AN ALTERNATIVE MODEL FOR TRANSPORT OF MOLECULES BETWEEN GAS AND SOLUTION

For almost 70 years, transport of molecules like N2 between gas and solution has been assumed to take place by diffusion through a surface film. We believe that gases and liquids have such different viscosities that such a film would be unstable to even modest stirring of the liquid. We propose instead that transport of a molecule between gas and surface takes place in a single translational step and that transport between surface and bulk solution is accomplished by physical mixing without significant contribution from molecular diffusion. The two models applied to data collected 50 years apart lead to identical limiting rates of transport at high stirring rates. Our model also predicts that limiting rate of transport is independent of apparatus, and it permits direct calculation of the accommodation coefficient, gamma, or probability that a molecule striking the surface will become dissolved in the bulk liquid. Because models of transport employed previously assumed diffusion through a surface film of unknown thickness, rates of transport could not be used to measure accommodation coefficients. If our model is valid, we find that gamma is in the neighborhood of 10(-9) for the hydrogen-bonded solvents water, aqueous glycerol, and concentrated sulfuric acid. It rises to 10(-7) for the nonpolar solvent heptane. These values of gamma are much less than those around 10(-1) which have been reported for solutions like HCl or SO2 in water where gas molecules ionize almost instantaneously upon dissolving. If our proposed model is indeed valid, measurement of rates of transport of gases to sufficiently well mixed solutions should provide data of direct applicability to many problems of interphase transport. We believe that for decades chemical engineers have been depriving themselves of techniques which could have been useful for their work.