The effect of solid micro particles on mass transfer in agitated dispersions.

Abstract

The industrial application of gas-liquid contactors has made effective design and optimisation of these processes a very important topic. In order to sustain a competitive advantage, rate limiting steps must be clearly understood. Hydrodynamics, heat transfer and mass transfer are complicated features of gas-liquid contactors and require a fundamental understanding. The mechanism of mass transfer in the presence of a small concentration of solid micro particles has been the subject of debate. The adsorption of gas by solid particles ("shuttle mechanism") is the traditional explanation. Recent experimental evidence suggests that the introduction of micro particles removes trace surface active impurities from the system and allows the true mass transfer coefficient to be measured. The objective of this study was to confirm the surfactant removal theory. Mass transfer is a field characterised by imprecise empirical relationships and difficult to obtain experimental parameters. This puts into context the significant challenge posed in preparing the careful set of measurements and analyses presented in this study to lend support to the surfactant removal mechanism. The study began with a review of mass transfer models. These models are based on concepts such as surface renewal and idealised turbulence. It is, however, difficult to choose between the models as they predict similar values despite being based on different mechanisms. The overall mass transfer coefficient is composed of the gas-phase coefficient (kGa) and liquid-phase coefficient (kLa). As the values of the coefficients are comparable and the solubility of oxygen or hydrogen is very Iow, the overall mass transfer coefficient is approximately equal to the liquid side coefficient. The relationship of kL with the diffusion coefficient (D) is one of the limited ways of choosing between the models. Mass transfer models predict k j • u:. D" . n is predicted to be % for a rigid surface (contaminated interface region) and Y2 for a mobile surface (clean interface region). If the surfactant removal mechanism applies, the introduction of solid particles will be accompanied by a reduction of n from % to 1/2. The effect of particles on n can be calculated from precise measurement of kL of gases with significantly different diffusion coefficients. A review of experimental methods was made to find precise methods to characterise mass transfer in the presence of solid micro particles. The chemical sulphite, gas-interchange and pressure step methods were identified as appropriate methods. These were implemented in a stirred cell (0.5 !) and an agitated tank (6 I). The chemical sulphite measurements were used to confirm that the enhancement of kLa is due to an enhancement of kL and not the specific interfacial area (a). Flat surface experiments were made using water and 0.8 M sodium sulphate batches. The reduction of n from % to Y2 was confirmed in both apparatuses after the addition of solid particles. The data were very well correlated and the dependence of kr on the energy dissipation rate per unit volume (e) is similar to the theoretically predicted value of 114 for the exponent. Observation of the reduction of n from % to Y2 was extended to agitated dispersions. The stirred cell kLa data were measured by the gas interchange method and are of excellent quality. The agitated tank results were measured by pressure step methods. The pressure dependence of the polarographic probes affected the precision of the results and the effect was within the experimental uncertainty. The effect of particles on n could not, therefore, be conclusively confirmed in the agitated tank. By relating precisely measured mass transfer coefficients to the diffusion coefficients; the surfactant removal theory is confirmed. The result is valid for a flat mass transfer area as well as for agitated dispersion where the nature of the interface region changes with time due to the accumulation of surfactants on an initially clean interface.