The measurement of electric quadrupole moments of gas molecules by induced birefringence.
Determining the electric quadrupole moments of gas molecules from measurements of birefringence induced in the gas by an applied electric field gradient is widely recognised as being the most direct experimental technique for this purpose. This thesis presents a new molecular theory of the experiment, and also describes the apparatus used in the measurement of the quadrupole moments of a number of gases and the revisions made to the manner in which the experiment was performed. A recent eigenvalue theory of light propagation in matter has been used to derive an expression relating the induced birefringence to the electric quadrupole moment of the molecule, whether this be dipolar or not. This expression is different from that which has previously been applied to dipolar molecules. It is shown that the new expression is independent of molecular origin only if it is cast in terms of the primitive quadrupole moment, as opposed to the traceless moment. Previous work using the same set of apparatus yielded results for carbon dioxide which were significantly lower than those reported by other workers. By using the Jones calculus to re-examine the cascade of optical components used in the experiment, it was found that the previous method of performing the experiment did not completely eliminate the effects of an imperfect retardance in the quarter-wave plate used, nor of strain-induced birefringence in the windows of the the cell containing the gas. These effects could cause results to be underestimated by as much as 8%. It is shown that a better method of performing the experiment is to amplify the optical signal by deliberately offsetting the analysing prism, rather than the quarter-wave plate as was previously done. The results of measurements made using this technique are reported for carbon dioxide, carbon monoxide, nitrogen, ethene, chlorine, and boron trifluoride. The measured quadrupole moment of carbon dioxide is in good agreement with the most recent values found by other workers using a similar experiment, and with the latest theoretical value of this quantity.