Milling data from sugar factories in Mauritius were examined from 1960 to 2004 to assess the trend in the quality of cane received at mills and the change in factory performance. A deterioration in overall quality was apparent due to the increased level of extraneous matter delivered in the cane supply. Comparison was made with available data from other countries in the world, notably those of South Africa and Australia. Controlled addition of extraneous matter to clean cane was effected under laboratory conditions to determine the relative impact of dry leaves, green leaves and cane tops on the quality of cane and the resulting juice, and to predict through derived equations, their impact on cane processing. The addition of dry leaves was found to have the most adverse effect followed by green leaves and cane tops. In the case of dry leaf addition to cane the detrimental effects were found to be masked by an increase in the concentration of solutes in the juice extracted. This phenomenon was thought to be due to the selective sorption of water (so-called Brix-free water) by dry leaves. To test this assertion, the sugar cane stalks of four different cane varieties aged 52, 44 and 36 weeks were separated into their component parts by means of a method devised in this work. There were nine component parts: stalk fibre, stalk pith, rind fibre, rind fines, top fibre, dry leaf fibre, dry leaf fines, green leaf fibre and green leaf fines which, on characterisation by Fourier transform infrared spectroscopy and scanning electron microscopy, were very similar except that stalk pith was more flaky and had a higher surface area than the others. Various analytical techniques were tested for the determination of Brix-free water. The most convenient method proved to be a refractometric method which was improved so as to be applicable to the wide range of cane components fibres studied. Statistical analysis of the Brix-free water content of the separated samples showed that when the combined effect of fibre and pith in the cane stalk of three ages was considered, the four cane varieties were not different. This was not the case for dry leaf, green leaf, top and rind. Of the nine cane components, stalk pith exhibited the highest Brix-free water value of about 20 g/100 g fibre, whereas all the other components exhibited values of about 15 g/100 g fibre, which are much lower than the traditionally accepted value of 25% for cane. The latter was found to be the fibre saturation point of bound water determined at 20 oC, which is the sum of dissolved and hydrated waters, and which is normally greater than the Brix-free water value as determined in this work. The water sorption characteristics of the various cane component parts were further investigated by making measurements to determine the equilibrium moisture contents at various water activity values. These data were used to construct adsorption isotherms. These were fitted to 17 existing isotherm models, of which two, namely, the Hailwood-Horrobin and Guggenheim-Anderson-de Boer models, gave the best fit. The sorbed water was subsequently characterised in terms of various parameters, namely, the monolayer moisture content, the number of adsorbed monolayers, the percentage of bound water, the total surface area for hydrophilic binding, the heats of sorption of the monolayer and multilayers, the net and total isosteric heats of sorption and the entropy of sorption. From the monolayer moisture content and the amount of “hydrated water” as calculated from the Hailwood-Horrobin model, it is clear that at EMC values between 0 and 5% (aw = 0 – 0.3), the non-freezable water is tightly bound to the surface of the fibre. The second region starts at EMC values from 5% to 10 – 15% (aw = 0.3 to 0.6 – 0.8) depending on the cane components, and the bound water in this region is termed the freezable water. The third type of water is essentially free water, it exists after the second region and ends at EMC values of about 25%. From this study, it is apparent that the Brix-free water as measured in this work measures the amount of water bound in the first two regions.

Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2008.