 
0,052 
0,048 
0,037 
0,039 
The relationship between ultrasonic intensity and water volume is presented in Figure AS, 
0,4 .--.-\----------- ------ 
• 
'" 5 0,3 l: 
j 
III 0,2 J-
 ,5 \ 
.~ 
~ 0,1 '-
 ::> 
o ------____________ ~ 
o = 
"'" Volume (mL) 
,= 1200 
Figure A.5 : Relationship bem'een ultrasonic intensity and water \'olume during sonication 
with the low intensity ultrasonic born 
Ultrasonic intensity, shown in Figure AS, decreases with increasing water volume according to a JXlwer 
relationship. The formula of the regression line (R2 "" 0,997) in Figure AS is 
y "" 25,6x-O·93 1 (A.5 ( 
Appendix A EQVIPMDli CIlARACTERlSA1l0N A-13 
The relationship between the initial rate of hydrogen peroxide formation and ultrasonic intensity is 
presented in Figure A6. 
o,cs 
:E 
E 
~ 0,04 
S 
< 
0 0,03 ~ 
E 
.. 
N 0,02 0 
N 
" -0 0,01 
" • 0: 
0 
0 o,cs 0,1 0.1 5 0,2 0,25 0,3 0,35 0,4 
Ultrasonic intensity (W/cm3) 
Figure A.6: Relationsbip between initial rate of bydrogen peroJide formation and ultrasonic 
intensity during soDieation It'itb tbe Iolt' intensity ultrasonic bom 
The initial rate of hydrogen peroxide formation, shown in Figure A6, increased linearly with increasing 
ultrasonic intensity. The formula afthe regression line (R2 = 0,998) in Figure A6 is 
y = 0,001 + 0,122x IA6] 
The direct relationship between the initial rate of hydrogen peroxide formation and ultrasonic intensity and 
between intensity and water volume indicates that a sonochemical effect such as hydrogen peroxide 
formation is dependant (with all other parameters kept constant) upon the sample volume being sonicated. 
A.1.4 Acoustic power reduction due to the ultrasonic cell lid design 
The point of contact between the ultrasonic horn and lid of the ultrasonic cell was initially sealed with an 
O-ring. The contact between the horn and the lid of the cell dampened the movement of the horn and 
reduced the transfer of acoustic power to a sample in the cell. The acoustic power transferred to a water 
sample in the cell was calculated by measuring the temperature profile during sonication. The temperature 
profile during the sonication of 500 mL of water in a beaker was also measured.. Temperature was recorded 
every' 2 min for a to min sonication period. experiments were performed in triplicate. The temperature 
profiles measured during sonication of water with the low intensity horn in the ultrasonic cell and in a 
beaker are compared in Figure A 7. 
Appendix A EQUIPMENT CBARACTf.RISA1l0N 
'" 
«> 
[) 
'-
 "'" ~ 
• ~20 
• I-
 10 
0 
~ 
~ 
r 
r 
0 2 
, 
• 
4 
, 
• 
6 
Time (min) 
• • 
• • 
8 10 12 
... beaker 
• cell 
A-14 
Figun: A.7 : Comparison of temperature profiles measured during toDication witb tbe low 
intensity born in 500 mL of water in the ultrasonic cell aDd in a beaker (If z: 3 per time period) 
The lower temperature profile measured in the cell during sonication, shown in Figure A. 7, indicates that 
less energy was transferred to the water in the cell than to the water in the beaker. The data presented in 
Figure A. 7 is nx:orded in Table A 12. 
Table A.12: Temperatures measured during sonication witb tbe low intensity born in 500 m.L of 
water in tbe ultrasonic cell and in a beaker 
Time Temperatun: (oq 
(min) 
ultrasonic cell beaker 
0 26.8 27,0 27,0 25,9 26,2 26,5 
2 27,0 27,3 27,3 28,0 29,0 29,0 
4 27.5 27,9 27,7 30.5 31.2 31.2 
6 28,0 28,3 28,0 32,5 33.2 33, 1 
8 28.4 28,6 28.3 34.8 35.4 35.3 
10 28,7 28,8 28,7 36.8 37,1 37,5 
Statistical analysis of the data presented in Figure: A7 is reported in Table F.6 in Appendix F. The 
maximum standard deviation in the temperatures measured in the ultrasonic cell and in a beaker are 
0.20 and 0.58 qc. respectively. 
Acoustic power transferred to both of the water samples was calculated using equation A2. The rate of 
temperature increase (~) was determined from regression over the 0 to 10 min data using the linear 
regression model 
Appendix A EQUIPMENT CHARACTERISATION A-IS 
y=a+bx [A7] 
where coefficient b, the gradient of the regression line, represents the slope of the temperature profile. The 
calculated gradients of the temperature profiles during sonication in the ultrasonic cell and in a beaker (and 
the standard errors of the calculated values) are recorded in Table A 13. 
Table A. 13 : ComparilOn of the gradients of the temperature profiles measured in SOO mL of water 
during sonication witb the low intensity horn in tbe ultruonic cdl and in a beaker 
beaker 
ultrasonic cell 
Gradient of temperature profiles If 
(OC min-1) 
1,08 ± 0,03 
0,19±0,01 
0,991 
0,957 
Density (for mass calculation) and heat capacity were calculated at the average temperature of each of the 
temperature profiles. The acoustic power transferred during sonication with the low intensity horn to 
500 mL of water in either the ultrasonic cell or a beaker is compared in Table A14. 
Table A.14: Comparison of acoostic power transferred during sonication with the low intensity 
horn to SOO mL of water in the ultrasonic cell and in a beaker 
beaker 
ultrasonic cell 
Water volume 
(mL) 
500 
500 
Acoustic power 
(W) 
37,7 
6,5 
The acoustic power uansferred to a water sample in the ultrasonic cell was approximately 17 % of that 
transferred to 500 mL of water in a beaker (Table A 14). 
A,1.5 Mixing 
The agitation induced in 500 mL of water by the ultrasonic horn was investigated using potassium 
pennanganate crystals. The crystals were screened with a 600 IJ.m sieve, three of the crystals that did not 
pass through the sieve were placed in 500 mL of water in a glass beaker. The water was sonicated wi\tL the 
horn at a transducer displacement of 11 J.Illl. Pictures of the water after sonication for 2 and 5 s <pe 
presented in Figure A8. 
Append~ A EQUIPMENT C1u.RACTERlSATlON A-16 
(a) (b) 
Figure A.S: Mixing of potassium pennanganate crystals in 500 mL of water during sonication 
with the high intensity born; (a) after 2 S and (b) after 5 s 
A control was performed for comparison by adding JX)tassium permanganate crystals to 500 mL of water in 
a beaker but without sonication. Pictures of the control after 10 and 60 min are presented in Figure A9. 
(a) (b) 
Figure A.9: Mixing of potassium permanganate crystals in 500 mL of water without 
sonication; (a) after 10 min and (b) after 60 min 
A.1.6 Rotameter calibration 
The flow rate of the gas bubbled through liquid samples in the ultrasonic cell was controlled by a rotameter. 
The rotameter was calibrated by measuring the now rate of gas (in quadruple) with a bubble meter for each 
Append.is. A EoUlrMViT CIlARACJT.RISA'JlON A-17 
rotameter setting from I to 10. The oxygen flow rate at each rotameter setting was measured at regulator 
pressures of 200 and 300 kPa. The oxygen Oowrates are presented in Figure AID. 
25 
• 
I 
~20 
• • Cl E- I 
• 15 
" 
I • 200 kPa ~ 
• 
.300 kPa ~ 
c 10 
• 
• ~ 0 
• 5 
• 
• 0 • 
, 
0 2 4 • 8 10 12 
Rotameter setting 
Figure A.l0: Rotameter calibrano. for oqgen at rqulator presJllra of 200 aDd 300 kPa 
(n = 4 per rotameter !ettiDg) 
There is no significant difference at the lower rotameter settings, shown in Figure A 10, between the oxygen 
flowrates measured at regulator pressures of 200 or 300 kPa. The flow rate, 6,0 mL s'\ used in gas 
experiments in the ultrasonic cell was obtained at the rotameter setting 4 with the regulator pressure set at 
200 kPa. The flow rate at the rotameter setting 4 with the regulator pressure set at 100 kPa was also 
measured to be 6,0 mL S-I, MinimaJ splashing occurrod at thls flow rate for a 500 mL volume of water. 
The data presented in Figure AID is recorded in Table AIS. 
Table A-1S: ROUmeter c.a..Libration for osygen at regulator pressures of 200 and 300 kPa 
Rotameter Flow rate (mL 5,1) 
setting 
200 kPa 300 kP. 
0,7 0,7 0,8 0,7 0,7 0,7 0,7 0,7 
2 1.7 1,7 1,8 1,8 1,7 1,7 1,7 1,8 
3 3,5 3,5 3,4 3,4 3,6 3,5 3,6 3,6 
4 6,1 6,0 6,0 6,0 6,2 6,1 6,2 
5 8,7 8,8 8,8 8,8 8,6 8, 1 8,6 
6 11 , 1 11 , 1 11,3 11,0 11 ,6 11,6 11, 1 11 ,5 
7 13,8 13,3 13,6 13,6 14,3 13,8 14,0 13,_ 
8 16,7 16,3 16,3 16,0 17,0 16,0 16,2 16,0 
9 19,5 18,7 18,9 19, I 20,5 ' 19,5 19,4 19,5 
10 2\,0 21 ,2 21 ,2 21 ,7 22,1 21 ,6 21,7 21 ,2 
AppeadiI A EQUJ.PMJ:NT CBAIlAC'J"DUSATJON A-IS 
Statistical analysis of the data presented in Figure AIO is reported in Table F.7 in Appendix F. The 
maximum standard deviation in the flowrates of oxygen at regulator pressures of 200 and 300 kPa are 
0.34 and 0,52 mL S·I , respectively. The rotameter calibration was also performed for nitrogen at a regulator 
pressure of 200 kPa. A flow rate of 6,0 mL s·\ as recorded in Table A 16, was obtained at a setting just 
below 4, approximately 3,7. 
Tab" A.16: Rotameter calibratioll for aitrogen .t • regulator pressatre 01200 kPa 
Rotameter Ji10w ",te 
setting (mL," ) 
2,0 2,0 1,9 2,0 2,0 2,0 
3,0 4,2 4,2 4,2 4,2 4,2 
3,5 5,2 5,1 5,1 5,1 
3,7 6,0 6,0 6,0 6,0 6,0 6,0 
4.0 6,5 6,5 6,5 6,5 6,5 
Statistical analysis of the data recorded in Table Al6 is rqx>rted in Table F.8 in Appendix F. The 
maximum standard deviation in the Oowrates of nitrogen is 0,05 mL S·I . 
A.2 OZONE EQUIPMENT 
Ozone production by the Sorbios generator is determined by oxygen Oow rate and power input. The Oow 
rate for all experiments was maintained at 2,4 mL s" and ozone production was changed by varying the 
input of power. The Sorbios ozone generator was characterised by measuring the amount of ozone produced 
in 20 min at voltage settings of 100; 110; 130; 150; 170 and 200 V. A potaSSium ioctide solution was used 
in the ultrasonic cell to react with all the ozone in the oxygen/ozone gas stream. Ozone concentration in 
solution was determined by titration with sodium thiosulphate as described in Section 5.3.2 of Chapter 5. 
Experiments were performed in triplicate. Statistical analysis of the dissolved ozone concentrations 
measured in solution during ozonation at voltages between 100 and 200 V is reported in Table F.9 in 
Appendix F. The maximum standard deviation in the measured ozone concentrations is 0,74 mg L-1• The 
concentration of ozone in the incoming gas stream was calculated from the amount of ozone measured in 
solution over 20 min and the average Oow rate of the oxygen/ozone gas stream. The concentration of ozone 
generated in an oxygen stream at voltages between 100 and 200 V is shown in Figure A.ll . 
Append.is. A .EQUIPMENT CBARACTI.RlSA11ON A-19 
• 
""" S c 0 
~ 
• 
• " 2D c • g 
0 
• u 
• c 10 0 
Cl 
• 
• • 0 
!Il 100 12D 1<> 100 l!1l 2DO ZlD 
Voltage CV) 
Figure A.ll: Ozoae CODCeD.tration in a 2,4 mL 1_1 oxygen gas stream generated by a Sorbios 
O'LOne generator at voltage settings between 100 and 200 V (n ,.. 3 per v~) 
Nebel reported that ozone generation, as shown in Figure All , varies exponentially with voltage applied to 
the high voltage electrode of ozone generators (Nebel, 1981). A voltage setting of 130 V was used as a 
standard voltage for ozone experiments. Oz.one was also generated at voltage settings of 100, 150 and 
170 V when the effect of ozone concentration on the degradation of atrazine was investigated. The data 
presented in Figure All is recorded in Table A 17. 
Table A.17: Ozone concentratiOD in a 2,4 mL S-1 oxygen gas stream generated by a Sorbio! ozone 
generator at VOltage settings between 100 and 200 V 
Voltage Ozone CODcentratioD 
(V) (mg ·L-I ) 
lOO 1,2 1,2 I , I 
110 2,0 2,8 2,7 
130 5,2 5,8 6, 1 
150 14,6 14,6 14,7 
170 20,6 215 20,1 
200 36,7 35,9 36, 1 
The weight percentage of ozone generated in the oxygen/ozone gas stream is reported in Table A18. 
Appendix A EQUIPMJ:NT CBARACTEIUS.A'IlON A-20 
Table A.18: Weigbt pel'ttll~ of 0Z0De in a 2,4 mL Si oxygen gas strum generated by a Sorbios 
ozone geoerator at voltage ItttinCS between 100 aad 200 V 
voltage Weight ptl'UDtage of 0Z0De ia gas 
M (-;.) 
100 0,1 0,1 0,1 
110 0,2 0,2 0,2 
130 0,_ 0,_ 0,5 
150 1,1 1,1 1,1 
170 1,6 1,6 1,5 
200 2,8 2,7 2,7 
B 
ANALYTICAL PROCEDURES 
Analytical procedures used to determine concentrations of hydrogen peroxide, ozone, dissolved oxygen and 
atrazine are presented in this appendix. A copper·DMP (2,9-dimethyl-l ,10-phenanthrotine) method, 
described in Section B.1. was used to measure hydrogen peroxide concentration. Iodometric and indigo 
calorimetric methods, described in Section B.2, were used to determine ozone concentration. Dissolved. 
oxygen concentration, described in Section B.3, was measured using an oxygen electrode. The high 
performance liquid chromatography technique used to determine atrazine concentration is presented in 
Section B.4. 
B.l HYDROGEN PEROXIDE MEASUREMENT 
The hydrogen peroxide concentration of a solution cou1d not be measured directly. a copper-DMP method 
was used to detennine hydrogen peroxide concentration. A copper(J)-DMP complex is fonned from the 
reaction between hydrogen peroxide and copper sulphate in the presence of an excess of DMP. The product 
is measured spectrophotometrically at a wavelength of 454 nm. A calibration curve was used to relate 
hydrogen peroxide concentration to absorbance at 454 nm. The calibration curve \'1'35 constructed by 
plotting concentration of known hydrogen peroxide solutions versus the absorbance al 454 run of the 
product fonned by the reaction of the solutions with copper su1phate and DMP. A calibration curve was 
performed for each new batch of copper su1phate and OMP solutions. Three separate calibration curves 
were used. throughout the investigation. 
B.l.l Calibration curve A 
Three hydrogen peroxide stock solutions were prepared by pipetting 25 mL of a 30 % hydrogen peroxide 
solution (analytical reagent grade (AR); Saarchem) into a 250 mL volumetric flask. Total volume was 
made up with uluapure water from a Millipore Milli Q System. The stock solutions were standardised by 
titration with a 3 840 mg L·1 potassium permanganate solution (AR: B. Owen Jones Chemicals) that had 
been standardised against a 6 SOl mg L'I sodium oxalate solution (AR; BOH Chemicals). Titrations were 
perfonned according to the standard methods reported in Vogel (Voge!. 1%1). The concentrations of the 
three hydrogen peroxide stock solutions were 3 021 , 3 045 and 3 031 mg L' \ respectively. The stock 
solutions were diluted. 500-fold to produce solutions of concentration 6,04; 6.09 and 6,06 mg L·'. 
Appendix B ANALYnCAL PRocEDURES B-2 
respectively. and 5 OOO·fold to produce solutions with concentrations of 0,604; 0,609 and 0,606 mg L-I , 
respectively_ A 3 078 mg L-t copper sulphate stock solution (laboratory reagent grade (LAB); Saarchem) 
was preplIed from the pentahydrate salt The solution was standardised by titration, using the fast sulphon 
black indicator (Vogel 1961), against a 3 723 mg L-t ethylenediaminetetraacetic acid (EDTA) solution 
(LAB; SDH Chemicals)_ A 10 001 mg L-t DMP solution was prepared in ethanol (AR; Saarchem)_ 
The calibration curve was produced by measuring the absorbance at a wavelength of 454 nm of a range of 
solutions of known hydrogen peroxide concentration. Absorbance was measured in a 10 mm glass cuvette 
with a Pharmacia Siotech Ultrospec 2000 S"(::leCbophotometer. The samples for absorbance measurement 
were prepared by mixing in a 10 mL volumetric flask 1 mL of the copper(D) sulphate solution, 1 mL of the 
ethanolic DMP solution and different volumes of the diluted hydrogen peroxide solutions. The volumes 
used of the 500·fold diluted hydrogen peroxide solutions were 1; 1,5; 2; 3 and 5 mL; volumes used of the 
5 OOO·fold diluted hydrogen peroxide solutions were 1; 2; 4; 6 and 8 mL. Total volume was made·up with 
ultapure water. The procedure was repeated for the diluted solutions of each of the three hydrogen peroxide 
stock solutions. A control was perfonned by mixing I mL of the copper(II) sulphate solution, 1 mL of the 
ethanolic DMP solution and making up to total volume (10 mL) with ultrapure water. The calibration curve 
of absorbance (in absorbance units, AU) measured at 454 nm for different hydrogen peroxide concentrations 
(in rng L- I ) is shown in Figure B.1. 
1,2 [ 
I 
" 
~ 0,6 
-
 • ~ a, • 
• 
. 
-e 
~ , 
;'i 0,4 
" 
0,2 -
 0 
0 
• 
a,s 
• 
1,5 2 2,5 
H202 concentration (mg/ L) 
Figure B.l: Hydrogen peroxide c.alibration c.urve A 
The equation of the regression line of the data presented in Figure 8.1 is 
Absorbance = (0,3597 ± 0,00(7) x (Hz02] 
• 
3 3,5 
[B. 11 
Appendi:s: B ANALYI1CAL PROCEDUR£S B-3 
The coefficient of determination (Rl) of the regression line is 0,987, thus, approximately 99 % of the 
variability in the data is accounted for with the regression equation 8.1 . The error (standard error of the y 
estimate divided by the mean of the sample data) in hydrogen peroxide concentrations calculated using 
equation B. I from absorbance measured. at 454 nm is 11 %. The data presented in Figure B. I is recorded in 
Table B.1 . 
Table B.l : nata for bydrogen perolide calibratioa curve A 
8 1<h stock IOlution A 
Hl<h Absorbance 
concentration (AD) 
(mg L ' I) 
0,000 
0,061 
0,121 
0,243 
0,364 
0,486 
0,607 
0,911 
I.214 
1,821 
2,428 
3.035 
0,000 
0,033 
0,053 
0,106 
0,162 
0,208 
0,274 
0,376 
0,485 
0,068 
0,887 
1,018 
B.1.2 Calibration curve B 
Bl<h stock solution B 
Bl<h Absorbance 
concentration (AU) 
(mg L-I ) 
0,000 
0,061 
0,122 
0,243 
0,365 
0,487 
0,609 
0,913 
1,217 
1,826 
2,434 
3,043 
0,000 
0,028 
0,058 
0,104 
0,157 
0,208 
0,258 
0,392 
0,477 
0,6% 
0.867 
1,021 
81<h stock solution C 
H,o, Absorbance 
cooceDtratiOn (All) 
(mg L·I ) 
0,000 0,000 
0,061 0,027 
0,121 0,061 
0,243 0,108 
0,364 0,153 
0,486 0,205 
0,607 0,269 
0,911 0,391 
1,215 0,477 
1,822 0,686 
2,430 0,869 
3.037 1.016 
Hydrogen peroxide stock solutions were prepared for calibration curve B by pipetting 25 mL of a 30 % 
hydrogen peroxide solution (AR: Saarchem) into a 250 mL volumetric flask. Tow volume was made up 
with ultrapure water. The stock solutions were standardised by tiuation with a 3 390 mg L'I potassium 
permanganate solution (AA: 8. Owen Jones Chemicals) that had been standardised against a 6 802 mg L' l 
sodiwn oxalate solution (AR; BOH Chemicals). The concentrations of the three hydrogen peroxide stock 
solutions were 2 995, 3 017 and 3 008 mg L' I, respectively. The stock solutions were diluted 500-fold to 
produce solutions of concentration 5,99; 6,03 and 6,02 mg L·1, respectively, and 5 OOO-fold to produce 
solutions of concentration 0,599: 0,603 and 0,602 mg L'\ respectively. A 3 075 mg L·1 copper sulphate 
stock solution (LAB: Saarchem) was prepared from the pentahydrate salt. The solution was standardised by 
Appendix B ANALYTICAL PRocEDURES 
titration, using the fast sulphon black indicator, against a 3 724 mg L·1 ethylenediaminetetraacetic acid 
solution (LAB; BOH Chemicals). A 10 009 mg L·1 OMP solution was prepared in ethanol (AR; Saarcbem). 
The procedure to prepare a hydrogen peroxide calibration curve is detailed. in Section B.1.1. The second 
calibration curve of absorbance (in absorbance units, AU) measured at 454 om for different hydrogen 
peroxide concentrations (in mg L-1) is shown in Figure B.2. 
1,2 
0,. , 
:> 
$ 
• g 0,6 
• e 
0 
• • It 
• 0,3 
1,5 2 2,5 3 3,5 
H202 concentration (mg/l ) 
Figure 8.2: Hydrogen peroxide calibration curve B 
The equation of the regression line of the data presented in Figure B.2 is 
Absorbance"" (0,3402 ± 0,0056) x [Hz92] [B.2) 
The coefficient of determination (R2) of the regression line is 0.979, thus, approximately 98 % of the 
variability in the data is accounted for with the regression equation 8 .2. The error in hydrogen peroxide 
concentrations calculated using equation 8.2 from absorbance measured at 454 nm is 14 %. The data 
presented in Figure 8.2 is recorded in Table 8.2. 
Appendix B ANALYnCAL PROCEDURES 8-S 
Table B.2 : Data for bydrogen peroxide calibration cuoe B 
B1(h stock solution A Hl<h stock solution B H1(h stock solution C 
H,o, Absorbance H,o, AbJOrbance H,o, AbJOrbance 
concentration (AU) coottotratioo (AU) cooceotratioo (AU) 
(mg L-') (mg L-') (mg L-t ) 
0,000 0,000 0,000 0,000 0,000 0,000 
0,060 0,028 0,060 0,030 0,060 0,043 
0,120 0,053 0,121 0,056 0,120 0,058 
0,240 0,105 0,241 0,103 0,241 0,124 
0.359 0,154 0,362 0,147 0,361 0,167 
0,479 0,198 0,483 0,205 0,481 0,218 
0.599 0,248 0,603 0,247 0,602 0,253 
0,899 0,346 0,905 0,364 0,902 0,389 
1,198 0,471 1,207 0,450 l,203 0,466 
1,797 0,641 1,810 0,660 1,805 0,643 
2,396 0,771 2,414 0,810 2,406 0,818 
2,995 0,903 3,017 0,928 3,008 1,034 
B,1.3 Calibration curve C 
Hydrogen peroxide stock solutions were prepared for calibration curve B by pipetting 25 mL of a 30 % 
hydrogen ~roxide solution (AR; Saarchem) into a 250 mL volumetric flask. Total volume was made up 
'with ultrapure water. The stock solutions were standardised. by titration with a 3 629 mg L'] potassium 
~nnanganate solution (~B. Owen Jones Chemicals) that had been standardised against a 6 800 rog L' t 
soctium oxalate solution (AR: BDH Chemicals). The concentrations of the three hydrogen peroxide stock 
solutions were 3 166, 3 065 and 3 047 mg L'\ respectively, The stock solutions were diluted 500-fold to 
produce solutions of concentration 6.33: 6,13 and 6,09 mg L't . respectively, and 5 OOO-fold to produce 
solutions of concentration 0,633; 0,613 and 0,609 mg L'], respectively. A 3 090 mg L'] copper sulphate 
stock solution (LAB: Saarchem) was prepared from the pentahydrnte salt. The solution was standardised by 
titration, using the fast sulphon black indicator, against a 3 723 mg L'] ethylenediaminetetraacetic acid 
solution (LAB; BDH Chemicals). A 10 006 mg L·t DMP solution was prepared in ethanol (AR; Saarchem), 
The procedure to prepare a hydrogen peroxide calibration curve is detailed in Section B.1.1. The second 
calibration curve of absorbance (in absorbance units. AV) measured at 454 nm for different hydrogen 
peroxide concentrations (in mg L") is shown in Figure B.3. 
Appendix B ANALYTICAL PROCEDURES B-<i 
1,2 r-------- -----------------, 
I· 
~ 0,6 
• g a,s 
• e 
0 
• a,' 11 
0,2 
° 
° 
0,5 1,5 2 2,5 3 3,5 
H202 concentration (mg/ l ) 
Figure D.3: Hydrogen peroxide calibration curve C 
The equation of the regression line through the data presented in Figure 8.3 is 
A bsorbance:= (0,3410 ± 0,0053) x rHAJ [B.3] 
The coefficient of determination (R2) of the regression line is 0,981 , thus, approximately 98 % of the 
variability in the data is accounted for with the regression equation B.3. The error in hydrogen peroxide 
concentrations ca1cu1ated using equation B.3 from absorbance measured at 454 nm is 13 %. The data 
presented in Figure B .3 is recorded in Table B.3. 
Append.ix B ANALYIlCAL PROCEDURES S-7 
Table B.3 : Data for bydrogen peroxide calibration curve C 
B2~ stock solution A B2<h stock !M)lution B H2<h stock solution C 
B,o, Absorbance B,o, Ab!JOrbance B,o, Ab!JOrbance 
concentration (AV) concentration (AV) concentration (AV) 
(mg L -') (mg L ·l ) (mg L -' ) 
0,000 0,000 0,000 0,000 0,000 0,000 
0,063 0,036 0,061 0,034 0,061 0,028 
0,127 0,080 0,123 0,073 0,122 0,054 
0,253 0,140 0,245 0,133 0,244 0,104 
0,380 0,182 0,368 0,170 0,365 0,158 
0,506 0,223 0,490 0,206 0,487 0,207 
0,633 0,271 0,613 0,256 0,609 0_238 
0,950 0,368 0,920 0,369 0,914 0,333 
1,266 0,493 1,226 0,482 1,226 0,453 
1,899 0,658 1,839 0,688 1,827 0,647 
2,532 0,866 2,452 0,866 2,436 0.811 
3,165 0,990 3,065 0,983 3,045 0,951 
8.2 OZONE MEASUREMENT 
Aqueous ozone concentration was determined using an iodometric (Eaton et al ., 1995a) and indigo 
colorimetric method (Bader and Hoigne, 1981 ; Eaton et al., 1995b). The iodometric method is detailed in 
Section B.2. 1 and the indigo colorimetric method in Section B.2.2. 
B.2.1 Iodometric method 
The iodometric method was used for daily characterisation to determine the rate of ozone generation. 
Experimental conditions were kept oonstanl., however. the rate of ozone generation (at a \'oltage of 130 V) 
varied typically between 0.014 and 0,020 mg S·l due to the daily fluctuation in oxygen Oowrate between 
2 and 3 mL S·l . The % error in the daily rate of ozone generation during eXTJerimentation was calculated 
from the ozone mass balances feJX)rted in Section 0 .4 to be 7,5%. The rate of ozone generation was 
determined b)' bubbling ozone for 20 nUn through 500 mL of a 10 000 mg L·1 potassium icxl.ide solution in 
the ultrasonic cell (AR; Merck). A 50 mL sample of the iodide solution post--ozonation was titrated against 
a 25 140 mg L·1 sodium thiosulphate solution (AR PAL Chemicals) that had been standardised against a 
4 904 mg L·1 potassium dichromate solution (G~ Merck). Titrations were performed according to 
Standard Methods (Eaton et al. , 1995a). The iodometric method was also used in the series of ozone mass 
balances to detennine the amount of ozone that passed through the ultrasonic ceU and reacted with the 2 % 
potassium iodide solution in the gas washing bottles. The mass balances were performed over a 45 min 
period of ozonation. The total amount of ozone generated during a 45 min period was measured using a 
AppeodiJ: B ANALYTICAL PROCEDURES B-8 
16 000 mg L'! potassium iodide solution in the ultrasonic cell. The iodometric method can be used to 
detennine ozone concentrations above I mg L·1 (Eaton et al., 1995a). 
B.2.2 Indigo colorimetric method 
Residual ozone in water or atrazine solutions in the ultrasonic cell was measured using the indigo 
colorimetric method. Indigo stock solution was prepared from potassium indigo trisulfonate (AA; Aidrich) 
and diluted IQ-fold and 50-fold to prepare indigo reagents for the detennination of different concentrations 
of ozone-rontaining solutions (Bader and Hoigne, 1981). The absorbance at a wavelength of 600 om was 
measured of a solution prepared by mixing, in a lOO mL volumetric flask. 10 mL of the indigo reagent, a 
known volume of the ozone-rontaining solution so that the indigo solution was partially decolourised and 
distilled. water to makeup total volume. Controls for the two reagents were performed by measuring 
absorbance at 600 run of a solution prepared by mixing 10 mL of the indigo reagent and 90 mL of distilled 
water in a lOO mL volumetric flask. Ozone concentration was calcuJated from the difference in absorbance 
at 600 om between that of the control and the sample, taking into account the volume of ozone-rontaining 
solution that used in the analysis (Eaton et aI. , 1995b) The indigo colorimetric method can be used to detect 
ozone concentrations down to 2 ~g L'! (Eaton et al., 1995b) 
B.3 DISSOLVED OXYGEN MEASUREMENT 
The oxygen electrode used to measure dissolved oxygen concentration was standardised daily for changes in 
aunospheric pressure and calibrated daily using oxygen-saturated water at 25 0c. The solubility 
concentration of oxygen in water at a pressure of 101,325 kPa and for temperatures between 0 and 40 °C is 
presented in Figure B.4 (International Critical Tables, 1928). 
00 ------------------------------__ __ 
• 
• 
• 
• 
• • • 
o --~----______ ~ ____________ ~ ____ ~ 
o 10 20 40 
Temperature (0C) 
Figure B.4: Solubility of oxygen in water at 101,325 kPa and temperatures between 0 and 
40 °C (International Critical Tables. 1928) 
Appendix B ANALYTICAL PRocEDURES 8-9 
The data presented in Figure B.4 is recorded in Table BA. 
Table B.4: Solubility of oxygen in water at 101,325 kPa and temperatures between 0 and 40°C 
(International Critical Tables, 1928) 
Temperature 
(OCl 
o 
10 
15 
20 
25 
30 
35 
40 
B.4 ATRAZINE MEASUREMENT 
O:rygen coocentration 
(mg L-') 
69,76 
54,40 
49,28 
44,16 
40,32 
37, 12 
34,88 
32,96 
Atrazine concentration was measured using high performance liquid chromatography (HPLC). The HPLC 
system used a LiChrospher'" RP-Select B (250 x 4 mm. 5 jJJ1l particles) column with a 70:30 (v/v) solvent 
mixture of methanol to water. The solvent was pumped at a fiowrate of 0,6 mL min·1 through the column. 
The sample injection volume was 25 ~ and absorbance was measured at a wavelength of 230 nm. The 
calibration curve was performed by injecting, in triplicate. atrazine (97 %; Sanachem) concentrations of I: 
2: 5: 10 and 20 mg L-1 through the HPLC system and recording the absorbance at 230 nm. The calibration 
curve is presented in Figure 8.5. 
5 ----------------__________________ --, 
"' 
_t 
5 " -
 ~ 
E. 3 -
 • ~ 
~ 2 _ 
11. 
':l 
"- 1 _ 
:L 
o ~~ ______________________________________ _" 
o , • 12 
" 
24 
Atrazine concentration (mg/ L) 
Figure B.5 : Atrazioe calibration cun'e 
Appendix B ANALYTICAL PROCEDURES 8-10 
The equation of the regression line through the data presented in Figure B.5 is 
HPLC peak area = (206 479 ± 2 443) x (Atrazine] [B.4] 
The coefficient of determination (R2) of the regression line is 0,996, thus, approximately 100 % of the 
variability in the data is accounted for with the regression equation 8.5. The error in concentrations of 
atrazine calculated using equation B.5 of the calibration cwve is 7 % . The data presented in Figure 8.5 is 
recorded in Table 8 .5. 
Table B.S: Atrazine calibration cUn'e data 
Hl~ concentration 
(mg L-1) 
0 0 
1 229707 
2 485799 
5 1 193211 
10 2044989 
20 4020687 
&PLC peak area 
(units) 
o 
225625 
476 326 
1 130174 
1 916789 
4191517 
o 
201 579 
439410 
1029 198 
1839291 
4291244 
Statistical analysis of the data presented in Figure 8.5 is reported in Table F.IO in Appendix F . 
c 
ULTRASOUND EXPERIMENTAL DATA 
Data for the ultrasound process investigation discussed in Chapter 6 is recorded in this Appendix. 
Dissolved oxygen measurements are reported in Section C.I , bydrogen peroxide formation in Section C.2, 
hydrogen peroxide degradation in Section C.3, experiments using commercial hydrogen peroxide in Section 
C.4 and interval experiments in Section C.S. 
C.l DISSOLVED OXYGEN CONCENTRATION 
The decrease in dissolved oxygen concentration in oxygen-saturated water was measured. during sonication 
and when nitrogen was sparged through the water without sonication. The reduction in oxygen 
concentration (without sonication) was measured as a controL Oxygen-saturated water was prepared by 
sparging oxygen through the \\'3ter at a flow rate of 6 mL 5.1 for IQ min. Dissolved oxygen concentration 
was measured at time periods 0[0; 5; 10; 20; 30; 45; 60; 90; 120; 150 and 180 min. Triplicate experiments 
were perfonned for each time period.. The data recorded in Table C.I is presented in Figure 6.1 in 
Chapter 6, statistical analysis of the data is reponed in Table F .11 in Appendix F. 
Table C.I : Dissoh'ed oxygen concentration in oxygen-saturated water during sonication, without 
sonication (control) and during nitrogen sparging witbout sonication 
Time Oxygen concentration (mg L-1) 
(min) 
control sonication nitrogen 
0 37,2 40,1 40,0 40,3 37.7 36,7 40,3 39.9 36,4 
5 35,3 37,7 35,3 30,3 29,2 28,1 2,8 3.0 3,1 
15 32,6 33.2 32,4 23.6 22,7 24.3 1,9 1.6 1.8 
30 29.7 29,9 20,8 20.7 21,2 1.8 1,5 1,4 
45 28.8 29,0 18.0 19,7 
60 27,6 27.7 28,0 17,2 17,2 17,7 1.7 1,5 
90 26,1 25,7 25,5 14,4 14,2 14.3 
120 25,1 25,3 13,4 12,8 13,3 2,0 1,8 
150 24,4 24,5 24,9 12,0 12,2 
180 24.2 23,8 11 ,2 1,5 1,4 
Appendix C ULTRASOUND EXPERlMElli'TAL DATA C-2 
The dissolved oxygen concentration of water saturated with air was measured for all solutions just before 
experimentation. Oxygen concentration varied between 8,3 and 12,3 mg L-1 (fable C.2) with an average of 
10,0 mg L·1. The standard deviation in the oxygen concentrations is 0,90 mg L·1, statistical analysis of the 
data is reported in Table F.12 in Appendix F. 
Table C.2: DillOlved oxygen eouuntration in water saturated with air 
Oxygen CQDCentration (mg L-1) 
8,6 9,3 8,3 9,' 8,5 8,' 10,0 8,8 
9,6 9,8 9,9 10,1 9,6 9,6 11,2 10,6 
10,8 9,' 9,8 9,2 8,7 9,9 10,1 9,7 
10,5 10,2 9,9 10,8 10,6 12,3 10,2 10,0 
10,' 10,3 12,3 10,6 10,5 10,3 10,9 9,4 
Dissolved oxygen concentration during sonication and saturation with oxygen. air and nitrogen was 
measured over a period of 40 min in the ultrasonic cell. The gases were sparged through 500 mL of water 
at a flow rate of 6 mL S-I for 10 min prior to and during sonication. Oxygen concentration was measured at 
time periods of 0; 5; 10; 20 and 40 min; duplicate experiments were perfonned for each time period. A 
control was perfonned. by measuring oxygen concentration during sonication of water with no gas sparging. 
The data recorded in Table C.3 is presented in Figure 6.2 in Chapter 6, statistical analysis of the data is 
reported in Table F.13 in Appendix F. 
Table C.l: Dissolved oxygen concentration in nitrogen-, air- and oxygen-saturated water during 
sonication in tbe ultrasonic cell at an acoustic power of 57 W 
Time Oxygen concentration (mg L·1) 
(min) 
control nitrogen "r oxygen 
0 9.8 9,6 0,9 0,8 10,' 11 ,5 40,0 39,8 
5 10,8 10,7 0,8 0,8 12,3 12,2 40,4 40,0 
10 11,0 10,9 0,6 0,9 13,3 12,2 40,0 39,5 
20 10,5 10,1 0,8 1,0 13,6 12,7 39,0 ' 0,0 
40 11 ,7 11,8 0,9 1,2 13,1 13,2 39,0 38,0 
C.2 HYDROGEN PEROXIDE FORMATION 
The formation of hydrogen peroxide during sonication in nitrogen-, air- and oxygen-saturated water was 
measured using the DMP method detailed in Section 5.3. 1 of Chapter 5, The gases were sparged through 
AppendiJ: C ULTRASOUND EXJ>ERlMENTAL DATA C-3 
the water at a flow rate of 6 mL s'\ for lO min prior to and during sonication (at an acoustic power of 57 W) 
to maintain a constant oxygen concentration in the water. Hydrogen peroxide concentration., reported in 
Table C.4, was measured at time periods of 0; 4; 8; 12; 16; 20; 30; 60; 120; 240; 480 and 960 min (16 h). 
Triplicate experiments were perfonned for each time period for each gas. A control was performed by 
measuring hydrogen peroxide concentration during the sonication of water with no gas sparging. The data 
recorded in Table C.4 is presented in Figure 6.3 in Chapter 6 , statistical analysis of the data is reported in 
Table F.15 in Appendix F. 
Table C.4 : Hydrogen peroxide formation in nitrogen-, air- and oxygen-saturated water during 
sonication io the ultrasonic c:cll for 16 h at an acoustic power of 57 W 
Time HIOt concentration (mg L· I ) 
(min) 
""".1'01 nitrogen 
0 0,00 0,00 0,00 0,00 0,00 0,00 
4 0,02 0,02 0,02 0,01 0,01 0,01 
8 0,04 0,03 0,05 0,02 0,02 0,01 
12 0,05 0,05 0,04 0,04 0,04 0,04 
16 0,06 0,06 0,75 0,05 0,05 0,04 
20 0,97 0,09 0,78 0,06 0,06 0,06 
30 0,12 0,11 0,11 0,05 0,04 0,05 
60 0,15 0,15 0,15 0,06 0,06 0,05 
120 0,24 0,28 0,27 0,07 0,06 0,07 
240 0,35 0,38 0,42 0,08 0,09 
480 0,77 0,76 0 ,78 0,09 0,08 
960 0,88 0,97 0,07 0,09 
Time BlOt concentration (mg L·I ) 
(min) 
,.;, oxygen 
0 0,00 0,00 0,00 0,01 0,01 0,01 
4 0,Q2 0,02 0,02 0,05 0,05 0,04 
8 0,04 0,04 0,04 0,06 0,07 0,06 
12 0,04 0,05 0,05 0,07 0,08 0,09 
16 0,06 0,07 0,08 0,08 0,09 0,09 
20 0,08 0,09 0,09 0,11 0, 11 0,11 
30 0, 10 0 , 11 0 , 12 0,16 0,15 0,16 
60 0,18 0,18 0,19 0,26 0,26 0,26 
120 0,25 0,28 0,29 0,44 0,42 0.42 
240 0,34 0 ,34 0,35 0,46 0,50 0,48 
480 0,71 0,82 0,81 0,93 0,74 0,92 
960 l , lO 0,92 1,07 1,15 1,28 
Appendix C ULTRASOUJ'IrI"D ExPERIMnlliAL DATA C-4 
Hydrogen peroxide formation during sonication at an acoustic )X)wer of 24 W was measured in 
nitrogen-saturated. water and a control. Water saturated. with rutrogen was prepared by sparging nitrogen 
through the water for 10 min prior to and during sonication at a flow rate of 6 mL S-l . A control was 
performed. by measuring hydrogen peroxide concentration in water with no gas sparging. Hydrogen 
peroxide concentration was measured at time periods of 0; 4; 8; 12; 16; 20; 30; 60; 120; 240; 480 and 
960 min. Triplicate experiments were performed for each time period. The data recorded in Table C.5 is 
presented in Figure 6.4 and Figure 6.5 in Chapter 6, statistical analysis oftbe data is reponed in Table F.16 
in Appendix F. 
Table C.S : Hydrogen peroxide formation in nitrogen-saturated water and the control during 
sonication in tbe ultratOGic cell at aD acoustic power of 14 W 
Time H]Oz coocentraoon (mg L-I ) 
(min) 
control a.itrogen 
0 0,00 0,00 0,00 0,00 0,00 0,00 
4 0,02 0,01 0,10 0,01 0,01 0,01 
8 0,02 0,02 0,02 0,02 0,03 0.02 
12 0 ,02 0,03 0,03 0,03 0,04 0,03 
16 0 ,04 0,04 0,04 0,05 0,05 0,05 
20 0,05 0,05 0,06 0,05 0,05 0,06 
30 0 ,07 0,07 0,07 0,06 0,06 0,05 
60 0 ,14 0,13 0,14 0,05 0,06 0,05 
120 0 ,24 0,24 0,25 0,06 0,06 0,Q7 
240 0,44 0,43 0,42 0,07 0,07 0,08 
480 0 ,77 0,72 0.73 
960 0,98 
C.3 HYDROGEN PEROXIDE DEGRADATION 
Hydrogen peroxide is tx>th formed and degraded by free radical reactions during sonication_ The rate of 
hydrogen peroxide degradation was determined by sonicating water for 16 h at an acoustic power of 57 W to 
generate hydrogen peroxide. Hydrogen peroxide degradation after sonication was stopped was measured 
over 3 h. Hydrogen peroxide concentration was measured at time periods of 0; 5; 10; 20; 30; 60; 00; 120: 
150 and 180 min. Nitrogen, air and oxygen were sparged. through the water during the 3 h degradation 
period to investigate the effect of gas sparging upon the rate of hydrogen peroxide degradation. A control 
was performed by measuring hydrogen peroxide degradation in water with no gas sparging. Triplicate 
e..'-periments were perfonned for the control and for each gas. Different hydrogen peroxide concentrations 
Appendix C ULTRASOUND EXPERIMENTAL DATA c-s 
were produced during the 16 h p:riod of sonication thus resulting in different initial concentrations for the 
degradation experiments. The variation in initial concentration was removed, for comparison purposes. by 
normalising the data and dividing each concentration by the initial concentration for that particular 
experiment. The data recorded in Table C.6 is presented in Figure 6 .7 in Chapter 6 , statistical analysis of 
the data is reported in Table F.17 in Appendix F. 
Table C.6 : Hydrogen peroxide degradation (nonnalited data) in nitrogen-, air- and oxygen-
 saturated watl:r after sonication in the ultruonic cell for 16 h at an acoustic power of 57 W 
Time (BzOt] / initial [BzOt] 
(min) (mg L-t/mg Lot) 
control nitrogen 
0 1.00 1.00 1.00 1.00 1,00 1,00 
5 1,01 1,00 1,00 0,98 0.% 0.% 
to 1.01 0.99 1,00 1,00 0,97 0,95 
20 1,02 0,98 1,00 0,% 0.91 0,97 
30 1,01 0,97 1.01 0,% 0,95 0.93 
60 1,01 0,97 0,98 0,90 0,87 0.90 
90 0,97 0.% 0,90 0,84 0.85 
120 0.99 0,95 0.87 0,81 0.85 
150 0,98 0,93 0,86 0,76 0,81 
180 0,95 0 ,77 0 ,75 0 ,76 
Time IB,o,) I ;.,;tialIB,o,) 
(min) (mg Lo1/mg LOI) 
,.; , oxygen 
0 1,00 1,00 1,00 1,00 1,00 LOO 
5 0.96 0.94 1,02 0,95 1,01 0.95 
to 0.% 0,94 0,98 0,97 1.02 0.95 
20 0.95 0,93 0,98 0.95 0.98 0.95 
30 0,96 0.93 0.94 0,95 0.99 0.95 
60 0.89 0,87 0,91 0,91 0,93 0,91 
90 0.87 0,86 0,87 0,90 0,88 
120 0.82 O,SO 0.84 0,84 0.88 0.83 
150 0,79 0,78 0,79 0,83 0,83 0.80 
180 0,75 0.74 0,78 0,79 0.79 0,78 
The actual hydrogen peroxide concentrations are presented in Table Co7. 
Appendix C ULTRASOUND EXPERIMENTAL DATA c.;; 
Table C.7 : Hydrogen peroxide degradation in nitrogen-. air- and oxygen-saturated water after 
sonication in the uhrasonic cell for 16 h at an acoustic power of 57 W 
Time [H,o,) I initial [8,0,) 
(min) (mg L-t/mg L-t) 
control nitrogen 
° 
0,38 0,44 0,47 0,71 1,02 1,25 
5 0,38 0,44 0,47 0,69 0,98 1,20 
10 0,38 0,44 0,47 0,71 0,99 1,18 
20 0,38 0,43 0,47 0,68 0,93 1,21 
30 0,38 0,43 0,48 0,68 0,97 1,16 
60 0,38 0,42 0,46 0,64 0,89 1, 12 
90 0,37 0,42 0,45 0,64 0,86 1,07 
120 0,38 0,42 0,41 0,62 0,83 1,06 
150 0,37 0,41 0,41 0,6 1 0,78 1,01 
180 0,36 0,39 0,41 0,55 0,77 0,95 
Time [H,o,) I initial [8,0,) 
(min) (mg L-t/mg L-1) 
air oxygen 
° 
1,28 1,16 0,40 0,37 0,31 0,42 
5 1,23 1,09 0,40 0,35 0,32 0.40 
10 1,23 1,10 0.39 0,36 0,32 0,40 
20 1,22 1,08 0,39 0,35 0,3 1 0.40 
30 1,23 1,08 0,38 0,35 0,3 1 0.40 
60 1,13 1,00 0,36 0,34 0,29 0.38 
90 1,16 0,94 0,3' 0,32 0,28 0.37 
120 1,05 0,93 0,33 0,3 1 0,28 0.35 
150 1,01 0,90 0,32 0,3 1 0,26 0.3' 
180 0,95 0,86 0,31 0,29 0,25 0.33 
CA INTERVALEXPERUMENTS 
Ultrasonic reactors, such as the HarM'e// sonochemica1 reactor, are based on the concept of a flow loop 
system in which the reaction solution is circulated through a sonicated zone and a holding tank or reservoir. 
A flow loop system is advantageous if the limiting reagent of a reaction is retained in the sonicated zone or 
if the reagents exhibit a memory effect and remain activated after passing through the sonicated zone. 
AppendiJ: C ULTRASOUND EXPERIMENTAL DATA C-7 
Interval experiments were performed to monitor hydrogen peroxide concentration when sonication was 
stopped for a certain time period and then restarted so as to simulate a flow loop system with the solution 
moving in and out of a sonicated zone. Hydrogen peroxide concentration was measured in oxygen~saturated 
water at time periods of 0; 5; 10; 15; 20; 25; 30; 35; 40 and 45 min. Saturation was achieved by sparging 
oxygen through the water for 10 min prior to and during the 45 min experiment at a flow rate of 6 mL S·l. 
The experiment was structured so that the water was sonicated for 15 min; sonication was stopped for the 
following 15 min and then restarted for a further 15 min. Sonication in the ultrasonic cell was performed at 
an acoustic power of 57 W. A control was performed in water with no gas sparging. All experiments were 
performed in triplicate. The data recorded in Table C.S is presented in Figure 6.S in Chapter 6, statistical 
analysis of the data is reported in Table F.tS in Appendix F. 
Table c.S: Hydrogen peroxide formation in oxygen-saturated water and a control in the ultrasonic 
ceD during 15 miD of IOnication at an acoustic power of 57 W. 15 miD without JODication and a 
further 15 min with sonication 
Time Ultrasound Hydrogen peroxide concentration (mg L+I) 
(min) 
control oxygen 
0 on 0,00 0,00 0,00 0,00 0,0 1 0,00 
5 on 0,04 0,04 0,03 0,05 0,06 0,05 
10 on 0,06 0,06 0,06 0,08 0,09 0,09 
15 on/off 0,08 0,08 0,08 0, 11 0, 11 0, 11 
20 off 0,08 0,07 0,07 0, 11 0, 11 0,10 
25 off 0,80 0,07 0,08 0,11 0, 12 0. 11 
30 off/on 0,08 0,07 0,08 0, 11 0, 10 0, 11 
35 on 0, 11 0,10 0,11 0, 14 0,13 0,13 
40 on 0,12 0,12 0,13 0,16 0. 16 0.17 
45 on 0,14 0,15 0,14 0,19 0.19 0, 19 
Interval experiments were also performed to monitor the effect of a change in gas during sonication. An 
oxygen~saturated solution was prepared by sp<uging oxygen through the solution for IO min prior to and 
during sonication at a flow rate of 6 rnL S·l . Sonication was performed with a 15 min period of oxygen 
saturation followed by a 15 mill period with nitrogen saturation and a further 15 min period ,vith oxygen 
saturation. Hydrogen peroxide concentration was measured at time periods of 0; 5; 10; 15; 20; 25 ; 30; 35; 
40 and 45 min. Experiments for each time period were performed in triplicate. The data recorded in 
Table C. 9 is presented in Figure 6.9 in Chapter 6, statistical analysis of the ~ta is reported in Table F.19 in 
Appendix F. 
Appendix C ULTRASOUND EXPERIME!'I'TAL DATA C-8 
Table C.9: Hydrogen peroxide formation in water during tonication in tbe ultrasonic ceU at an 
acoustic power of 57 W with 15 min of oxygen saturation, 15 min of nitrogen saturation aDd a 
further 15 min of oxygen saturation 
Time 
(min) 
Gas 
° oxygen 
5 oxygen 
10 oxygen 
15 oxygen/nitrogen 
20 nitrogen 
25 nitrogen 
30 nitrogen/oxygen 
35 oxygen 
40 o),.'ygen 
45 oxygen 
Hydrogen peroside concentration 
(mg L- I ) 
0,00 0,01 
0,04 0,04 
0,09 0,08 
0,11 0,11 
0,13 0,12 
0,14 0,14 
0,15 0,15 
0,16 0,16 
0,19 0,18 
0,20 0,21 
C.S COMMERCIAL HYDROGEN PEROXIDE EXPERIMENTS 
0,01 
0,04 
0,08 
0,11 
0,13 
0,14 
0,14 
0,16 
0,18 
0,20 
Hydrogen peroxide is used commercially for the oxidation of pollutants during water treatment. The use of 
ultrasound in combination with hydrogen peroxide facilitates the formation of hydroxyl radicals from 
hydrogen peroxide. Hydrogen peroxide concentration was measured during sonication for 1 h (at an 
acoustic power of 57 W) of approximate 0,25; 0,50 and 0,75 mg L- I commercial hydrogen peroxide 
solutions (AR grade; Saarchem). The effect of gas on hydrogen peroxide concentration was investigated by 
sparging nitrogen., air and oxygen through the solutions for 10 min prior to and during sonication at a flow 
rate of6 mL S-I . A control was performed in water with no gas sparging. Hydrogen peroxide concentration 
was measured at time periods ofO; 5; 10; 20; 30 and 60 min. Experiments for each time period for each gas 
were performed in triplicate. The hydrogen peroxide concentrations measured during the sonication of a 
0,28 mg L- I hydrogen peroxide solution are recorded in Table C. IO and presented in Figure 6.10 in 
Chapter 6. Statistical analysis of the data is reported in Table F.20 in Appendix F. 
Appendis C ULTRASOUND EXPERIMENTAL DATA C-9 
Table C.I0 : Hydrogen peroside formation in nitrogen-, air- and oxygen-saturated 0,28 mg L-1 
bydrogen peroside !tOlUtiODS during sonication in the ultrasonic cell for 1 b 
Time H2o,. concentration (mg L-I ) 
(min) control nitrogen 
0 0,28 0,27 0,28 0,28 0,27 0,28 
5 0,28 0,29 0,28 0,25 0,26 0,26 
10 0,30 0,30 0,29 0,25 0,26 0,25 
20 0,33 0,32 0,33 0,25 0,24 0,24 
30 0,34 0,35 0,35 0,26 0,26 0,27 
60 0,40 0,40 0,41 0,25 0,26 0,26 
Time H20:z concentration (mg L-I ) 
(min) air oxygen 
0 0,28 0,27 0,27 0,29 0,29 0,29 
5 0,29 0,28 0,27 0,30 0,30 0,29 
10 0,29 0,30 0,29 0,33 0,33 0,31 
20 0,33 0,34 0,32 0,37 0,36 0,37 
30 0,35 0,35 0,34 0,42 0,41 0,40 
60 0,41 0,42 0,42 0,56 0,57 0,55 
The hydrogen peroxide concentrations measured during the sonication of a 0,43 mg L-\ hydrogen peroxide 
solution are recorded in Table C.II and presented in Figure 6.11 in Chapter 6. Statistical analysis of the 
data is reputed in Table F.21 in Appendix F. 
Table C.ll: Hydrogen peroside formation in nitrogen-. air- and oxygen-saturated 0,43 mg L· 1 
hydrogen peroxide solutions during sonication in tbe ultrasonic cell for 1 b 
Time 
(min) 
o 
5 
10 
20 
30 
60 
Time 
(min) 
o 
5 
10 
20 
30 
60 
0,42 
0,46 
0.46 
0,48 
0,50 
0,58 
0,42 
0,45 
0,46 
0,48 
0,51 
0,57 
control 
0,42 
0,45 
0,45 
0,48 
0,50 
0,58 
air 
0,42 
0,45 
0,45 
0,47 
0,50 
0,57 
8 20:z concentration (mg L-I ) 
0,45 0,42 
0,45 0,44 
0,45 0,43 
0,48 0,43 
0,51 0,4 1 
0,59 0,40 
B10:z concentration (mg L-I ) 
0,42 0,45 
0,45 0,47 
0,44 0,47 
0,47 0,51 
0,49 0,52 
0,59 0,61 
nitrogen 
0,44 
0,44 
0,44 
0,42 
0,42 
0,40 
osygen 
0,43 
0,46 
0,48 
0,50 
0,52 
0,62 
0,43 
0,44 
0,42 
0,42 
0,42 
0,39 
0,44 
0,47 
0,47 
0,50 
0,53 
0,63 
Appendis. C ULTRASOUND EXPERIMENTAL DATA C-IO 
The hydrogen peroxide concentrations measured during the sonication of a 0,72 mg L- 1 hydrogen peroxide 
solution are recorded in Table C.l2 and presented in Figure 6.12 in Chapter 6. Statistica1 analysis of the 
data is reported in Table F.22 in Appendix F. 
Table c.n : Hydrogen peroxide formation in nitrogen-, air- and oxygen-saturated 0,72 mg L-1 
bydrogen peroxide solutions during sonication in the ultrasonic ceU for ] b 
Time Hl<h concentration (mg L-1) 
(min) control nitrogen 
° 
0,73 0,73 0,72 0,73 0,73 0,73 
5 0,71 0,73 0,71 0,71 0,71 0,70 
10 0,74 0,73 0,73 0,69 0,69 0,69 
20 0,75 0,74 0,78 0,68 0,68 0,67 
30 0,80 0,79 0,78 0,69 0,68 0,68 
60 0,82 0,81 0,82 0,64 0,63 0,63 
Time Hl<h concentration (mg L-t) 
(min) 
.u, OIygen 
° 
0,71 0,71 0,72 0,71 0,72 0,73 
5 0,72 0,73 0,74 0,77 0,76 0,76 
10 0,74 0,75 0,75 0,77 0,78 0,77 
20 0,77 0,76 0,76 0,76 0,75 0,77 
30 0,78 0,78 0,78 0,78 0,78 0,79 
60 0,80 0,81 0,81 0,83 0,86 0,85 
o 
OZONE EXPERIMENTAL DATA 
Data for the ozone process investigation discussed in Chapter 7 is reGOrded in this Appendix. The dissolved 
ozone concentration in water is reponed in Section 0 .1, ozone decomposition in Section 0 .2, the formation 
of hydrogen peroxide during ozonation in Section 0 .3 and ozone mass balances in Section 0 .4. 
D.1 DISSOLVED OZONE CONCENTRATION 
The dissolved ozone concentration in water was measured during ozonation, ozonation combined with 
sonication (at an acoustic power of 57 W). ozonation combined with hydrogen peroxide (40 mg L'I) and 
ozonation combined with sonication and bydrogen peroxide. Water was saturated with ozone by sparging 
the ozone/oxygen gas stream through the diffuser of the ultrasonic cell. The ozone concentration in the 
2,4 mL 5' \ gas stream was 5,7 mg L't and the ozone production rate was 0,014 mg 5.1 (fable 5.6 in 
Section 5.2). Dissolved ozone concentration in the solutions was measured at time periods 0[0; 1; 2,5; 5; 
10; 20; 40 and 60 min. Duplicate experiments were performed for each time period of the different 
experimental conditions. The dissolved ozone concentrations measured in water during ozonation, 
ozonation combined with sonication, ozonation combined with hydrogen peroxide and ozonation combined 
with sonication and hydrogen peroxide is recorded in Table 0 .1. The data recorded in Table D. l is 
presented in Figure 7.1 in Chapter 7, statistical ana1ysis of the data is reported in Table F.23 in Appendix F. 
Table 0.1: Dissolved ozone concentration in water during ozonation (0,014 mg 51), O'LOnation 
combined with sonic.ation (57 W), ozonltion combined witb bydrogen peroxide (40 mg L- I ) and 
ozonation combined with sonication and bydrogen peroxide 
Time Ozone concentration (mg L-I ) 
(min) OZODe O'LODe/uIS 
° 
0.00 0,00 0,00 0,00 0,00 0,00 0,00 0.00 
1 0,38 0,39 0,48 0,46 
2,5 0,77 0,92 0,77 0,70 
5 1, 15 1.24 0,74 0,76 0,01 0,00 0,02 0.Q2 
10 1,41 1,27 0,67 0,60 0,00 0.00 0.Q2 0,02 
20 1,22 1,03 0,63 0.59 0,01 0.01 0,02 0.Q2 
40 1, 14 1,02 0,56 0.54 0,0 1 0,01 0.Q2 0,02 
60 0,94 1,08 0,60 0,58 0,02 0.Q2 0,07 0,02 
Appendix D OZONE EXPERIMENTAL DATA D-2 
D.2 OZONE DECOMPOSITION 
Ozone decomp:>sition was measured. in water saturated with ozone prepared by sparging the oxygen/gas gas 
through water in the ultrasonic cell for 20 min. The ozone concentration in the 2,4 mL sol gas stream was 
5,7 mg L·l and the ozone production rate was 0,014 mg s'\ (Table 5.6 in Section 5.2). Dissolved ozone 
concentration was measured. over 40 min (after the 20 min saturation period) during ozonation, during 
sonication and when ozonation was combined with sonication. A control was performed by measuring 
dissolved ozone concentration after the 20 min saturation period without ozonation or sonication. Duplicate 
experiments were performed for each time period. The data recorded in Table 0 .2 is presented in 
Figure 7.3 in Chapter 7, statistical analysis of the data is reported in Table F.24 in Appendix F. 
Table D.2: Dissolved ozone concentration in water in tbe ultruonic cell after a 20 min saturation 
period during ozonation (0,014 mg SI), 8ODic.ation (57 W) and ozooation combined with sonication 
Time Ozone concentration (mg LOI) 
(mia) 
control ozone ul, ozone/uls 
0 1,44 1,57 1,22 1,03 0,97 0,92 1,57 1,44 
2,5 0,44 0,46 
5 1,40 0,28 0,27 1,40 
10 1,37 1,40 0,13 0,13 1,37 1,40 
20 1,22 1,27 1,14 1,02 0,07 0,04 1,22 1,27 
40 1,40 1,29 0,94 1,08 1,40 1,29 
D.3 HYDROGEN PEROXIDE FORMATION 
Hydrogen peroxide formation in water was measured during ozonation, sonication and when ozonation was 
combined with sonication in the ultrasonic cell. Ozonation was performed at an ozone production rate of 
0,014 mg S·l and sonication at an acoustic power of 57 W. Hydrogen peroxide concentration was measured 
at time periods of 0; 5; 10; 20; 40 and 60 min. Duplicate experiments were performed for each time period. 
The data recorded in Table 0 .4 is presented in Figure 7.4 in Chapter 7, statistical analysis of the data is 
reported in Table F .25 in Appendix F. 
Appendix D OZONE EXPERIMENTAL DATA D-3 
Table D.l: Hydrogen peroxide concentration in water in tbe ultrasonic cell during ozonation 
(0,014 mg S-I), sonicatioD (57 W) and OZODatioD combined with sonication 
Time Hydrogen peroxide concentration (mg L-1) 
(min) 
ozone _Is ozone/uls 
° 
0,00 0,00 0,00 0,00 0,00 0,00 
5 0,00 0,00 0,04 0,04 0,03 0,03 
10 0,00 0,00 0,05 0,05 0,08 0,07 
20 0,01 0,00 0,Q7 0,08 0, 17 0,18 
40 0,00 0,00 0, 13 0,13 0,35 0,36 
60 0,00 0,00 0,21 0, 19 0,48 0,47 
D.4 MAss BALANCES 
Mass balances were performed to compare ozone decomposition over 45 min during ozonation, ozonation 
combined with sonication, ozonation combined with hydrogen peroxide and ozonation combined with 
sonication and hydrogen peroxide. Ozone decomposition was determined from the difference between the 
amount of ozone generated and the unreacted ozone in the system at the termination of the 45 min 
experiment. Ozonation was performed at an ozone production rate of 0,014 mg s·\ sonication at an 
acoustic power of 57 Wand hydrogen peroxide was added so as to prepare a 40 mg L·1 solution. Ozone and 
hydrogen peroxide roncentrations were measured after the 45 min ozonation period except in solutions with 
added hydrogen peroxide (40 mg L· I ) where only ozone concentration was measured. The upper limit of the 
hydrogen peroxide concentrations that can be measured using the DMP method (described in Section 5.4.1) 
is 4,08 mg L·1. 
The amount of ozone generated in 45 min was determined from the reaction of ozone with a 16000 mg L·\ 
potassium iodide solution in the ultrasonic cell and back titration with sodium thiosulphate as described in 
Section 5.4.2. The potassium iodide solution was sufficiently concentrated to ensure that aU the ozone 
reacted and none passed through the ultrasonic cell . The amount of ozone generated in each mass balance 
(ozonation; ozonation combined with sonication; ozonation combined with hydrogen peroxide and 
ozonation combined with sonication and hydrogen peroxide) is recorded in Table D.4, statistical analysis of 
the data is reported in Table F.26 in Appendix F. The amount of ozone generated in 45 min was measured 
for each mass balance as fluctUations in daily conditions s1ightly aJtered the daily ozone production rate. 
Appendis. 0 OZONE EXFERIMENTAL DATA 
Table 0.4: Amoont of ozone generated in 4S miD in the ozonatioD. ozonation combined with 
sonication, ozoDation combined with hydrogen peroJ.ide and ozonation combined with sonication 
and hydrogen peroDde mass balance experiments 
Ozooe mass (mg) 
ozone ozone/uls ozoneJB1<h ozone/uls/Bl<h 
55,8 55,1 53,5 54,2 
55,5 48,8 47,9 49,9 
52,5 47,9 45,7 48,3 
45,1 
average 54,6 49,2 49,0 ~,8 
The amount of ozone generated in 45 min by the Sorbios ozone generator should have been the same for the 
same experimental conditions irrespective of when an experiment was performed. Fluctuations in daily 
conditions were, however, found to slightly alter the daily ozone production rate. The overall mean and 
standard deviation of the amount of ozone generated in 45 min as reported in Table 0.4 was 50,8 and 
3,82 mg L· l , respectively (Table 5.26). Thus the % error in the daily amount of ozone generated in the 
Sorbios ozone generator is 7,5 %. The % error is calculated by dividing the standard deviation by the mean 
of the samples. The % error within the ozone generated in each mass balance is 3,3; 8,5; 8,3 and 6,1 0/0, 
respectively, for the ozonation, ozonation combined with sonication, ozonation combined with hydrogen 
peroxide and ozonation combined with sonication and bydrogen peroxide mass balances. 
The total amount of unreacted ozone at the termination of an experiment was determined from the dissolved 
ozone in solution in the ultrasonic cell, the gas-phase ozone in the gas spaces in the system and the ozone 
that had passed through the system and reacted in the potassium· iodide gas traps. The gas spaces in the 
system consists of the volume of the incoming gas line, the space above the liquid in the ultrasonic cell and 
the outgoing gas line to the potassium iodide traps. The volume of the gas spaces in the system is recorded 
in Table D.7. 
Table 0.5: Volume of the gas spaces in tbe ultrasonic cell experimental setup 
incoming gas line 
outgoing gas line 
gas space in the ultrasonic cell 
total volume 
Volume 
(mLl 
17 
21 
460 
498 
Appendix D OZONE EXPERIME!'ro'TAL DATA D-5 
The residual ozone concentration in solution in the ultrasonic cell was measured according to the indigo 
calorimetric method described. in Section 5.3.2. The ozone that passed through the u1trasonic cell and 
reacted with the potassium iodide in the iodide traps, as shown in Figure 5.8 in Section 5.2, was measured 
according to the sodium thiosulphate method described. in Section 5.3.2. The ozone concentration in the 
incoming gas line was determined separately from the concentration in the gas space in the u1trasonic cell 
and the outgoing gas line since ozone is transferred into solution where it can react or decompose. The 
ozone concentration in the incoming gas line is thus higher than the ozone concentration in the outgoing 
gas line since ozone is used up in solution. The ozone concentration in the incoming gas line was 
calculated from the amount of ozone generated over 45 min (recorded in Table 0.4), the total volume of gas 
that passed through the system at a flow rate of 2,4 mL s'] and the volume of the incoming gas line (17 mL). 
The ozone concentration in the gas space in the u1trasonic cell and the outgoing gas line was calculated 
from the amount of ozone reacted in the potassium iodide traps at the termination of the experiment 
(recorded in Table 0.6), the total volume of gas that passed through the system at a flow rate of 2,4 mL S-I 
and the volume of the space in the u1trasonic cell and the outgoing gas line (481 mL). Thus the total 
amount of ozone in the gas phase was calculated from the sum of the ozone in the incoming gas line and 
that in the gas space in the u1trasonic cell and the outgoing gas line. 
The ozone and hydrogen peroxide measurements in water after ozonation, ozonation and sonication, 
ozonation and hydrogen peroxide and ozonation, sonication and hydrogen peroxide for 45 min are recorded 
in Table 0.6. Experiments were performed in duplicate. Statistical analysis of the data recorded in 
Table 0 .6 is reported in Table F.27 in Appendix F. 
Table D.6: Ozone and bydrogen peroxide measurements in water after tbe 45 min ozonation, 
ozonation combined witb sonication, ozonation combined with bydrogen peroxide, and ozonation 
combined with sonication and bydrogen peroxide mass balance experiments 
Ozone mass (mg) 
<WIne ozone/uls ozone/B2~ ozone/ulsIH l O2 
dissolved ozone in solution 0,64 0,60 0,21 0,23 0,00 0,08 0,04 0,05 
gas·phase ozone 4,18 4,18 3,36 3,36 3,30 3,30 3,07 3,07 
ozone in the Kl traps 47,09 46,73 35,56 36,25 35.25 35,27 34,37 33,8' 
total 51,91 51,51 39,13 39,84 38,55 38,65 37,48 36,96 
Hydrogen peroxide concentration (mg L-t ) 
<WIne ozone/uis 
Hz02 in solution 0,00 0,01 0,24 0,25 
E 
ATRAZINE EXPERIMENTAL DATA 
Data for the atrazine experiments discussed in Chapter 8 is recorded in this Appendix. Atrazine chemistry 
during sonication, ozonation and in the presence of hydrogen peroxide is discussed in Section E. I . Mass 
balances are also reported in Section E.l. The investigation of atrazine degradation during sonication, 
ozonation and in the presence of hydrogen peroxide is reported in Section E.2. Identification of degradation 
products is detailed in Section E.4. 
E.} A TRAZINE CHEMISTRY 
Atrazine concentration in a standing 5 mg L'I solution was measured over 3 h as a degradation control at 
time periods of 0; 45; 90; 135 and 180 min. Atrazine concentration was also measured in a 5 mg L'l 
atrazine solution when oxygen was sparged through the solution at a flow rate of 6 mL 5.1. Experiments 
were performed in triplicate. The data reported in Table E.l is presented in Figure 8. I in Chapter 8, 
statistical analysis of the data is reported in Table F.28 in Appendix F. 
Table E.1 : Atrazine concentration in a 5 mg L-l atrazine solution in the ultrasonic cell o,:er 3 h in 
tbe absence of ultrasound without and ,rith oxygen sparging 
Time Atrazine concentration (mg L-1) 
(miD) control control witb oxygen sparging 
0 5,4 5,1 5,9 4,8 4.9 4,9 
45 5.3 4,9 5,6 5,0 4,7 4,9 
90 5,0 5,1 5,5 4.9 4,5 5,1 
135 5,2 5,0 4,9 4,9 5.2 5,2 
180 5,4 5.5 5,2 4,8 4,9 4.7 
Atrazine chemistry during ozonation is detailed in Section E. l .l and with hydrogen peroxide addition in 
Section E.l .2. Mass balances are reported in Section E.l .3. 
E.l.1 Ozone 
The dissolved ozone concentration in a 5 mg L-1 atrazine solution was measured during ozonation, 
ozonation combined with sonication (at an acoustic power of 57 W), ozonation combined with hydrogen 
AppendiI E ATRAZIN'E EXPERIMEl'IiAL DATA E-2 
peroxide (40 mg L· I ) and ozonation combined with sonication and hydrogen peroxide. The atrazine 
solution was saturated with ozone by sp.nging the ozone/oxygen gas stream through the diffuser of the 
ultrasonic cell. The ozone concentration in the 2,4 mL s'\ gas stream was 5,7 mg L·I and the ozone 
production rate was 0,014 mg S-I (Table 5.6 in Section 5.2). Dissolved ozone concentration in the solutions 
was measured at time periods of 0; I; 2,5; 5; 10; 20; 40 and 60 min. Duplicate experiments were performed 
for each time period of the different experimental conditions. The dissolved ozone concentrations measured 
in a 5 mg L- I atrazine solution cluring ozonation, ozonation combined with sonication, ozonation combined 
with hydrogen peroxide and ozonation combined with sonication and hydrogen peroxide is recorded in 
Table E.2. The data recorded in Table E.2 is presented in Figure 8.4 in Chapter 8, statistical analysis of the 
data is reported in Table F.29 in Appendix F. 
Table £.2: Dissolved ozone concentration in a 5 mg L' I atrazioe !OIution during ozoDation 
(0,014 mg 5.1), ozonation combined with sonication (S7 W), OZODaOOn combined witb bydrogen 
peroxide (40 mg L-I) and ozonation combined witb sonication and bydrogen peroxide 
Time Ozone concentration (mg L'I) 
(min) 
OZODe ozonelub ozoneIB2~ ozone/ulsIH2~ 
0 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 
0,31 0,27 0,37 0,34 
2,5 0,56 0,58 0,41 0,38 0,23 0,24 0,14 0,12 
5 0,71 0,67 0,35 0,40 0,40 0,36 0,23 0,20 
10 0,58 0,58 0,37 0,35 0,39 0,32 0,16 0,18 
20 0,58 0,56 0,36 0,33 0,48 0,36 0,17 0, 12 
40 0,57 0,59 0,40 0,36 0,3 1 0,35 0,10 0,08 
60 0,43 0,50 0,35 0,3 1 0,43 0,35 0,05 0,04 
E.1.2 Hydrogen peroxide 
Hydrogen peroxide fonnation in a 5 mg L-1 atrazine solution was investigated during ozonation, sonication 
and when ozonation was combined with sonication. Ozonation was performed at an ozone production rate 
of 0,014 mg S'I and sonication at an acoustic power of 57 W. Hydrogen peroxide concentration was 
measured at time periods of 0; 5; 10; 20; 40 and 60 min. Duplicate experiments were perfonned for each 
time period. The data recorded in Table E.3 is presented in Figure 8.6 in Chapter 8, statistical analysis of 
the data is reponed in Table F.30 in Appendix F. 
Appendix E ATRA21N£ EXPERIM£I'fTAL DATA E-J 
Table E.3: Hydrogen peroside concentration in a 5 mg L- I atrazine solution in tbe ultrasonic cell 
during ozonation (0,014 mg S i), sonication (57 W) and ozonation combined witb sonication 
Time Hydrogen peroside concentration (mg L- I ) 
(min) 
..... e "Is ozone/uls 
0 0,00 0,00 0,00 0,00 0.00 0,00 
5 0,18 0,18 0,03 0,Ql 0,23 0,20 
10 0,34 0,34 0,04 0,05 0,40 0,41 
20 0,67 0,67 0,09 0,09 0,73 0,69 
40 1,17 1,18 0,19 0,16 1,21 1.22 
60 1,62 1,66 0,20 0,19 1,53 1,61 
E.l.3 Mass balances 
Mass balances in a 5 mg L-' atrazine solution were perfonned over a 45 min period of ozonation, 02onation 
combined with sonication, ozonation combined with hydrogen peroxide and 02onation combined wim 
sonication and hydrogen peroxide to relate atrazine degradation to ozone decomposition_ The amount of 
ozone decomposed or consumed in the oxidation of atrazine is calculated from the difference between the 
amount of ozone generated in 45 min and the unreacted ozone in the system at the termination of the 
experiment. The total amount of unreacted ozone consists of the dissolved ozone in solution in the 
ultrasonic cell, the ozone in the gas spaces of the system and the 020ne that passed through the system and 
reacted with the potassium iodide in the gas traps. Ozonation was performed at an ozone production rate of 
0,014 rug S-I , sonication at an acoustic power of 57 Wand hydrogen peroxide was added so as to prepare a 
40 mg L- I solution. Ozone and hydrogen peroxide concentrations were measured after the 45 min period 
except in solutions with added hydrogen peroxide where only ozone concentration was measured The 
upper limit of the hydrogen peroxide concentrations that can be measured using the OMP method 
(described in Section 5.4 .1) is 4,08 mg L- I . 
The measurement of the amount of ozone generated in 45 min for the different mass balances is detailed in 
Section 0 .4 and summarised in Table 0 .4. The measurement of unreactod ozone in the '}'stem at the 
termination of me experiments is also detailed in Section 0 .4. The ozone, hydrogen peroxide and atrazine 
measurements in a 5 mg L-I atrazine solution after 45 min of ozonation, ozonation combined 'with 
sonication, ozonation combined with hydrogen peroxide and ozonation combined with sonication and 
hydrogen peroxide are recorded in Table E.4. Statistical anal)'sis of the data recorded. in Table 0 .4 is 
reponed in Table F.31 in Appendix F. The mass balance measurements in water are recorded in Table 0 .6 
in Section 0 .4. 
H 
SUMMARY OF RADICAL REACTIONS 
Free radical reactions are repeated from the literature chapters to facilitate the discussion and comparison of 
the reactions occurring sonication, ozonation and with the addition of hydrogen peroxide. 
H.t RADICAL REACTIONS: ULTRASOUND 
H 2 0 -+ HO· + H" 
Scheme 2.1 
Ho + HO' -+ H 2 0 
HO' + HO' ..... H 10 1 
Ho + Ho -+ H2 
Scheme 2.2 
H" + O2 -+ HO; 
HO; + H" -10 H20 Z 
HOi + HO; -+ H2 0 2 + O2 
Scheme 2.3 
HO' + HOi -+ H 2 0 + 0 1 
HO · + HO· - H, O + O· 
O' + H 20 -+ H 20 1 
HO' + HO' -+ H 2 + O 2 
Scheme 2.4 
[a] 
[b] 
[cl 
[a] 
[b] 
[cl 
[a] 
[bl 
[cl 
Id] 
Appendix H RADICAL REAcnONS 
H.2 RADICAL REACTIONS: OZONE 
initiatio n step 
0 3 + HO- ~ 0;- + HO; 
Propagation steps 
HOi ~ 0 ; - + HT 
0 3 + Oi - ~ O j + O2 
0 ; - + R I- !& HOj 
HO; ~ HO' + O 2 
0 3 + HO' ~ HO~ 
HO~ !t HOi + O 2 
Temlination steps 
HO; + HO; ...... H20 2 + 20 3 
HO; + HO; ...... H 2 0 2 + 0 3 + O 2 
Scheme 3.3 
pK. = 4,8 
k , = 1,6 X 109 rvr' 5' \ 
k2 = 5,2 X 1010 M"" ' 5,1 
k3 = 1, 1 X 103 M-I s" 
14= 2 x l o9 M ' S'I 
k3 = 2,8 X 104 Ml 50 \ 
H.3 RADICAL REACTIONS: OZONE/HYDROGEN PEROXIDE 
Initiation steps 
0 3 + HO- ~ Oi- + HO; 
HOi ~ Oi - + H+ 
H 20 2 ~ HOi' + H+ 
0' 0 3 + HOi ~ 0 ;- + HOi 
Propagation steps 
ki=70 M I so l 
pK. = 4 ,8 
pK = 11,8 
H-2 
raj 
[b] 
[cl 
[dJ 
[eJ 
[I] 
[g] 
[hJ 
[i J 
[a] 
[bJ 
[cl 
[dJ 
0 3 + 0 ;- ~ 0 ; - + O2 kl = 1,6 X 109 M'I sol le] 
0 ; - + H+ !! HO; k2 = 5,2 X 10 10 MO' so l [f] 
HOj ~ HO' + O2 k3 = 1, 1 X 103 M-I 5,1 [g] 
0 ) + HO' ~ O 2 + HOi ~ = 2,0 X 109 M"' S' l [h] 
HO' + H 2 0 2 ~ 0;- + H2 0 + H+ k3 = 2,7 X 107 M-I 5, 1 [i] 
HO' + HOi ~ 0 ;- + H2 0 ~ = 7 , 5 X 109 M ' I S· l U1 
Tennination steps 
HO' + B ~ B~ 
0 3 + B ~ BoD! 
Scbeme3.6 
[kJ 
OJ 
Appendix E ATRAZIl'iE EXPERlMEl'fl'AL DATA E-4 
Table E.4: Ozone, hydrogen peroxide and atrazine measurements in as mg Lol atrazine solution 
after tbe 45 miD ozonation, ozonation combined with sonication, ozonation combined witb bydrogen 
peroxide and ozonation combined with sonication and bydrogen peroxide ma.ss balance experiments 
dissolved ozone in solution 
gas-phase ozone 
ozone in the K.I traps 
total 
H~ in solution 
initial concentration 
final concentration 
ozone 
0,16 0,23 
3,82 3,82 
43 ,42 45,56 
47,40 49,61 
ozone 
1,14 1,25 
ozone 
5,2 
3,8 
5,1 
3,8 
E,2 ATRAZINE DEGRADATION 
Ozone mass (mg) 
ozone/uls ozone/H20z 
0,12 0, 11 0,14 0,22 0,20 
2,89 2,89 3, 18 3, 18 3,18 
30,65 30,34 33,61 34,39 33,68 
33,66 33,34 36,93 37,79 37,06 
Hydrogen peroxide concentration (mg Lot) 
ozone/uls ozonelH20z 
1,41 1,36 
Atrazine concentration (mg LO') 
ozone/uls 
5,3 
3,4 
5,0 
3,5 
6,4 
3,7 
6,5 
3,5 
6,0 
3,8 
ozooe/uls/lbOz 
0 ,02 
2,67 
29,42 
3Z,11 
6,1 
4 ,0 
0,01 
2,67 
29,64 
32,32 
6,1 
3,9 
Atrazine degradation was investigated during sonication. ozonation and with hydrogen peroxide. 
Degradation during sonication is reported in Section E.2 .1, during ozonation in Section E.2.2 and with 
hydrogen peroxide in Section E.2.3. 
E.2.1 Ultrasound 
The ultrasonic degradation of atrazine was investigated by sonicating solutions of initial concentration 
5, 10 and 20 mg L'\ in the ultrasonic cell for 3 h at an acoustic power of 57 w. Atrazine concentration was 
measured using the HPLC method described in Section 5.4.4 . The data reported in Table E.5 is presented 
in Figure 8.7 in Chapter 8, statistical analysis of the data is reported in Table F.32 in Appendix F. 
Appendix £ ATRAZZIIo'E EXPERlMEr."fAL DATA E-5 
Table £.5: AtraziDe concentration in a 5, 10 and 20 mg L-1 atraziDe solution during sonication in 
the ultrasooic cell at an acoustic power of 57 W over 3 h 
Time Atrazine concentration (mg L-1) 
(min) 5 mg L-1 solution 10 mg L-1 solution 20 mg L I solutioo 
0 5,4 5,9 6,1 10,9 11,9 11,5 20,9 20,4 20,6 
45 5,6 5,0 4,8 10,7 11,0 10,9 21 ,8 20,3 21 ,5 
90 5,1 4,7 4,9 10,6 10,7 11,2 20,6 20,1 20,0 
135 4,9 4,7 4,2 10,6 9,8 10,5 21 ,6 20,5 21 ,1 
180 4,2 4,4 4,2 10,1 9,6 9,6 20,8 18,8 18,4 
A concentration of 5 mg L-1 was used as the initial concentration for all further atrazine experiments. 
Atrazine degradation is through a radica1 reaction process. The degradation of atrazine in the presence and 
absence of oxygen was investigated by sparging oxygen and nitrogen at a flow rate of 6 mL S·I through the 
atrazine solution during sonication. The data reported in Table E.6 is presented in Figure 8.8 in Chapter 8, 
statistical analysis of the data is reported in Table F.34 in Appendix F. 
Table E.6: Effect of oxygen and nitrogen sparging on the degradation of atrazme in a 5 mg L ' l 
atrazine solution during sonication in the ultrasonic cell at an acoustic po""er of 57 W over 3 h 
Time Atrazine concentration (mg L ' l) 
(min) nitrogen oxygen 
0 5,1 5,1 5,6 5,3 5,3 5,2 
45 5,4 5,1 4,9 4,8 4,9 5,0 
90 4,6 4,9 5,6 4,7 4,7 4.8 
135 5,4 5,4 4,9 4,5 4,5 4,6 
180 4,7 4,8 5,3 4,1 3,8 4,2 
E.Z.2 Ozone 
The atrazine degradation during ozonation in the ultrasonic cell was investigated by measuring atrazine 
concentration over 3 h al ozone production rates of 0,003; 0,014; 0,030 and 0,047 mg S'l (Table 5.6 in 
Chapter 5). Atrazine concentration was measured al time periods of 0; 15; 30; 45 ; 90; 135 and 180 min. 
Experiments were performed in duplicate. The data for the degradation of atrazine during ozonation is 
recorded in Table E. 7. The data recorded in Table E.7 is presented in Figure 8.9 in Chapter 8, statistical 
analysis of the data is reported in Table F.35 in Appendix F. 
AppendiI. E ATIlAZINE ExPEIUMEJ'ITAL D ATA 
Table E.7 : Atrazine concentration in a 5 mg L-1 atrazine solution during ozonation over 3 b at 
ozone production rates of 0,003; 0,014; 0,030 and 0,047 mg 5-1 
Time Atrazine concentration (mg L-1) 
(min) 100 V I30V 150 V 170V 
0 5,0 4.9 5,2 5,1 5,0 5,2 5,3 4,9 
15 4,7 4,7 4,7 4,7 4,1 4,1 3,8 3,5 
30 4,4 4,4 4,5 4,5 3,6 3,9 2,3 2,3 
45 4,2 4,0 4,2 4,4 2,5 2,8 1,8 1,8 
90 3,5 3.5 3,8 3,4 2,3 2,1 1,0 0,9 
135 3,1 3,2 3,0 3,0 1,3 1,4 0,5 0,7 
180 2,8 2,7 2,4 2,5 0,9 1,0 0,1 0,2 
Alrazine degradation was also investigated when ozonation of a 5 mg L-1 atrazine solution was combined 
with sonication at an acoustic power of 57 W. Atrazine concentration was measured at time periods of 
0; 15; 30; 45; 90; 135 and 180 min. Experiments were performed in triplicate. The data fOT the 
degradation of auazine during ozonation combined with sonication in Table E.8. The data recorded in 
Table E.8 is presented in Figure 8.11 . statistical analysis is reported in Table F.36 in Appendix F. 
Table E.8 : Atrazioe coocentratioD in a 5 mg L-1 atrazine solution during ozonation o\-'er 3 b at 
ozone production rates of 0,003; 0,014; 0,830 and 0,047 mg 5-1 combined witb sonication at an 
acoustic power of 57 W 
Time Atrazine concentration (mg L-1) 
(min) 100 V I30V 
0 4.8 4,8 5,3 5,0 4,9 
15 3,9 3,8 4,2 4.4 4.5 
30 3,7 3,7 3,9 4,1 4,0 
45 3,5 3,4 3,3 3.4 3,5 
90 3.2 3,2 2, 1 2,3 2,3 
135 2.9 3.0 1,7 2,0 1.9 
180 2.8 2,5 1,3 1,6 1.7 
Time Atrazine concentration (mg L-1) 
(min) 150 V 170 V 
0 5.3 5,2 5,1 4,9 4.9 4,9 
15 3,7 3,7 3,7 2,9 3,2 3.0 
30 2,8 2.9 2,7 2,0 2,2 2.2 
45 2,3 2,2 2,2 1,4 1,5 1.5 
90 1,6 1,6 1,8 0,5 0,5 0.5 
135 1,0 I, I I, I 0, I 0, 1 0,2 
180 0,6 0,6 0,6 0,0 0.0 0.0 
E-7 
Comparisons of atrazine degradation in a 5 mg L-1 atrazine solution during ozonation at ozone production 
rates of 0,003; 0,014; 0,030 and 0,047 mg 5.1, sonication at an acoustic power of 57 W and ozonation 
combined. with sonication are shown in Figure E.1 to Figure E.4 . 
6 
::; • 
"'s • 5 • c 
0 4 ~ 
• • 
" 
• 
• uls c 
• 3 g 
0 
u 
• 0 ,003 mg/ s 03 
• uls;O,OO3 mg/ s 03 
• 2 .~ 
;\: 
0 
0 30 ro 150 
''''' 
210 
Time (min) 
Figure Lt: Atrazine degradation in as mg L·1 atrazine solution over 3 b during ozonatioD 
(0,003 mg S' I), sonication (57 W) aDd ozonation combined witb JODieatioD 
I 
6 I 
::; • 
~s 
• • c 
0 • ~ 4 " 
• 
." 
• g 3 -
 0 
u 
• 
• 
• u ls 
• 0 ,014 mg/s 03 
... uls;D,014 mg/ s 03 
• 2 . ~ -
 ;\: 1 -
 0 
0 ro 120 150 
''''' 
210 
Time (min) 
Figure E.2 : Atrazine degradation in a 5 mg L" atmine solution o\-er 3 b during ozonatioD 
(0,014 mg 5,1), sonication (57 W) and ozonation combined with sonication 
Appendix E ATR.A2Pf£ Exp£RIM:£NTAL DATA 
6 
• 
5 
• ... • 
.§.4 • I • • uls g ~. 0 r • 0,030 mg/ s 03 u 3 a • uls;O,030 mg/ s 03 it 2 r 
0 I 
0 30 00 
"" ."" 
"., .00 2.0 
Time (min) 
Figure E.3: Atrazine degradation in a 5 mg L-l .trazine IOlution en'er 3 b during ozonation 
(0,030 mg 8 1), sonication (57 W) and ozonation combiaed. with sonication 
6 
• uls 
• 0,047 mg/ s 03 
• uls;0,047 mg/ s 03 
I : r-~ ~: --:--. --------!-. -----.J 
~ 3 r 
• ~ 2 
it 0 ____ '_~ ______ ::lI.===_~. _ ___' 
o 
."" "" 
.80 2.0 
Time (m in) 
Figure E.4: Atrazine degradation in a 5 rng L·I atrazine solution over 3 h during ownation 
(0.047 mg SI), sonication (57 W) and ownation combined with sonication 
E.2.3 Hydrogen peroxide 
The oxidation of organic pollutants during water treatment with advanced oxidation processes such as 
ultrasound or ozone is enhanced with the addition of hydrogen peroxide due to the generation of additional 
free racticaIs. The molar ratio of ozone to hydrogen peroxide for optimal atrazine oxidation is reported in 
literature to vary between 1,5 and 3 (Beltr.in et al., 1998: Meijers et al., 1993: Paillard et al .. 1991). The 
hydrogen peroxide concentration used in the atrazine investigation was selected based on the atrazine 
degradation obtained over 45 mID in solutions during ozonarion with different hydrogen peroxide 
Appendix E ATRAm'iE EXPERIMENTAL DATA E-. 
concentrations. A hydrogen peroxide concentration of 27 rng L-1 is equivalent to a 2: 1 molar ratio of ozone 
to hydrogen peroxide at an ozone production rate of 0,014 mg sol. A 5 mg L- I atrazine solution was 
ozonation i.n the ultrasonic cell at an ozone production rate of 0,014 mg sol with hydrogen peroxide 
concentrations of 0, 1, 5, 10, 25 and 50 mg L-1• The different hydrogen peroxide concentrations were 
prepared by dilution from a 3 072 mg L-1 peroxide stock solution that had been titrated against potassium 
pennanganate. The experiments were performed in duplicate. The fraction of undegraded atrazine after 
ozonation for 45 min with the different bydrogen peroxide concentrations is presented in Figure E.5. The 
data presented in Figure E.5 is recorded i.n Table E.9, statistical analysis of the data is reported in 
Table F.37 in Appendix F. 
lOO 
eo • • • • c 
• . ~ • • II eo • 
~ 
• ~ 
i;. .., 
• ~ 
c , 
'" 20 
0 
0 10 20 3J .., 
'" H202 concentrat ion (mg/ L) 
Figure E.5: Fraction of' undegraded atr azine in a 5 mg L-1 atrazine solution after ozonation 
(0,014 mg 5-1) f'or 4S min with hydrogen peroxide concentrations of' I , S, 10, 25 and 50 mg L-1 
Table E.9 : Fraction of' undegraded atrazine iD a 5 mg L-1 atrazine solution afte r ozonation 
(0,014 mg 5-1) for 45 min with hydrogen peroxide concentrations of 1, 5, 10, 25 and 50 mg L-1 
8,0, Atraziue concentration (mg L-I ) 
concentration Initial Final % Initial Final % 
(mg L-l ) cone CODC undegraded cone cone undegraded 
0 5,0 3,' 78 4,4 3,7 84 
4,9 3,3 67 4,7 3,2 68 
5 4,5 3,0 66 4,9 3,2 65 
10 4,8 3, I 64 5,0 3,2 64 
25 4,5 3,7 82 4,7 3,6 77 
50 4,' 3,9 80 5,6 4,0 71 
Appendis E ATRAlJNE EXPElUMEIIIIAL D ATA E-IO 
Alrazine degradation over a period of 45 min was greatest. shown in Figure E.S, when ozonation was 
performed with a hydrogen peroxide concentration of 10 rng Lol . A hydrogen peroxide concentration of 
40 rng L- I was thus used for ozonation and sonication experiments over a period of 3 h. Ozonation was 
performed at an ozone production rate of 0,014 mg s-I and sonication at an acoustic power of 57 W. 
Alrazine concentration was measured at time periods of 0; 15; 30; 45 ; 90; 135 and 180 min with hydrogen 
peroxide addition, hydrogen peroxide combined. with sonication, bydrogen peroxide combined with 
ozonation and hydrogen peroxide combined with both sonication and ozonation. The experiments were 
perfonne<t in triplicate. The data recorded in Table E.IO is presented in Figure 8.14 in Chapter 8, statistical 
analysis of the data is reponed in Table F.38 in Appendix F. 
Table E.IO : Atrazine concentration in a 5 mg Lol atraziDe JOIutiOD with 40 mg L-t hydrogen 
peroxide, hydrogen peroxide combined with JOmcation (57 W). hydrogen peroxide combined with 
ozonation (0,014 mg 1.1) and bydrogen peroxide combined with lODieation aDd ozonation 
Time Atranne concentration (mg LOl) 
(min) 8,0, ulsIH10t 
0 ' ,9 5,2 5,0 5,0 5,0 
' 5 4,8 5,1 5,0 4,4 4,4 
90 4,8 4,5 ' ,7 4,1 4,0 
135 4,5 4,9 4,9 3,3 3,3 
180 4,6 4,8 4,5 3,0 2,8 
Time Atrazine concentration (mg LOI) 
(min) OYll,o, ulslOJllhOt 
0 5,5 5,4 5,7 5,2 5, 1 5, I 
15 4,4 4,5 4,1 4,4 4,2 4, I 
30 3,9 3,8 3,8 4, I 4,0 4,0 
45 3,7 3,3 3,4 3,9 3,7 3,5 
90 2,' 2,7 2,7 3,0 2,8 2,7 
135 2, 1 2,1 2,2 2,8 2,5 3,0 
180 1,4 1,7 1,7 2,7 2,6 2,7 
Comparisons of atrazine degradation in a 5 mg L-1 atrazine solution during sonication with and without 
hydrogen peroxide, ozonation with and without hydrogen peroxide and sonication combined with ozonation 
with and without hydrogen peroxide are shown in Figure E.6 to Figure E.8. 
Appendix E ATRAZIl'Io"E EXPERIMENTAL D ATA E-II 
6 
~s D 0 
• • g ! 24 0 --0 g • H202 
c 
• 
... 
o uls • 3 g 
• uls/ H202 0 
" • 2 c 
.~ 
ii: 
0 
0 30 El) 90 120 150 11ll 210 
Time (min) 
Figure E.6: Atrazloe degradation in a 5 mg L-1 atrazine solution witb 40 mg L-1 hydrogen 
perorld~ sonication (57 W) and !lOoication combined with bydrogen peroIide 
6 f 
~s E • 
c • • 
,g 4 r 
'E 
• 3 g 
0 
" • 2 -•  ~ 
ii: , • 
o ------------__________________ ~ 
o 
'" 
120 150 11ll 210 
Time (min) 
• H202 
~ 03 
• 031H202 
Figure E_7: Atrazine degradation in a 5 mg L-1 atrazine solution witb 40 mg L-1 bydrogen 
peroIide. monation (0.014 rng S-I) and ozonation combined witb bydrogen peroxide 
Appendix E A~'E EXPERlMEl\"TAL D ,-\TA 
o eo 
• 
-':---,lIi,,,-__ ->< 
120 150 160 
Time (min) 
210 
• H202 
X uls/ 03 
E-12 
'" uls/ 03/ H202 
Figure E.8: Atrazine degradation in a 5 mg Cl solution with 40 mg L-1 hydrogen peroxide, 
sonication (57 W) combined with ozonation (0,014 mg S-l) and sonication combined with 
ozonation and hydrogen peroxide 
E.3 PRODUCT IDENTInCATION 
The HPLC method described in Section 5.4.4 was used to measure atrazine CODCenlIation during the 
degradation experiments discussed in Section 8.2. Product peaks appeared on the HPLC chromatograms as 
the concentration of atrazine decreased. The chromatogram of an unreacted 5 mg L- t atrazine solution is 
presented in Figure E.9. 
=1 if I 11 11 
=>- 11 
li 1\ I \ ~. -_ ,I ___ !, ~ L,-./-r ~: I __ , 
. , 
" 
Figure E.9: HPLC chromatogram of a 5 mg L-t atrazinc solution 
The retention time of atrazine on the chromatograms for the HPLC analysis system, as shown in Figure E.9, 
is approximately 6,5 min, although some variation occurred on a daily basis. The blimp at a retention time 
of 8 min was an impurity in the atrazinc feedstock (97 %) since it was evident on all chromatograms. even 
for a new atrazine feedstock obtained from Sanachem. 
Appendix E ATRAZIJIro'E EXP£RIME1<oiAL DATA E-l3 
Gas chromatography-mass spectrometry (GC-MS) was used for the identification of degradation products. 
Analysis of atrazinc samples after 3 h of sonication, ozonation and sonication combined with ozonation was 
performed at the CSIR aDd the Rand Afri.kaans University (RAU). GC-MS analysis at the CSIR was 
performed on a Varian 2000 instrument with Saturn software. GC-MS analysis at RAU was performed on a 
Saturn 3 ion trap instrument with a J& W DB I column. The flow rate of ultrapure helium. the carrier gas, 
was 0,94 mL min- I . a split ratio of 13: I helium to sample was used. The injection sample volume was 3 f.ll.. 
and the temperature was ramped from 40 QC to 270 QC at a rate of 10 QC min-} (Vorster. 2000). The 
GC-MS chromatograms (analysis performed at RAU) of the sonication,. ozonation and sonication combined 
with ozonation samples are presented in Figure E.IO. 
11'18:< 
l 
I 
'" 
'''1 
1 
I 
, 
<0. 
b . bb 
OB' 
'3.'3'3 
I 
'" 9 .99 
ea, 
13 . 33 
Ieee 
1&.66 
(a) ultrasound 
I 
'BB 
13.33 
(b) IYLOne 
I 
Il1(1e 
l b.bb 
12ee 
2f1.1I11 
I 
12(111 
Appeom E ATRA2lNE EXPERIMENTAL DATA 
, 
'" 
" .''3 '" 'J . '3'J 
'OB 
11 .''3 
(c) ultrasouadlozooe 
"'. ZB.BB 
£-14 
Figure £.10: GC-MS cbromatograms of a 5 mg L-1 atrazine solution after sonication (57 W), 
ozonatioo (0,047 mg 5-1) and sooicatioo combiocd with 07,onatiOn for 3 b 
F 
STATISTICAL ANALYSIS 
Statistics is a valuable tool for engineers and scientists to use in the development of new products or the 
investigation of operation variables to optimise product and process quality (Kyle, 1995). Good engineering 
practise combines statistical thinking and tools with sound engineering theory. Statistical tools include 
experimental design to identify a data collection strategy, descriptive statistics such as sample mean and 
standard deviation, regression models to indicate correlation between variables and tests indicating 
variability and uncertainty (Vining, 1998). 
lmp:lnant statistics of a test sample are the sample mean x and the sample standard deviation, s. The 
sample mean is a measure of the central tendency of the data, the average value, and the standard deviation 
a measure of the spread or variation in the sample data. it reflects the average distance between a particular 
point and the sample mean (Kyle, 1995). The sample mean and standard deviation are defined as 
x = 
s = 
l" n :E X i 
i=1 
where n is the sample size and Xi a sample point (Vining, 1998). 
[F. I] 
[F.2] 
The confidence interval of a calculated statistic is the range around the calculated value, at a particular 
confidence level (typically 95 or 99 %), that contains the true value. A confidence level of 95 % indicates 
that the range will contain the true value 95 % of the time, The width of the confidence interval is 
determined by the level of confidence (the higher the required confidence the wider the interval) and the 
variability of the data (the higher the data variability the wider the interval), The confidence interval 
- s 
x ± I r¥..n. ----;= 
, ,n [F.3 ] 
is defined as the interval with 100 x (1 - 0.) % confidence (a is 0,05 for a 95 ,% confidence) of a population 
mean where i is the sample mean, $ the sample standard deviation, n the sample size and t<¥, ..rl the value 
Appendix F STATISnCAL ANALYSIS F-2 
from a I-table that corresponds to the right-hand tail of 0)2 for a l-distribution with dl degrees of freedom 
(Vining, 1998). The term fr is defined as the standard error of the estimate of the population mean. 
The adequacy of a regression model of a set of experimental data is determined by the coefficient of 
determination, R2 (Vining, 1998). The coefficient of determination is the proportion of total variability 
explained by the regression model and has a va1ue 0 !5: If :s; 1. R2 is close to 1 for models that fit 
experimental data well and close to 0 for models that fit experimental data poorly. 
= 1 [F.4] 
where SSlCfDl is the overall variability in data, SSrwg the variability explained by the regression model and 
SSru the variability left unexplained and usually attnOuted to error (Vming, 1998). The values of SS/OlQJ and 
SS,..... are calculated from the measured dependent variable y . the variable mean y and the prediction of the 
variable using the regression model y (Vining, 1998). 
ss~ [F.S] 
" - , SS_ = E (y, -yl 
~, 
[F.6] 
Statistical analysis performed in the investigation includes the calculation of means and standard deviations, 
regression models to determine relationships between variables and experimental designs to determine the 
interaction between independent variables on the response of the dependent variable. Tables of calculated 
sample means, 95% confidence limits of the means and standard deviations are recorded. in this appendix. 
Regression models and experimental designs are discussed in the appropriate chapters. 
Statistical analysis of experiments perfonned for equipment characterisation is presented in Section F. l and 
for the analytical procedures in Section F.2. Statistical analysis of the ultrasound experiments is presented 
in Section F. 3, ozone experiments in Section FA and atrazine experiments in Section F.5. 
F_l EQUIPMENT CHARACTERISATION 
Statistical analysis oftbe experimental data for the characterisation of equipment (recorded. in Appendix A) 
is presented in this section. 
Appendix F STATISTICAL ANALYSIS F-3 
F.l.l Evaluation of machining tbe surface of a born tip 
Table F.I: Statistical analysis of tbe formation of bydrogen peroxide during sonication witb a new 
and a machiDed horD tip (Table A.I; Figure A.I) 
Transducer New horn tip 
vibration validN mean -95 % confidence +95 % confidence std deviation 
()lm) (mg L " ) limit (mg L") limit (mg L-I ) (mg L " ) 
0 3 0,00 0,000 
2 3 0,02 0,01 0,03 0,003 
4 3 0.04 0,02 0,06 0,009 
6 3 0,06 0,04 0.08 0,009 
8 3 0,08 0,Q7 0,09 0,005 
10 3 0.10 0,09 0,10 0,003 
11 3 0,10 0,09 0, 11 0,004 
12 3 0,11 0,10 0,12 0,003 
Machined bOrD tip 
0 3 0.00 0,000 
2 3 0,Q2 0,00 0,03 0,006 
4 3 0,05 0,Q3 0,06 0,004 
6 3 0,Q7 0,05 0,08 0,005 
8 3 0,08 0,07 0,09 0,003 
10 3 0,10 0,09 O, lO 0,002 
11 3 0,11 0,10 0,12 0,004 
12 3 0,11 0,10 0,10 0,004 
F.1.2 Acoustic power measurement 
Table F.2 : Statistical analysis of tbe measured acoustic power of tbe sonication experiments 
(Table A.3) 
S~m 
Time valid N mean -95 % confidence +95 % confidence std deviation 
(min) (W) limit (W) limit (W) (W) 
0 3 24,6 23,7 25,5 0,36 
3 24,4 24,0 24,8 0,17 
2 3 24.4 24,0 24,8 0,17 
3 3 24,3 23,6 24,9 0,25 
4 3 24,3 23,6 24,9 0,25 
5 3 24,3 23,6 24,9 0,25 
6 3 24,2 23,8 24,5 0,15 
7 3 24,2 23,8 24,5 0,15 
8 3 24,1 23,8 24,4 0,12 
9 3 24,0 23,6 24,3 0,15 
10 3 23,9 23,3 24,4 0,23 
AppeDd.iI: F STAnsTICAL ANALYSIS F-4 
Table F.2 cont. 
8~m 
Time valid N mean -95 % coDfidence +95 ~. coafideoce std deviation 
(min) (W) Umit (W) limit (W) (W) 
0 3 39,8 39,1 40,6 0,29 
3 39,4 39,0 39,8 0,17 
2 3 39,3 38,6 40,1 0,29 
3 3 39,3 38,6 39,9 0,25 
4 3 39,2 38,6 39,9 0,25 
5 3 39,2 38,6 39,9 0,25 
6 3 39,2 38,4 39,9 0,29 
7 3 39,2 38,4 39,8 0,29 
8 3 39,1 38,7 39,5 0,17 
9 3 39,2 38,4 39,9 0,29 
10 3 39,0 38,4 39,7 0,25 
11 J1m 
0 3 57,7 53,7 61 ,7 1,61 
3 57,6 53,9 61 ,3 1,49 
2 3 57,4 54,0 60,9 1,40 
3 3 57,3 53,5 61 ,1 1,53 
4 3 57,1 53 ,4 60,8 1,49 
5 3 56,8 53,0 60,6 1,52 
6 3 56,5 52,4 60,6 1,67 
7 3 56,2 51,6 60,9 1,86 
8 3 56,2 51 ,8 60,5 1,76 
9 3 55,6 50,7 60,5 1,97 
10 3 55.5 51 ,0 60,0 1,80 
Table F.3 : Statistical analysis of the measured temperatures in SOO mL water in an insulated 
"'essel during sonication at transducer displacement amplitudes of 5, 8 and 11 Ilm (Table A.5; 
Figure A.3) 
5~m 
Time valid N mean -95 % confidence +95 % cODfidence std deviation 
(min) ('Cl limit (0C) limit (0C) (0C) 
0 3 23,7 23 ,6 23,9 0,06 
3 24,2 23 ,8 24,5 0,15 
2 3 24,8 24,6 25,0 0, 10 
3 3 25,4 25,1 25,7 0,12 
4 3 26,0 25,8 26,1 0,06 
5 3 26,6 0,00 
6 3 27,1 27,0 27,3 0,06 
7 3 27,8 27,6 27,9 0,06 
8 3 28,3 28,1 28,5 0,10 
9 3 28,9 28,8 29,1 0,06 
10 3 29,4 29,1 29.8 0, 15 
12 3 30,3 29,7 31,0 0,25 
I. 3 31,3 30,5 32.0 0,31 
16 3 32,1 31.3 32,8 0.31 
Appendix F STATISI1CAL ANALYSIS F-5 
Table F.3 cont. 
5~m 
Time valid N ...... -95 % confidence +95 e;. confidence std deviation 
(min) ("Cl IimU ("Cl IimU (OCl (OCl 
18 3 32,8 32,1 33 ,S 0,26 
20 3 33,S 32,7 34,2 0,29 
25 3 35,S 34,3 36,8 0,49 
30 3 37,S 36,1 38,8 0,55 
35 3 39,3 38,0 40,6 0,53 
40 3 41 ,0 39,6 42,4 0,57 
8~m 
0 3 23,9 23,6 24,2 0,11 
3 24,6 23 ,9 25,3 0,31 
2 3 25,6 24,7 26,S 0,36 
3 3 26,6 25,7 27,S 0,36 
4 3 27,6 26,S 28,6 0,41 
5 3 28,5 27,6 29,4 0,36 
6 3 29,4 28,4 30,4 0,40 
7 3 30,4 29,4 31,4 0,40 
8 3 31,3 30,S 32,2 0,35 
9 3 32,3 31 ,6 33,0 0,30 
10 3 33,2 32,3 34,0 0,35 
12 3 35, 1 34,2 36,0 0,36 
14 3 36,9 36,0 37,8 0,36 
16 3 38,7 37,8 39,7 0,38 
18 3 40,6 39,7 41 ,S 0,36 
20 3 42,4 41 ,S 43,3 0,36 
25 3 46,6 45,7 47,5 0,35 
30 3 50,1 49,9 51,S 0,32 
35 3 54,S 53,6 55,S 0,38 
40 3 58,3 58,1 58,4 0,06 
11 J.lm 
0 3 23,5 22,2 24,7 0,50 
I 3 24,7 23 ,6 25,9 0,45 
2 3 26,1 24,7 27.5 0,56 
3 3 27,S 26,2 28,7 0,50 
4 3 28,9 27,5 30,3 0,56 
5 3 30,3 29,2 31,5 0,45 
6 3 31,7 30,7 32,7 0,40 
7 3 33, 1 32,1 34,1 0,40 
8 3 34,4 33.6 35,3 0,35 
9 3 35,8 35,0 36,7 0,35 
10 3 37,2 36.2 38,2 0,40 
12 3 39,9 39,3 40,6 0,25 
I' 3 42,6 41 ,8 43,3 0.3 1 
16 3 45, 1 44,S 45,6 0,25 
18 3 47,6 47, 1 48,2 0,21 
20 3 SO, I 49,9 50,3 0,10 
25 3 55,9 55,6 56,2 0,11 
30 3 61 ,4 61 ,1 61 ,7 0,12 
35 3 66,4 66.1 66,7 0,12 
40 3 70,8 0,00 
Appendix F S TATISTICAL ANAU'SIS 
F.1.3 Volume experiments with tbe low intensity ultrasonic born 
Table F.4 : Statistical analysis of the formation of bydrogen peroxide in water volumes of 100 to 
1 000 mL during sonication witb tbe low intensity ultruonic born (Table A. 7; Figure A.4) 
lOOmL 
Time valid N mean -95 % confidence +95 % confidence std deviation 
(min) (mg L" ) limit (mg L-
 '
 ) limit (mg L" ) (mg L" ) 
0 2 0,00 0,000 
2 2 0,11 0,06 0,16 0,006 
4 2 0,19 0,17 0,21 0,002 
6 2 0,27 0,10 0,44 0,019 
8 2 0,32 0,21 0,45 0,013 
10 2 0,37 0,000 
15 2 0,46 0,34 0,58 0,013 
20 2 0,55 0,47 0,64 0,009 
200mL 
0 2 0,00 0,000 
2 2 0,07 0,00 0,16 0,0 11 
4 2 0, 11 0,05 0,18 0,007 
6 2 0,15 0,13 0,17 0,002 
8 2 0,19 0,16 0,22 0,004 
10 2 0,23 0,20 0,26 0,004 
15 2 0,27 0,16 0,39 0,013 
20 2 0,33 0,21 0,45 0,013 
JOOmL 
0 2 0,00 0,000 
2 2 0,04 0,01 0,08 0,004 
4 2 0,08 0,05 0,11 0,004 
6 2 0,10 0,00 0,20 0,011 
8 2 0,13 0,12 0,15 0,002 
10 2 0,15 0, 10 0,20 0,006 
15 2 0,20 0,05 0,35 0,017 
20 2 0,23 0,000 
400mL 
0 2 0,00 0,000 
2 2 0,05 0,03 0,07 0,002 
4 2 0,07 0,000 
6 2 0,10 0,03 0,16 0,007 
8 2 0,11 0,09 0,13 0,002 
10 2 0,13 0,000 
15 2 0, 17 0,14 0,21 0,004 
20 2 0,20 0,15 0,25 0,006 
SOOmL 
0 2 0,00 0,000 
2 2 O,QJ 0,000 
4 2 0,04 0,00 0,07 0,004 
6 2 0,07 0,00 0,13 0,007 
8 2 0,08 0,06 0,09 0,002 
10 2 0,10 0,000 
15 2 0,14 0,07 0,21 0,()()7 
20 2 0,17 0,14 0,20 0,004 
AppendiI. F STAnsTICAL ANALYSIS F-7 
Table F.4 conL 
600mL 
Time valid N mean -~ -10 confidence +95 % confidence std deviation 
(mid) (mg L -') limit (mg L -') limit (mg L -') (mg L -' ) 
° 
2 0,00 0,000 
2 2 0,03 0,00 0,06 0,004 
• 2 0,04 0,000 
6 2 0,05 0,Q3 0,Q1 0,002 
8 2 0,07 0,000 
10 2 0,09 0,02 0,17 0,009 
IS 2 0,11 0,08 0, 15 0,004 
20 2 0,15 0,06 0,23 0,009 
700mL 
° 
2 0,00 0,000 
2 2 0,02 0,00 0,09 0,007 
4 2 0,04 0,02 0,06 0,002 
6 2 0,05 0,02 0,08 0,004 
8 2 0,06 0,02 0,09 0,004 
10 2 0,08 0,04 0, 11 0,004 
IS 2 0,10 0,08 0, 12 0,002 
20 2 0,12 0,04 0,20 0,009 
800mL 
° 
2 0,00 0,000 
2 2 0,02 0,00 0,05 0,004 
4 2 0,03 0,00 0,06 0,004 
6 2 0,04 0,00 0,09 0,006 
8 2 0,06 0,00 0,14 0,009 
10 2 0,Q1 0,00 0, 14 0,007 
IS 2 0,10 0,000 
20 2 0,13 0,08 0, 18 0_006 
900mL 
° 
2 0,00 0,000 
2 2 0,03 0,00 0,06 0,004 
• 2 0,03 0,00 0,07 0,00' 
6 2 0,05 0,02 0,08 0,004 
8 2 0,05 0,00 0,10 0,006 
10 2 0,07 0,03 0,10 0,004 
IS 2 0,10 0,000 
20 2 0,11 0,07 0,14 0_004 
1 000 mL 
° 
2 0,00 0,000 
2 2 0,01 0,00 0,03 0,002 
• 2 0,03 0,000 
6 2 0,04 0,00 0,Q1 0,004 
8 2 0,04 0,00 0,09 0,006 
10 2 0,06 0,00 0, 14 0,009 
IS 2 0,08 0,00 0,21 0,014 
20 2 0,10 0,000 
Appendix F STATIS'DCAL ANALYSIS 
Table F.S : Statistical analysis of the temperature mcuuremeats in water volumes of 100 to 
1 000 mL during sonication with the low intensity ultrasonic horn (Table A.I0) 
lOOmL 
Time validN mean -95 -/0 confidence +9S % confidence std deviation 
(mid) ('C) limit ('C) limit (0C) ('C) 
° 
3 24,8 23,8 25,9 0,42 
2 3 36,6 35,4 37,7 0,47 
4 3 45,0 44,1 45,9 0,35 
6 3 52,1 49,9 54,2 0,86 
8 3 57,9 55,1 60,8 1,14 
10 3 64,2 60,4 67,9 1,51 
200mL 
° 
3 24,2 23 ,1 25,3 0,44 
2 3 30,5 28,5 32,5 0,81 
4 3 35,3 32,9 37,7 0,95 
6 3 39,8 37,8 41,8 0,79 
8 3 44,5 42,8 46,2 0,69 
10 3 48,6 46,7 50,5 0,78 
300mL 
° 
3 24,0 23 ,4 24,7 0,25 
2 3 28,0 26,9 29,1 0,46 
4 3 31,4 29,7 33,2 0,70 
6 3 34,7 33,0 36,4 0,70 
8 3 38, 1 36,6 39,6 0,60 
10 3 41 ,0 38,6 43,4 0,95 
400mL 
° 
3 25,6 24,6 26,6 0,40 
2 3 28,4 27,3 29,5 0,46 
4 3 31,1 30,3 31,9 0,32 
6 3 33,7 33,0 34,4 0,26 
8 3 36,0 35,3 36,7 0,26 
10 3 38,2 37,3 39,1 0,35 
500mL 
° 
3 26,2 25,5 26,9 0,30 
2 3 28,7 27,2 30,1 0,58 
4 3 31,0 30,0 32,0 0,40 
6 3 32,9 32,0 33,9 0,38 
8 3 35,2 34,4 36,0 0,32 
10 3 37,1 36,3 38,0 0,35 
600mL 
° 
3 26,3 25,3 27,4 0,42 
2 3 28,5 27,4 29,7 0,46 
4 3 30,3 29,6 31, 1 0,29 
6 3 32,2 31,3 33,2 0,38 
8 3 33,8 32,5 35,0 0,49 
10 3 35,7 34,4 37,0 0,52 
Appendix F STAnsTICAL ANALYSIS F-9 
Table F.S CODt. 
700mL 
Time valid N mean -95 % coDfideace +9! % COIlfidence std deviation 
(min) ("C) limit (0C) limit ("C) ("C) 
0 3 26,5 26,0 27,0 0,20 
2 3 28,2 21,7 28,8 0,21 
4 3 29,5 28,4 30,6 0,44 
6 3 31,1 30,8 31 ,5 0,15 
8 3 32,7 32,1 33,2 0,23 
10 3 34,1 33,9 34,2 0,06 
800mL 
0 3 24,5 23,8 25,2 0,30 
2 3 25,7 25,4 26,0 0,12 
4 3 27,3 26,7 27,8 0,21 
6 3 28,7 28,4 29,0 0,12 
8 3 29,9 29,6 30,2 0,12 
10 3 31,1 30,8 31,5 0,15 
900mL 
0 3 24,4 23,5 25,2 0,35 
2 3 25,7 24,9 26,5 0,32 _ 
3 26,5 25,_ 27,6 0,46 
6 3 28,2 27,3 29,0 0,35 
8 3 29,4 28,1 30,7 0,53 
10 3 30,5 29,4 31,6 0,44 
1000 mL 
0 3 25,9 25,7 26,1 0,10 
2 3 27,0 26,8 27,2 0,10 _ 
3 28,2 27,6 28,7 0,21 
6 3 29,2 28,6 29,9 0,25 
8 3 30,3 29,5 31 ,0 0,31 
10 3 31,3 30,5 32,0 0,31 
F.1.4 Acoustic power reduction due to tbe ultrasonic ceU lid design 
Table F.6: Statistical analysis of the comparison of meuured temperatures during sonication with 
tbe lo~' intensity horn in SOO mL of water in the ultrasonic cell and in a beaker (Table A.12; 
Figure A.7) 
mtruonic cell 
Time valldN me .. -95 % confidence +95 % confidence std deviation 
(min) ('C) limit COC) limit (0C) (0C) 
0 3 26,9 26,6 27,2 0,12 
2 3 27,2 26,8 27,6 0, 17 _ 
3 27,7 27,2 28,2 0,20 
6 3 28,1 27,7 28,5 0,17 
8 3 28,_ 28,1 28,8 0, 15 
10 3 28,7 28,6 28,9 0,06 
Appendix F STATISTICAL ANALYSIS F-\O 
Table F.6 cont. 
Beaker 
Time valid N ..... ·95 % confidence +95 % coafidence std deviation 
(min) ("Cl limit ('Cl limit ('Cl ('Cl 
0 3 26,2 25,S 26,9 0,30 
2 3 28,7 27,3 30,1 0,58 
4 3 31,0 30,0 32,0 0,40 
6 3 32,9 32,0 33,9 0,38 
8 3 35,2 34.4 36,0 0,32 
10 3 37,1 36,3 38,0 0,35 
F.l.S Rotameter calibration 
Table F.7 : Statistical analysis of the rotameter calibration for oxygen at regulator prasures of 200 
and 300 kPa (Table A.15; Figure A.I0) 
lOO kPo 
Rotameter valid N mean -95 % coDfidence +95 % cODfideace std deviation 
settlog (mL ,.1) limit (mL ,-I) limit (mL ,-I) (mL 5. 1) 
0 1 0,0 
1 4 0,7 0 ,7 0,8 0,01 
2 4 1,7 1,7 1,8 0,03 
3 4 3,4 3,4 3,5 0,05 
4 4 6,0 6,0 6,1 0,05 
5 4 8,8 8,7 8,8 0,04 
6 4 11,1 10,9 11,3 0,13 
7 4 13,6 13,2 13,9 0,21 
8 4 16,3 15,9 16,8 0,29 
9 4 19,1 18,5 19,6 0,34 
10 4 21 ,3 20,8 21 ,8 0,30 
300 kP. 
0 I 0,0 
4 0,7 0 ,7 0,7 0,01 
2 4 1,7 1,7 1,8 0,03 
3 4 3,6 3,5 3,7 0,06 
4 4 6,2 6 ,1 6,2 0,03 
5 4 8,4 7,8 9,1 0,25 
6 4 11 ,5 11, 1 11 ,8 0,24 
7 4 13,9 13,3 14,5 0,38 
8 4 16,3 IS ,S 17, 1 0,48 
9 4 19,7 18,9 20,6 0,52 
10 4 21 ,7 21 ,1 22,2 0,37 
AppendiJ: F STATIS11CAL ANALYSIS F-U 
Table F.8: Statistical analysi. of the rotameter calibration for nitrogen at a regulator pressure of 
200 kPa (fable A.16) 
Rotameter Flow rate 
setting valid N mean -95 % confidence +95 % confidence std deviation 
(mL , -') limit (mL , -') limit (mL S-I) (mL S-I) 
2,0 5 2,0 1,9 2,1 0,05 
3,0 5 4,2 4,2 4,2 0,04 
3,5 4 5,1 5,1 5,2 0,03 
3,7 6 6,0 6,0 6,0 0,03 
4,0 5 6,5 6,5 6,5 0,01 
F.l.6 Ozone generator calibration 
Table F.9: Statistical analysis of the dissolved ozone concentration in solution during ozonation 
with ozone generated by a Sorbios OZOne generator in a 2,4 mL ._1 oxygen gas stream at voltage 
settings between 100 and 200 V (Table A.I'; Figure A.ll) 
Voltage Ozone conceDtration 
(V) validN mean -95 % confidence +95 % confidence std deviation 
(mg L -') limit (mg L-t ) limit (mg L -') (mg L-I ) 
100 3 7,7 7,2 8,3 0,23 
110 3 14,7 14,0 15,3 0,27 
130 3 34,5 32,6 36,3 0,74 
150 3 73,2 72,6 73 ,7 0,23 
170 3 111,9 110,3 113,6 0,68 
200 3 178,4 172,5 184,3 2,39 
F,2 ANALYTICAL PROCEDURES 
Statistical analysis of information recorded in Appendix B is presented in this section. Means and standard 
deviations were not calculated for the hydrogen peroxide calibration curves since three stock solutions with 
slightly different concentrations were used to prepare each calibration curve. The calibration curves were 
not prepared from triplicate injections of the same solution. 
Table F.IO: Statistical analysis of the atrazine calibration cUn'e (Table 8.5; Figure 8.5) 
Atrazine Peak area 
concentration valid N mean -95 % confidence +95 % confidence std deviation 
(mg L ol) (units) limit (units) limit (units) (units) 
0 3 0 0 
1 3 218170 181214 256727 15 199 
2 3 467178 406 292 528065 24510 
5 3 1117528 912004 1 32.3 052 82 734 
10 3 1933690 1675624 2 191 755 103 885 
20 3 4 167816 3827919 4507713 136827 
Appendix F STAnsnCAL ANALYSIS F-12 
F_3 ULTRASOUND EXPERIMENfS 
Statistical analysis of the ultrasound experimental data recorded in Appendix C is presented in this section. 
F_3_1 Dissolved oxygen concentration 
Table F.Il : Statim.cal analysis of the dissolved oxygen concentration in oxygen-saturated water 
during sonication, without sonication (control) and during nitrogen spargiDg without sonication 
(Table C.I; Figure 6.1) 
Control 
Time validN mean -95 % confidence +95 % confidence std deviation 
(min) (mg L-1) limit (mg rl) limit (mg L-1) (mg L" ) 
0 3 39, 1 35,0 43,2 1,65 
5 3 36,1 32,7 39,5 1,39 
15 3 32,7 32,0 33,8 0,42 
30 2 29,8 28,5 31, 1 0, [4 
45 2 28,9 27,6 30,2 0, [4 
60 3 27,8 27,2 28,3 0,21 
90 3 25,8 25,0 26,5 0,31 
120 2 25,2 23 ,9 26,S 0, [4 
150 3 24,6 23,9 25,3 0,26 
180 2 24,0 2[ ,5 26,5 0,28 
Sonication 
0 3 38,2 33,6 42,8 1,85 
5 3 29,2 26,S 3[,9 1, 10 
15 3 23,5 2[ ,5 25,5 0,80 
30 3 20,9 20,2 2[ ,6 0,26 
45 2 [8,9 8,0 29,7 1,20 
60 3 [7,4 16,6 [8,[ 0,29 
90 3 [4,3 14,1 14,5 0,10 
120 3 13,2 12,4 14,0 0.32 
150 2 [2 ,[ 10,8 13,4 0,14 
180 JJ ,2 
Nitrogen 
0 3 38.9 33,5 44,2 2, 15 
5 3 3,0 2,6 3,3 0, [5 
15 3 1,8 I,' 2,1 0, 15 
30 3 1,6 1,0 2,1 0,21 
45 0 
60 2 1,6 0.3 2,9 0, 14 
90 0 
120 2 1,9 0,6 3,2 0,14 
150 0 
180 2 1,5 0,8 2,1 0,07 
Appendix F STATISnCAL ANALYSIS F-13 
Table F.12: Statistical analysis of the dissolved oxygen concentration in water saturated with air 
(Table C.2) 
valid N 
40 10.0 
-95 % confidence 
limit (mg L'l) 
9,7 
~ % confidence 
limit (mg L-I ) 
10.3 
std deviation 
(mg L " ) 
0,90 
Table F.IJ Statistical analysis of the dissolved oxygen concentration in nitrogen-, air- and 
oxygen-saturated water during sonication in the ultrasonic ceU at an acoustic power of 57 W 
(Table c.Jj Figure 6.2) 
Control 
Time validN m .... -95 % confideDce +95 % confideace std deviation 
(miD) (mg L" ) limit (mg L 'l) limit (mg L") (mg L") 
0 2 9,7 8,4 11,0 0,14 
5 2 10,8 10,1 11,4 0,07 
10 2 11.0 10,3 11,6 0,07 
20 2 10,3 7.8 12.8 0,28 
40 2 11,8 11,1 12,4 0,07 
Nitrogen 
0 2 0,9 0,2 1,5 0,07 
5 2 0.8 0.00 
10 2 0.8 0,0 2.7 0.21 
20 2 0,9 0,0 2,2 0, 14 
40 2 1.1 0,0 3,0 0.21 
Air 
0 2 11,0 4,0 17,9 0,78 
5 2 12,3 11,6 12,9 0,07 
10 2 12,8 5,8 19,7 0,78 
20 2 13,2 7,4 18.9 0,64 
40 2 13.2 12,5 13,8 0.07 
Oxygen 
0 2 39,9 38,6 41 .1 0.1' 
5 2 40.2 37,7 42,7 0,28 
10 2 39.8 36,6 42,9 0.35 
20 2 39.5 33, I 45,9 0.7 1 
40 2 38.5 32.1 44,9 0,71 
Table F.14: OveraU statistical analysis of the dissolved oxygen concentration in nitrogen-, air- and 
oxygen-saturated 'ft'ater during sonication in the ultrasonic cell (Table C.l; Figure 6.2) 
G ..... valid N mean -95 % confidence +95 % confidence std deviation 
(mg L'I) limit (mg L'l) limit (mg L-I ) (mg L " ) 
control 10 10,7 10.2 11,2 0,73 
nitrogen 10 0,9 0,8 1,0 0, 16 
air 10 12,5 11,8 13, I 0,96 
oxygen 10 39.6 39.1 40.1 0,71 
AppencliI F STA1lS11CAL ANALYSIS F-14 
F.3.2 Hydrogen peroxide formation 
Table F.15 : Statistical analysis of the formatioa of hydrogen peroxide iD Ditrogeo-, air- od 
oxygen-saturated water during sonication in the ultrasonic ceU for 16 h at an acoustic power of 
57 W (Table C.4; Figure 6.3) 
CODtrol 
Time valid N mean -95 % confidence +95 % coafidence std deviation 
(min) (mg L") limit (mg L-1) limit(mg L -') (mg L " ) 
0 3 0,00 0,000 
4 3 0,02 0,000 
8 3 0,04 0,02 0,06 0,008 
12 3 0,05 0,04 0,06 0,003 
16 3 0,07 0,05 0,08 0,006 
20 3 0,09 0,06 0,11 0,010 
30 3 0,11 0,10 0,13 0,007 
60 3 0,15 0,15 0,15 0,001 
120 3 0,26 0,21 0,32 0,023 
240 3 0,38 0,29 0,47 0,035 
480 3 0,77 0,75 0,79 0,008 
960 2 0,92 0,39 1,46 0,060 
Nitrogen 
0 3 0,00 0,00 0,01 0,002 
4 3 0,01 0,00 0,10 0,001 
8 3 0,02 0,01 0,oJ 0,004 
12 3 0,04 0,03 0,05 0,004 
16 3 0,04 0,04 0,05 0,002 
20 3 0,06 0,05 0,07 0,004 
30 3 0,05 0,04 0,06 0,005 
60 3 0,06 0,04 0,Q7 0,006 
120 3 0,06 0,06 0,07 0,002 
240 2 0,09 0,02 0,15 0,007 
480 2 0,09 0,04 0,14 0,005 
960 2 0,08 0,00 0, 19 0,012 
Ai. 
0 3 0,00 0,00 0,01 0,002 
4 3 0,02 0,01 0,02 0,001 
8 3 0,04 0,04 0,05 0,00 1 
12 3 0,05 0,04 0,05 0,003 
16 3 0,07 0,06 0,09 0,006 
20 3 0,09 0,08 0,09 0,002 
30 3 0, 11 0,08 0, 13 0,010 
60 3 0,18 0, 17 0,20 0,006 
120 3 0,28 0,23 0,32 0,020 
240 3 0,34 0,33 0,36 0,006 
480 3 0,78 0,63 0,93 0,061 
960 3 1,03 0,79 1,27 0,097 
Appendix F STAnsTICAL ANALYSIS F-IS 
Table F.IS coot. 
Oxygen 
Time valid N mean -9S -J. confidence +9!5 .,4 coofidence std deviatiOD 
(min) (mg L-' ) limit (mg L-1) limit(mg r') (lOg L-') 
° 
3 0,Q1 0,01 0,02 0,002 
4 3 0,04 0,03 0,06 0,006 
8 3 0,06 0,05 0,08 0,007 
12 3 0,08 0,06 0,10 0,008 
16 3 0,09 0,08 0,10 0,005 
20 3 O,ll 0,10 0,11 0,00 1 
30 3 0,16 0,14 0,17 0,006 
60 3 0,26 0,25 0,26 0,003 
120 3 0,43 0,41 0,45 0,009 
240 3 0,48 0,44 0,52 0,016 
480 3 0,86 0,59 1, I3 0,108 
960 2 1,21 0,38 2,05 0 ,093 
Table F,16 : Statistical aaalysiJ of the formatioll of bydrogen peroxide iD nitrogeo-saturated water 
and the control duriog tonicatioo in the ultruonic cell at an acoustic power of 14 W (Table C.S; 
Figure 6.4; Figure 6.5) 
Control 
Time valid N mean -9S ./e coofideoce +95 -J. toofideKe std deviatioo 
(min) (lOg L" ) limit (lOg L" ) limit (mg L-') (mg L-1) 
° 
3 0,00 0,000 
4 3 0,Q1 0,00 0,Q2 0,004 
8 3 0,02 0,000 
12 3 0,Q3 0,02 0,04 0,004 
16 3 0,04 0,Q3 0,05 0,003 
20 3 0,05 0.04 0,07 0.006 
30 3 0,Q7 0,06 0,08 0,003 
60 3 0,14 0,13 0. 14 0.003 
120 3 0,24 0.23 0.26 0,006 
240 3 0,42 0,40 0,45 0,009 
480 3 0,74 0,67 0,80 0,026 
960 0.98 
Nitrogen 
° 
3 0,00 0,000 
4 3 0,01 0,000 
8 3 0,Q2 0,02 0.Q3 0,002 
12 3 0,03 0,03 0,04 0,003 
16 3 0,05 0,04 0,05 0,002 
20 3 0,05 0,04 0.06 0,004 
30 3 0,06 0,05 0,07 0,004 
60 3 0,05 0,05 0,06 0,003 
120 3 0.06 0,05 0,07 0.003 
240 3 0,Q7 0,06 0,09 0,004 
Appendit. F STATISTICAL ANALYSIS F-16 
F.3.3 Hydrogen peroxide degradation 
Table F.17 : Statinica1 analysis of the degradation of hydrogen peroxide (normalised data) in 
nitrogen-, air- and oxygen-saturated water after sonication in the ultrasonic ceU for 16 h at an 
acoustic power of 57 W (fable C.6; Figure 6.7) 
Control 
Time valid N ..... -95 % confidence +93 % confidence std deviation 
(min) limit Hmit 
0 3 1,00 0,000 
5 3 1,00 0,99 1,01 0,004 
to 3 1,00 0,99 1,02 0,006 
20 3 1,00 0,% 1.05 0,018 
30 3 1,00 0,94 1,05 0,021 
60 3 0,98 0,93 1,04 0,021 
90 2 0,97 0,81 1,14 0,019 
120 2 0,97 0,69 1,25 0,03 l 
150 2 0,95 0,61 1,29 0,038 
180 I 0,95 
Nitrogen 
0 3 1,00 0,000 
5 3 0,97 0,94 1,00 0,012 
to 3 0,97 0,91 1,03 0,025 
20 3 0,95 0,87 1,02 0,030 
30 3 0,95 0,91 0,98 0,015 
60 3 0,89 0,86 0,92 0,014 
90 3 0,86 0,79 0,94 0,030 
120 3 0,84 0,76 0,92 0,032 
150 3 0,81 0,70 0,92 0,046 
ISO 3 0,76 0,73 0,79 0,012 
Air 
0 3 0,00 0,000 
5 3 0,97 0,87 1,07 0,040 
to 3 0,% 0,92 1,01 0,018 
20 3 0,95 0,89 1,01 0,025 
30 3 0,94 0,90 0,99 0,0 18 
60 3 0,89 0,83 0,95 0,024 
90 2 0,87 0,80 0,94 0,008 
120 3 0,82 0,77 0,86 0,018 
150 3 0,79 0,77 0,81 0,009 
180 3 0,76 0,70 0,81 0,023 
OIygen 
0 3 1,00 0,000 
5 3 0,97 0,89 1,05 0,032 
10 3 0,98 0,89 1,06 0,036 
20 3 0,% 0,92 1,00 0,016 
30 3 0,96 0,90 1,02 0,025 
60 3 0,92 0,89 0,94 0,012 
90 3 0,88 0,83 0,93 0,020 
120 3 0,85 0,79 0,92 0,026 
150 3 0,82 0,77 0,86 0,018 
ISO 3 0,78 0,78 0,79 0,003 
Appendix F STATISTICAL ANALYSIS F-17 
F.3.4 Interval experiments 
Table F.18: Statistical analysis of tbe formation of bydrogen peronde in oxygen-saturated water 
and a control in the ultrasonic cell during 15 min of sonication, 15 min without sonication and a 
further 15 min witb sonication (Table CS; Figure 6.8) 
Control 
Time valid N mean -95 % confidence +95 % confidence std deviation 
(min) (mgL-') limit (mg L-I ) limit (mg L- I ) (mg L -' ) 
0 3 0,00 0,000 
5 3 0,04 0,Q3 0,04 0,003 
10 3 0,06 0,05 0,Q7 0,003 
15 3 0,08 0,07 0,08 0,002 
20 3 0,08 0,Q7 0,08 0,003 
25 3 0,08 0,07 0,09 0,004 
30 3 0,08 0,Q7 0,08 0,003 
35 3 0,10 0,10 0,11 0,002 
40 3 0, 12 0,11 0, 14 0,006 
45 3 0,14 0,14 0, 15 0,003 
Oxygen 
0 3 0,0 1 0,00 0,Q3 0,008 
5 3 0,05 0,05 0,06 0,003 
10 3 0,09 0,08 0,10 0,004 
15 3 0, 11 0 ,10 0,12 0 ,003 
20 3 0,11 0,10 0, 12 0,004 
25 3 0,11 0,10 0, 13 0,006 
30 3 0, 11 0,09 0,12 0,006 
35 3 0,13 0,12 0,14 0_003 
40 3 0, 16 0,15 0,18 0,006 
45 3 0, 19 0,18 0,20 0,004 
Table F.19 : Statistical analysis of the formation of hydrogen peronde in water during sonication 
in the ultrasonic cell with 15 min of oxygen saturation, 15 min of nitrogen saturation and a further 
15 min of oxygen saturation (Table C.9; Figure 6.9) 
Time Gas valid N mean -95 % confidence +95 % confidence std deviation 
(min) (mg L-I ) limit (mg L-I ) limit (mg L-I ) (mg L-I ) 
0 0, 3 0,01 0_00 0,01 0,003 
5 0, 3 0,04 0,03 0,05 0,003 
10 0, 3 0,08 0,07 0,09 0,003 
15 Q,fN, 3 0, 11 0,10 0,12 0,004 
20 N, 3 0,13 0_ 12 0, 14 0,004 
25 N, 3 0,14 0,1 3 0, 15 0,003 
30 N,/o, 3 0,15 0,14 0,16 0,004 
35 0, 3 0,16 0,15 0,17 0,004 
40 0, 3 0,18 0,17 0,19 0,004 
45 0, 3 0,20 0,19 0_2 1 0_004 
Appendix F STATISTICAL ANALYSIS F-18 
F-3.5 Commercial bydrogen peroxide experiments 
Table F.20 : Statistical analysis of the formation of bydrogen peroxide in nitrogen-, air- and 
oxygen-saturated 0,28 mg L-1 bydrogen peroxide solutions during sonication in tbe ultrasonic ceU 
for 1 b at an acoustic power of 57 W (Table C.lO; Figure 6.10) 
Control 
Time valid N mean -95 % confidence +95 % confidence std deviation 
(min) (mg L-') limit (mg L·l) limit (mg L-I ) (mg L-') 
0 3 0,28 0,27 0,29 0,003 
5 3 0,28 0,26 0,30 0,008 
10 3 0,30 0,29 0,30 0,003 
20 3 0,33 0,31 0,34 0,004 
30 3 0,25 0,32 0,37 0,009 
60 3 0,40 0,40 0,41 0,003 
Nitrogen 
0 3 0,28 0,26 0,29 0,006 
5 3 0,25 0,25 0,26 0,002 
to 3 0,25 0,24 0,27 0,006 
20 3 0,25 0,23 0,26 0,006 
30 3 0,26 0,24 0,28 0,008 
60 3 0,25 0,24 0,27 0,004 
Air 
0 3 0,27 0,27 0,28 0,003 
5 3 0,28 0,26 0,29 0,006 
10 3 0,29 0,28 0,31 0,007 
20 3 0,33 0,31 0,35 0,009 
30 3 0,34 0,33 0,36 0,006 
60 3 0,41 0,40 0,43 0,007 
Oxygen 
0 3 0,29 0,28 0,30 0,003 
5 3 0,29 0,29 0,30 0_003 
to 3 0,32 0,30 0,35 0,009 
20 3 0.37 0,36 0,37 0_003 
30 3 0,41 0,40 0,43 0,006 
60 3 0,56 0,54 0,58 0,009 
Table F.ll Statistical analysis of tbe formation of bydrogen peroxide in nitrogen-, air- and 
oxygen-saturated 0.43 mg L"I hydrogen peroxide solutions during sonication in the ultrasonic cell 
for 1 b at an acoostic power of 57 W (Table C.ll; Figure 6.11) 
Control 
Time ,'alid N mean -95 % confidence +95 % confidence std de,iation 
(miD) (mg L-I ) limit (mg L-I ) limit (mg L- I ) (mg L-I ) 
0 3 0,43 0,39 0,47 0_017 
5 3 0,45 0,43 0,47 0,008 
to 3 0_45 0,44 0,46 0,004 
20 3 0,48 0,47 9,48 0,001 
30 3 0,50 0,49 0,52 0_007 
60 3 0,58 0,57 0,60 0_005 
Appendix F STA11S11CAL ANALYSIS F-19 
Table F.21 cont. 
Nitrogen 
Time validN mean -95 % confidence +95 % confidence std deviation 
(min) (mg L-') Hmi! (mg L-') Hmi! (mg L-') (mg L-') 
0 3 0,43 0,42 0,45 0,006 
5 3 0,44 0,43 0,45 0,002 
10 3 0,43 0,41 0,46 0,010 
20 3 0,42 0,40 0,45 0,009 
30 3 0,42 0,40 0,43 0,008 
60 3 0,40 0,39 0,41 0,004 
Air 
0 3 0,42 0,42 0,43 0,001 
5 3 0,45 0,44 0,46 0,003 
10 3 0,45 0,43 0,47 0,007 
20 3 0,48 0,46 0,49 0,005 
30 3 0,50 0,49 0,52 0,007 
60 3 0,58 0,54 0,61 0,014 
Oxygen 
0 3 0,44 0,42 0,46 0,007 
5 3 0,47 0,46 0,48 0,004 
10 3 0,47 0,46 0,49 0,005 
20 3 0,50 0,48 0,52 0,008 
30 3 0,52 0,52 0,53 0,004 
60 3 0,62 0,60 0,65 0,010 
Table F.22 : Statistical analysi5 of tbe formation of bydrogen peroxide in nitrogen-. air- and 
oxygen-saturated 0,72 mg L- t bydrogen peroxide solutions during sonication in tbe ultrasonic cell 
for] b at an acoustic power of 57 W (Table C.12; Figure 6.12) 
Control 
Time valid N mean -95 % confidence +95 % confidence std deviation 
(min) (mg L-') limit (mg L-t ) limit (mg L- t ) (mg L-t ) 
0 3 0,73 0,71 0,75 0,008 
5 3 0,72 0,69 0,75 0,011 
10 3 0_73 0,73 0,74 0,003 
20 3 0,75 0,71 0,80 0,020 
30 3 0,79 0,76 0,82 0,0 11 
60 3 0,82 0,8 1 0,82 0,003 
Nitrogen 
0 3 0,73 0,72 0,74 0,004 
5 3 0,71 0,68 0,73 0,009 
10 3 0,69 0,68 0.70 0,003 
20 3 0,68 0,66 0,69 0.005 
30 3 0,68 0,66 0,70 0,007 
60 3 0,63 0,62 0,65 0.004 
Appendix F STA.TIS'I1CAL ANALYSIS F-lO 
Table F.22 cont. 
Air 
Time valid N mean -95 % confidence +95 ~o confidence std deviation 
(min) (mg L-') limit (mg L-l ) limit (mg L-l ) (mg L-') 
0 3 0,71 0,70 0,72 0,004 
5 3 0,73 0,71 0,75 0,009 
10 3 0,75 0,74 0,75 0,002 
20 3 0,76 0,75 0,78 0,005 
30 3 0,78 0,77 0,79 0,003 
60 3 0,81 0,78 0,83 0,009 
Oxygen 
0 3 0,72 0,70 0,74 0,008 
5 3 0,76 0,75 0,78 0,006 
10 3 0,78 0,77 0,78 0,003 
20 3 0,76 0,74 0,78 0_007 
30 3 0,78 0,77 0,79 0,004 
60 3 0,85 0,81 0,88 0,015 
FA OWNE EXPERIMENTS 
Statistical analysis of the ozone experimental data recorded in Appendix 0 is presented in this section. 
F.4.1 Dissolved ozone concentration 
Table F.2J : Statistical analysis of the measured dissolved ozone concentration in water during 
ozonation (0,014 mg S i), ozonation combined witb sonication (57 W), 07,onation combined witb 
bydrogen peroxide (40 mg L-l ) and OZOIlation combined with sonication and bydrogen peroxide 
(Table D.l ; Figure 7.1) 
Ozone 
Time valid N mean -95 % confidence +95 % confidence std deviation 
(min) (mg L -') IiDHt (mg L -') limit (mg L-I ) (mg L-') 
0 2 0,00 0_000 
2 0.]9 0,33 0,44 0,006 
2.5 2 0,85 0,00 1,82 0,108 
5 2 1,20 0,62 1,77 0,063 
10 2 1,34 0,45 2,23 0,099 
20 2 1,13 0,00 2,33 0.134 
40 2 1,08 0.32 1,84 0,084 
60 2 1.01 0,10 1.92 0,101 
Ozone/uhruound 
0 0 0,00 0,000 
I 2 0.47 0,35 0,59 0,013 
2,5 2 0,73 0,25 1,22 0,054 
5 2 0,75 0,63 0,87 0,013 
10 2 0,63 0,21 1,06 0,047 
20 2 0_61 0,36 9,86 0.028 
40 2 0,55 0,43 0,67 0,013 
60 2 0,59 0.47 0,71 0,013 
Appeodix F STATImCAL ANALYSIS F-21 
Table F.23 CODt. 
Ozooe/bydrogeo peroxide 
Time valid N mean -95 % confidence +95 % coofidence std deviation 
(min) (mg L " ) limit (mg L't) limit (mg L't) (mg L " ) 
0 2 0,00 0,000 
5 2 0,00 0,00 0,02 0,001 
10 2 0.00 0,00 0,02 0,002 
20 2 0,01 0,000 
40 2 0,01 0,000 
60 2 0,02 0,01 0,03 0,001 
OzoneJultrasc:MlodlbydrogeD peroxide 
0 2 0,00 0,000 
5 2 0,02 0,0 1 0,03 0,001 
10 2 0,02 0,01 O,QJ 0,001 
20 2 0,02 0,000 
40 2 0,02 0,000 
60 2 0,05 0,00 0,31 0,030 
F.4.2 Ozone decomposition 
Table F.24: Statistical analysis of tbe dissolved ozone CODceDtratioD iD water iD tbe ultrasonic cell 
after a 20 miD saturation period during ozonatioo (0,014 mg s,t), sonication (57 W) and ownation 
combined witb sooicatioo (Table D.2; Figure 7.3) 
Cootrol 
Time valid N mean -95 % coofideoce +95 0/. coalideoce std deviation 
(min) (mg L-I ) limit (mg L't) limit (mg L't) (mg L't) 
0 2 1,51 0,70 2,31 0,090 
5 1,40 
10 2 1,38 1,18 1.58 0,023 
20 2 L25 0,94 1,55 0.034 
40 2 1,34 0,64 2,05 0.078 
Ozone 
0 2 1,13 0,00 2,33 0, 134 
20 2 1.08 0,32 1,84 0,085 
40 2 1.01 0, 10 1,92 0,101 
Ultrasound 
0 0 0,94 0,65 1,24 0.033 
2.5 2 0,45 0,33 0,57 0,0 13 
5 2 0,28 0.21 0.34 0,007 
10 2 0,13 0,12 0, 14 0.001 
20 2 0,06 0.00 0,25 0.022 
OzooeJultra.'IOUnd 
0 2 0,00 0,54 1,75 0,067 
2.5 2 0,53 0,00 1,07 0,060 
5 2 0.52 0,00 1,55 0. 115 
10 2 0,46 0,34 .0,58 0,013 
20 2 0,46 0,43 0,49 0,004 
40 2 0.40 0.22 0.59 0.021 
Appendix F STA.nsnCAL ANALYSIS F·22 
F.4.3 Hydrogen peroxide fonnation 
Table F.2S : Statistical analYlil of tbe formation of hydrogen pero:ude in water in the ultruonic 
eelI during ownation (0,014 mg I-I), !Onication (57 W) and own.tion combined with sonication 
(Table D.4i Figure 7.4) 
OzoDe 
Time valid N meaD -95 % coDfidence +95 0/. confidence std deviation 
(miD) (mg L " ) limit (mg L-l ) limit (mg LOI) (mg L " ) 
0 2 0,00 0,000 
5 2 0,00 0,000 
10 2 0,00 0,00 0,02 0,002 
20 2 0,00 0,00 0,02 0,002 
40 2 0,00 0,000 
60 2 0,00 0,000 
Ultruound 
0 2 0,00 0,000 
5 2 0,04 0,02 0,06 0,002 
10 2 0,05 0,Q3 0,07 0,002 
20 2 0,07 0,04 0,11 0,004 
40 2 0,13 0,09 0,17 0,004 
60 2 0,20 0,Q1 0,33 0,0 15 
OzoneJultruound 
0 2 0,00 0,000 
5 2 0,03 0,00 0,07 0,004 
10 2 0,08 0,04 0,11 0,004 
20 2 0,18 0,08 0,27 0,010 
40 2 0,35 0,26 0,45 0,0 10 
60 2 0,48 0,44 0,51 0.004 
F.4.4 M a ss balances 
Table F.26: Statistical analysis of the measured ozone input in tbe 45 min ownation, ownation 
combined l\'itb sonication, ozonation combined l\'ith hydrogen pero:ude and ozonation combined 
with sonication and hydrogen peroxide mau balance experiments (Table D.4) 
Mass balance Ozone input 
experiment valid N me .. -95 % confidence +95 % confidence std deviation 
(mg) limit (mg) limit (mg) (mg) 
ozone 3 540 6 50, ] 59,1 1,82 
ozoneJuls 4 49.2 42.5 55,9 4.21 
ozone/H~ 3 49,0 38,9 59,1 ' ,07 
ozoneJulsIH2~ 3 50,8 43,2 58,5 3,08 
total 13 50,8 48,5 53,1 3,82 
AppendiI F STA11SI1CAL ANALYSIS F-23 
Table F.27: Statistical aaalysil of tbe ozone and bydrogen peroxide meuuremeats iD water after 
tbe 45 miD ownation, ozonation combined witb sonication, ozonation combined witb bydrogen 
peroxide, aad ozonatioD combined witb sonication aad bydrogen peroxide mus balaace 
experiments (fable D.6) 
OzoD. 
valid N ...... -95 e;. confidence +95 ./. confidence std deviation 
(mg L " ) limit (mg L-1) limit (mg L-t ) (mg L " ) 
dissolved. ozone (mg) 2 0,62 0,42 0,81 0,021 
gas-phase ozone (mg) 2 4,84 0,000 
ozone in KI traps (mg) 2 46,92 44,56 49,27 0,262 
H~ in solution (mg L-1) 2 0,01 0,00 0,Q7 0,007 
OzoDelultruound 
dissolved ozone (mg) 2 0,22 0,07 0,37 0,0 17 
gas-phase ozone (mg) 2 4,55 0,000 
ozone in KI traps (mg) 2 35,91 31,56 40,26 0,484 
H~ in solution (mg Lol ) 2 0,23 0,00 0,49 0,029 
OzoDeIB,o, 
dissolved ozone (mg) 2 0,04 0,00 0,53 0,054 
gas-phase ozone (mg) 2 4,53 0,000 
ozone in KJ traps (mg) 2 35,26 35, 13 35,39 0,0 14 
Ozooe/ultruoundIB20:z 
dissolved ozone (mg) 2 0,04 0,00 0,09 0,006 
gas-phase ozone (mg) 2 4,50 0,000 
ozone in K1 traps (mg) 2 34,10 30,77 37,44 0,371 
F.S A TRAZINE EXPERIM£NTS 
SlalisticaJ analysis of the atrazine experimental dala recorded in Appendix E is presented in this section. 
F.S.l Atrazine chemistry 
Table Fo28: StatisticaJ aaalysis of tbe atrazine concentration in a 5 mg L-I atrazine solution in tbe 
ultrasonic cell o,"er J b in tbe absence of ultrasound witbout and witb oxygen spa rging (Table E.l; 
Figure 8 01) 
Control 
Time nlid N meao -95 % confidence +95 % confidence std deviation 
(min) (mg L-1) limit (mg L-' ) limit (mg L-1) (mg L-' ) 
° 
3 5,5 4,4 6,5 0,42 
45 3 5,3 4,5 6.1 0,3 1 
90 3 5,2 4,5 '5,9 0,27 
IJ5 3 5,0 4,7 5,4 0, 15 
180 3 5,' 5,0 5,7 0,14 
AppendiI F STATISTICAL ANALYSIS F-24 
Table F.28 cont. 
Oxygen 
Time valid N mean -95 % confidence +93 % confidence std deviation 
(min) (m! L·I ) limit (m! L·I ) limit (m! L· I ) (ms L·I ) 
0 3 4,9 4,7 5,1 0,08 
45 3 4,9 4,6 5,2 0,12 
90 3 4,8 4,0 5,7 0,34 
135 3 5,1 4,_ 5,_ 0,21 
ISO 3 4,8 4,5 5,0 0,10 
F.3.1.1 Ozone 
Table F.29 : Statistical analysis of tbe measured dissolved owne concentration in a 3 mg L· I 
atrazine solution during ozonation (0,014 mg . ,1), ozoDatioo combioed witb sonication (37 W), 
ownation combined witb bydrogen peronde (40 mg L'I) and ozonation combined with sollication 
and hydrogen peronde (Table £,2; Figure 8.4) 
Ozone 
Time valid N mean -95 °/0 confidence +95 °/. coufidence std deviatiOD 
(min) {mE L'll limit {mE L·ll limit {!!& L'll {mE L·ll 
0 2 0,00 0,000 
2 0,29 0,00 0,59 0,033 
2,5 2 0,57 0,45 0,69 0,01 3 
5 2 0,69 0,44 0,93 0,027 
10 2 0,58 0,000 
20 2 0,57 0,45 0,69 0,013 
40 2 0,58 0,46 0,70 0,013 
60 2 0,46 0,04 0,88 0,047 
Ozonelultruound 
0 0 0,00 0,000 
2 0,35 0,12 0,59 0,026 
2,5 2 0,39 0,19 0,60 0,023 
5 2 0,38 0,04 0,71 0,037 
10 2 0,36 0,26 0,46 0,011 
20 2 0,34 0, 12 0,56 0.025 
40 2 0,38 0,15 0,62 0,026 
60 2 0,33 0.06 0,60 0.030 
Ozone/bydrogcn peronde 
0 2 O,DO 0.000 
2.5 2 0,24 0, 17 0,3 1 O,D08 
5 2 0,38 0,15 0,62 0,026 
10 2 0,36 0,00 0,81 0,05 1 
20 2 0.42 O,DO 1.17 0.084 
40 2 0,33 0, 11 0,55 0,025 
60 2 0,39 0.00 0,92 0,059 
Ozone/uitruoundlhydrQ2en peroxide 
0 2 O,DO 0,000 
2,5 2 0,13 0,01 0,25 0,013 
5 2 0.2 1 0,00 0,43 0,024 
10 2 0,17 0,09 0,25 0,009 
20 2 0,15 0,00 0,42 0,030 
40 2 0,09 0.00 0,21 0.013 
60 2 0,04 0,02 0,06 0.002 
Appendix F STA.nsnCAL ANALYSIS F-25 
F.5.1.2 Hydrogen peroxide 
Table F.30 : StatiJtical analysis or tbe formation or bydrogen peroxide in a S mg Lo1 atnzine 
solution in the ultr'UOllic ttll during ozonatioo (0,014 mg Si), sonication (57 W) and orooation 
combined with !OBicatioD (Table Eo3; Figure 8.6) 
Ozon. 
Time valid N mean -95 ./. confidence +93 % confidence std deviation 
(min) (mgL" ) limit (mg Lol) limit (mg L" ) (mg Lot) 
° 
2 0,00 0,000 
5 2 0,18 0,14 0,22 0,004 
10 2 0,34 0,32 0,36 0,002 
20 2 0,67 0,63 0,71 0,004 
40 2 1,18 1,14 1,21 0,004 
60 2 1,64 1,38 1,90 0,029 
Ultrasound 
° ° 
0,00 0,000 
5 2 0,03 0,01 0,05 0,002 
10 2 0,04 0,01 0,08 0,004 
20 2 0,09 0,000 
40 2 0,17 0,00 0,38 0,023 
60 2 0,19 0,14 0,25 0,006 
OzoneJultruound 
° 
2 0,00 0,000 
5 2 0 ,22 0,05 0,38 0,019 
10 2 0,40 0,33 0,48 0,008 
20 2 0,71 0,45 0,97 0,030 
40 2 1,22 1,18 1,25 0,004 
60 2 1.57 1,09 2,06 0,054 
F.5.1.3 Mass balances 
Table Fo31: Statistical analysis of tbe ozone, hydrogen peroxide and atnzine measurements in a 
5 mg Lot atrazme !Olution after the 45 min O'LOnation, ozonation combined with sonication, 
O'LOnation combined with bydrogen peroxide and ozonation combined with sonication and hydrogen 
peroxide mass balance experiments (Table E.4) 
o-LODe 
valid N mean -95 % confidence +95 % confidence std deviation 
(mg Lol) limit (mg Lot) limit (mg Lol) (mg Lol) 
dissoh'ed ozone (mg) 2 0,20 0,00 0,64 0,049 
gas-phase ozone (mg) 2 4,84 0,000 
ozone in K.l traps (mg) 2 44,49 30,95 58,02 1.506 
H:z02 in solution (mg Lol) 2 1,20 0,50 1,89 0,078 
initial atrazine (mg Lol) 2 5,16 4,91 5,41 0,028 
final atrazine (mg Lol) 2 3,83 0,000 
Appendix F STA.TIS'I1CAL ANALYSIS F-26 
Table F.31 cont. 
Ozone/ultrasound 
valid N ..... -95 % confidence +95~. confidence std deviation 
(mg L-') limit (mg L-' ) limit (mg L-') (mg L-') 
dissolved ozone (mg) 2 0,00 0,00 0,20 0,045 
gas-phase ozone (mg) 2 4,55 0,000 
ozone in Kt traps (mg) 2 30,61 29,96 31 ,26 0,262 
Hz02 in solution (mg L- l ) 2 1,41 1,29 1,52 0,045 
initial atrazine (mg L- I ) 2 5,12 4,80 5,43 0, 127 
finaJ atrazine (mg L- I ) 2 3,38 3,10 3,66 0,114 
Ozonc/B1Ch 
dissolved ozone (mg) 2 0,18 0,08 0,29 0,041 
gas-phase ozone (mg) 2 4,53 0,000 
ozone in KI traps (mg) 2 33,89 32,81 34,97 0,434 
initial atrazine (mg L- I ) 2 6,33 5,67 7,00 0,270 
final atrazine (mg L-1) 2 3,66 3,27 4,05 0,158 
Ozonc/ult ruoundIB1<h 
dissolved ozone (mg) 2 0,02 0,00 0,05 0,004 
gas-phase ozone (mg) 2 4,50 0,000 
ozone in Kl traps (mg) 2 29,53 28,16 30,89 0,152 
initial atrazine (mg L-1) 2 6,08 5,87 6,28 0,023 
final atrazine (mg L- I ) 2 3,96 3,79 4,12 0,019 
F,5,2 Atrazine degradation 
F_5_2_1 Ultrasound 
Table F_32 : Statistical anal)'sis of tbe atrazinc concentration mcasured in a 5. 10 and 20 mg L-1 
atrazine solution during sonication in tbe ultrasonic cell at an acoustic powcr of 57 W over 3 b 
(Table E.5; Figure 8 _7) 
5mgL-1 
Time nlid N mean -95 % confidcncc +95 % confidcnce std deviation 
(min) (mgL-') limit (mg L-I ) limit (mg L-t ) (mg L-1) 
° 
3 5,8 4,9 6,7 0,36 
45 3 5,1 4,0 6,2 0,44 
90 3 4_8 4.3 5,4 0,21 
135 3 4,6 3,6 5,6 0,39 
180 3 4,3 3_9 4,6 0,15 
10 mg L-1 
° 
3 11 ,4 10,2 12,6 0,48 
45 3 10,8 10,5 11 ,2 0, 16 
90 3 10,8 10, 1 _11,5 0,30 
135 3 10,3 9,3 11 ,4 0,42 
180 3 9,8 9,0 10,6 0,32 
Appendix F STAnsTICAL ANALYSIS F-27 
Table F032 cont. 
20 mgLo1 
Time valid N mean -95 % confidence +95 % confidence std deviation 
(min) (mg L -') limit (mg Lol) Iinllt (mg L " ) (mg L " ) 
0 3 20,6 20,0 21 ,2 0,24 
45 3 21 ,2 19,2 23 ,2 0,80 
90 3 20,2 16,5 21 ,0 0,32 
135 3 21 ,1 19,6 22,S 0,57 
180 3 19,3 16,2 22 ,S 1,27 
Table F.33: Statistical analysis of the initial concentration of the 5 mg Lo1 atrazine !lOlutions used 
in tbe atrazine investigation 
valid N 
66 
mean 
(mg Lol) 
5,2 
-95 % confidence 
limit (mg L-I ) 
5,1 
+95 0/. confidence 
limit (mg L " ) 
5,3 
std deviation 
(rng Lol) 
0,43 
Table F.34: Statistical analysis of tbe effect of oIygen and nitrogen sparging on tbe degradation of 
atrazine in a 5 mg L-t atrazine solution during sonication in tbe ultrasonic cell at an acoustic power 
of 57 W over 3 b (Table E.6j Figure 8.8) 
Nitrogen 
Time valid N mean -95 % confidence +9S % confidence std deviation 
(min) (mg L" ) limit (mg L-I) limit (mg LOI) (mg Lol) 
0 3 5,3 4,5 6,0 0,30 
45 3 5,1 4,6 5,7 0,21 
90 3 5,0 3,8 6,3 0,51 
Il5 3 5.2 4,5 5.9 0,28 
180 3 4,9 4,2 5,6 0,29 
OI)'gen 
0 0 5,3 5,0 5,5 0,08 
45 3 4.9 4,6 5,2 0,12 
90 3 4,8 4,6 4,9 0.05 
135 3 4,5 4,4 4.7 0.06 
180 3 4,0 3,5 4,5 0.20 
F.S.2.2 Ozone 
Table F035: Statistical analysis of tbe atrazinc concentration in a 5 mg Lo1 atrazine solution during 
ozonation at ozone production rates of 0,003; 0,014; 0,030 and 0,047 mg st (Table E.7; Figure 809) 
0,003 mg 5 1 
Time valid N mean -95 % confidence +9S % confidence std deviation 
(min) (mg LOI) limit (mg LOI) limit (mg Lol) (mg L " ) 
0 2 5,0 4,9 5,1 0,01 
45 2 4,1 2,5 5,7 0, 18 
90 2 3.5 3.0 ' 4,0 0,06 
I3S 2 3,2 2,5 3,9 0,08 
180 2 2. 7 2.6 2,8 0.0 1 
Appendix F STAnsTICAL ANALYSIS F-28 
Table F.35 cont. 
0,014 mg S-l 
Time valid N me .. -95 % confideoce +95 % confidence std deviation 
(miD) (mg L " ) limit (mg L-l ) limit (mg L " ) (mg L " ) 
0 2 5,1 _,7 5,6 0,05 
_5 2 _ ,3 3,1 5, _ 0, 13 
90 2 3,6 0,9 6,3 0,30 
135 2 3,0 2,5 3,5 0,05 
180 2 2,5 1,_ 3,5 0,11 
0,030 mg sol 
0 2 5,1 4,0 6,2 0,12 
45 2 2,6 0,6 4,6 0,22 
90 2 2,2 0,9 3,_ 0,14 
135 2 1,3 0,6 2,0 0,08 
180 2 0,9 0,5 1,_ 0,05 
0,047 mg S-I 
0 2 5,1 2,9 7,3 0,25 
_5 2 1,8 1,6 2,0 0,02 
90 2 0,9 0,2 1,_ 0,07 
135 2 0,6 0,0 1,9 0,15 
180 2 0,2 0,0 0,3 0,02 
Table F.36 : Statiltica.l analysis of tbe atrazme conceotratioD iD a 5 mg L-I atrmne solution during 
ozonation over 3 b at ozone production rates of 0,003; 0,014; 0,030 and 0,047 mg 5' and sonication 
at an acoustic power of 57 W (Table E.8; Figure 8.11) 
U1truoundlO,003 mg 51 ozone 
Time valid N mean -95 % confidence +95 % confidence std deviation 
(miD) (mg L-t) limit (mg L-t) limit (mg L-l ) (mg L " ) 
0 2 4,8 4,7 4,8 0,01 
45 2 3.4 2,9 3,9 0,06 
90 2 3,2 3,1 3,3 0,01 
135 2 3,0 2,5 3,4 0.05 
180 2 2,7 0,5 4,9 0.25 
UltruoondlO,Ol4 mg s · OZODe 
0 3 5,1 4,5 5,6 0.21 
45 3 3,4 3,1 3,7 0, 11 
90 3 2,2 1,9 2,6 0,14 
135 3 2,0 1,8 2,1 0,05 
180 3 1,5 l.l 2.0 0.19 
UltrasouDdlO,OJO mg s-t O'.lOne 
0 3 0,0 5,0 5,_ 0.09 
45 3 2,3 2.1 2,_ 0,07 
90 3 1,7 1,4 1,9 O, ll 
135 3 1,1 0,9 1,3 0,08 
180 3 0,6 0,6 0,6 0,01 
Appendi.I F Sn.nsTICAL ANALYSIS F-29 
Table F.36 cont. 
IDtruouDdIO,047 mg sot OZODe 
Time vatidN mean -95 ·1. cODfideoce +95 "I. confidence std deviation 
(mid) (mg L " ) limit (mg Lol) limit (mg Lol) (mg Lol) 
0 3 0,0 4,9 4,9 0,02 
45 3 1,4 1,3 1,6 0,06 
90 3 0,5 0,4 0,5 0,02 
135 3 0,1 0,0 0,2 0,05 
ISO 3 0,0 0,00 
F.5.2.3 Uydrogea peroxide 
Table F.37: Statistical a.aalysis of the fractioa of UDdegraded atrazine id a 5 mg Lot atrazine 
lOIution after OZODahoD (0,0014 mg SI) for 45 miD witb bydrogen peroxide CODCeDtratiOdS of 1, 5, 
10, 25 a.ad 50 mg L-I (Table E.9; Figure £.5) 
B.o. validN mean -95 % confidence +95 'Yo CODfideoee std deviation 
(mg L-I ) (%) limit (0/. ) limit ("10) (-to) 
0 2 8 1 42 120 4,3 
2 68 63 72 0,5 
5 2 66 57 75 1,0 
10 2 64 61 68 0,4 
25 3 79 44 115 4,0 
50 3 76 24 127 5,8 
Table F.38 : Statistical analysis of tbe atrazine concentration id a 5 mg Lot atrazme solution witb 
40 mg L-I hydrogen peroxide, hydrogen peroxide combined with sonication (57 W), bydrogen 
peroxide combined "'itb ozonation (0,014 mg Si) and bydrogen peroxide combined with sonication 
and ozonation (Table E. IO; Figure 8.14) 
B.o. 
Time valid N mean -95 % confidence +95 % confidence std deviation 
(min) (mg Lot) limit (mg Lol) limit (mg L 'I) (mg L " ) 
0 3 5,0 4,7 5,4 0, 14 
45 3 5.0 4,6 5,3 0,14 
90 3 4,7 4,3 5, 1 0, 17 
135 3 4,8 4,2 5,4 0,23 
180 3 4.6 4,3 5,0 0,14 
Ultr8!OUndIB1(h 
0 0 0,0 4,_ 5,2 0,02 
45 2 4.4 4,3 4,5 0,01 
90 2 4,1 3,2 4,9 0,0_ 
135 2 3,3 3.1 3,5 0,02 
ISO 2 2,9 1,4 4,4 0,17 
AppeDdix F STA11STICAL ANALYSIS F-30 
Table F.l8 CODt. 
OzoDelBlOt 
Time valid N me .. -95 ./. cODfideDc:e +9S ./. coafideDCe std dcviatioa 
(miD) (mg L-') limit (mg L-I ) limit (mg L-I ) (mg L-') 
0 3 0,0 5,2 5,8 0,12 
15 3 4,4 3,9 4,8 0,19 
30 3 3,8 3,6 4,1 0,09 
45 3 3,4 3,0 3,9 0,20 
90 3 2,6 2,3 2,9 0,13 
135 3 2,1 2,0 2,2 0,04 
180 3 1,6 1,1 2,1 0,21 
Ultruo,"ullOZODe!B,o, 
0 0 5,2 5,0 5,3 0,05 
15 3 4,2 3,9 4,5 0,12 
30 3 4,0 4,0 4,1 0,03 
45 3 3,7 3,1 4,3 0,22 
90 3 2,8 2,3 3,2 0,19 
135 3 2,7 2,1 3,4 0,24 
ISO 3 2,6 2,5 2,8 0,Q7 
G 
FREE RADICAL REACTIONS 
Free radicals contain an unpaired electron, they are thus extremely reactive and in general exist only for a 
shon time period Free radical chemistry is described in Section G. l and the free radical reactions that 
occur during photolysis and radioJysis and may also occur during sonication are listed in Section 0.2. 
G.t FREE RADICAL CHEMISTRY 
Free radical reactions can be broadly classified as radical combination processes. where two radicals react to 
produce non-radical products in which all electrons are paired, and radical transfer processes which 
generate a new radical that reacts fwther with molecules around it, often propagating a chain of radical 
reactions (Huang et al., 1974). 
Combination or coupling: 
The combination of two radicals, shown in Scheme G. l , leads to bond formation. The process is extremely 
fast and requires little or no activation energy. Radical concentrations in solution are extremely low and 
combination reactions are thus, in general, not as important as transfer reactions with the solvent. Reaction 
conditions, however, are favourable when free radicals are formed in pairs in close proximity to each other 
within the solvent cage (Huang et al., 1974). 
R" + B" - - •• R-B 
ScbemeG.l 
Disproportionarion: 
The transfer of a p-hydrogen atom from one radical to another generating non-radical products is known as 
disprop:>rtionation. The process is energetically favourable since two bonds are fonned for one that is 
broken (Huang et al. , 1974). 
" . / H-C-C + 
/ " 
" ./ H-C-C 
/ " 
Scheme G.2 
" / H-C-C-H 
/ " 
AppeDdix G FRu RADICAL Ib.AcnONS G-2 
Redox reactions: 
Radicals are both formed and destroyed in one-electron transfer redox reactions. Anions are produced 
during the reduction of radicals by transition metal ions while cations., shown in Scheme G.3, are produced 
during oxidation (Huangd al., 1974). 
Displacement: 
D HO· + Fe 
AI· + Cui 
m 
--•• HO· + Fe 
• 
AI" + euO 
ScbcmeG.3 
Displacement s:im.ilar to ionic displacement (SN reactions) rarely occurs in radical chemistry except for the 
abstraction of hydrogen and halogen atoms (Huang et al., 1974). 
R· + A-8 --•• R-A + B' 
ScbemeG.4 
Addition: 
A more stable radical is formed from the addition of a radical to an unsaturated compound, as shown in 
Scheme G.S. Radical polymerisation is initiated when the generated radical reacts with another molecule of 
the unsaturated substrate (Huang et al., 1974) . 
• 
Scheme G.S 
Fragmentation: 
A smaller radical and an unsaturated molecule are formed when a radical breaks down by the scission of a 
bond 13 to the radical centre (Huang et al. . 1974). 
Me 
I , 
Me-C-O 
I 
Me 
• 
Scheme G.6 
Me 
'C--O ' + Me 
Me/ 
AppeadU. G FREE RADIcAL RucnoNS G-3 
Insertion: 
The insertion reaction is peculiar to carbenes and nitrenes. The insertion of methylene (H2C:) into the C-H 
bonds of pentane, as shown in Scheme G.7, produces isomeric hexanes (Huang et al., 1974). 
CH, -(CHi)'-CH, H:!C: 
--"'=<.. CH, - (CHi),- CH(CH,h 
+ 
CH, CH,CHCH,<:H, 
I 
CH, 
ScbemeG.7 
G.2 RADICAL REACTIONS 
The conditions in the cavitation bJbbles during sonication are similar to that initiated. during radiolysis and 
photolysis. Radical reactions occuning during radiolysis and photolysis can thus occur during sonication. 
The reactions proposed by Spinks and Woods that occur during the radiolysis of water include the reactions 
listed in Scheme G.8 (Spinks and Woods, 1976). 
H20 + hv .... Ho + HO" 
H 20 + hv .... H2 + 0 
H" + HO' -+ H2 0 
H" + H" .... H2 
HO· + HO· - H,O, 
H" + HO· - H, + 0 
HO· + HO· + M - H,O + 0 
0 + HO" .... H" + O2 
HO' + H2 .... H 20 + H" 
0 + H20 .... H20 2 
0 + H,O - 2HO· 
Ho + H 20 2 .... H20 + HO' 
HO" + H20 2 .... H20 + HOi 
HOi ;: Oi- + H-
 HO· .. 0·- + H" 
H 20 2 ~ HOi' + H-
 H- + H20 .... HO- + H2 
HO" + H 20 .... H30 '" + 0 
H" + O2 .... HOi 
raj 
[hJ 
[cJ 
[dJ 
[eJ 
[n 
[gJ 
[hJ 
[iJ 
UJ 
[kJ 
[mJ 
[nJ 
[oJ 
[PJ 
Iq] 
['J 
(sJ 
[tJ 
Appendix G FREE RAmCAL RucnoNS 
202 + 2H"' - H 20 2 + O2 
H" + H" - H2 
HO· + HO- - H,O + 0 -
 SCb .... G.8 
[u] 
[v) 
[w] 
The radical reactions JXOIXI5Od by Lii and ccrworkers that occur in a pulse irradiated. Ar-H~ system 
during radiation with a high-pressure xenon lamp are listed. in Scheme G.9 (Lii et al., 1980), 
HO" + HOi - O 2 + H2 O [a] 
HO· + HO· 
- 0 + H2O [b] 
HO· + HO· 
- H20 2 [cl 
O2 + 0 - 0 3 [d] 
H" + O 2 - HOi [e] 
HOi + HOi - O2 + H20 2 [0 
HO" + H20 2 - HOi + H2O [g] 
0 3 + HO" - 0 2 + HOi [b] 
o + HOi - 0 2 + HO" [;] 
H" + HO; 
- O2 + H2 m 
H" + HOi - HO· + HO· [It] 
H· + HO; 
- 0 + H2 O [I] 
H" + HO' - H2O [m] 
o + HO' - H" + O2 [n] 
Scbeme G.9 
The radica1 reactions proposed by Gonzale:z and Martire that occur during the UV photolysis of alkaline 
solutions of hydrogen peroxide are listed in Scheme G.IO (Gonzalez and Martire. 1997). 
0' - + O2 - 0 ;  
HO· • 0·- + H" 
0 ;- - 0' - + O2 
0'- + 0 ;- - Oi- + Oi  
H20 2 • HO; + H'" 
HO' + HOi - Oi- + H20 
0· - + HO; - 0;- + HO 
HO' + HO· - HO; + H" 
0 · - + HO' - HO; 
0- - + 0' - + H20 - HOi + HO-
 [a) 
[b] 
[cl 
[d] 
[e] 
[0 
[g] 
[hI 
[;] 
m 
Append.iI G FR££ IlU)lCAL R£A.cnoss 
HO· + 0;- - 0, + HO 
HO· + 0;- - H+ + 20i 
HO· + 0;- - HO- + 0 , 
0·- + 0;- + H,O - 0, + 2HO-
 0;- + 0;- + H20 - O2 + HO- + HOi 
0;- + H20 2 - HO- + HO· + O2 
0;- + HOi - 0·- + HO- + O2 
0 3- + Oi- + H20 - 2HO- + 20 2 
0 , + HO- - 0, + HO, 
0 ) + 0· - - O2 + Oi-
 0 ) + Oi- - O2 + 0 3-
 5< ....... G.IO 
G-5 
[kJ 
IrnJ 
InJ 
10J 
(PJ 
Iq] 
[rJ 
[sJ 
ItJ 
[uJ 
[v] 
The radical reactions proposed by Podolske and Johnston that occur during the photolysis of ozone are listed 
in Scheme G. 11 (Podolske and Johnston, 1983). 
0 + 0 )- 0 2 +02 
o + O2 + H 20 - 0 ] + H20 
o + O2 + He - 0 ] + He 
o + O2 + O2 - 0 ] + O2 
o + O2 + 0 3 - 0 3 + 0 3 
° + HO· - H" + 0 , 
° + HO; - HO· + 0 , 
o + H 20 2 - HO· + HOi 
HO· + 0 ] - HO; + O2 
HO· + HO· - H,O + ° 
HO· + HO· - H,O, 
HO· + H20 2 - HO; + H20 
HOi + 0 3 - HO· + O2 + O2 
HOi + HOi - O2 + H20 2 
H" + 0 3 - HO· + O2 
H· + O2 + He - HOi + He 
H· + O2 + H20 - HOi + H20 
H· + O2 + O2 - HOi + O2 
HO· + HOi - O2 + H20 
Scheme G.ll 
[a] 
fb] 
[cl 
Id] 
leJ 
If) 
IgJ 
[hi 
li] 
~] 
[k] 
[IJ 
ImJ 
[n] 
10J 
(PJ 
Iq] 
IrJ 
is]