ALL CURRENT methods for the determination of diffusion coefficients of gases dissolved in liquids suffer from two main disadvantages. Firstly, each measurement requires a period of hours or even days to complete, and secondly, the methods are somewhat limited in their application [ 11. A critical review of these methods has been given by Himmelblau [2].
In view of the influence of diffusion on the overall rate of many chemical engineering operations involving gas-liquid systems, it is highly desirable that a method of determining diffusion coefficients which offers speed, accuracy and general application be developed. In this note, we report the first stage in the development of such a technique.
The method is based upon that used by Bohemen and Purnell[3], and Giddings and Seager[4,5] for measurements of gas phase binary diffusion coefficients. This is a flow method in which a pulse of component A is injected into a stream of component B, passing in laminar flow through a long tube of circular cross-section. Dispersion of the injected pulse is then effected by longitudinal molecular diffusion and by Taylor dispersion, the latter being a result of the parabolic velocity profile and radial molecular diffusion [6]. At a distance L, downstream from the point of injection, the temporal variation of the mean concentration of the sample component over the cross-section of the tube conforms to a Gaussian distribution function, provided that [6-91