No abstract

The mutual diffusion coefficients of H2 with Ne, Ar, and Xe have been determined by the two-bulb technique of Ney and Armistead in the temperature range —30° to 68°C. Diffusion was allowed to take place through a precision capillary tube connecting the two diffusion bulbs and samples of the gas were analyzed at different times with the help of a previously calibrated thermal-conductivity analyzer. A least-square method was then followed to calculate the force constants on the Lennard-Jones (12:6) potential model from the experimental values of the diffusion coefficients. Also, using experimental values of the mutual-diffusion coefficient, thermal conductivity, and viscosity of pure components, the thermal conductivity of various mixtures were calculated and a good agreement with the experimental data was obtained.

The temperature dependence of the isotopic thermal-diffusion factor α0 for ammonia has been investigated with the help of an equimolar mixture of 14NH3 and 15NH3. A 10-tube metal swing-separator as well as a two-bulb glass apparatus were used for these measurements. The value of α0 increases with temperature, its value being 0.00±0.02 at 242°K to 0.09±0.03 at 496°K, in disagreement with the values obtained by Watson and Woernley, who observed a definite inversion in the sign of α0 at temperature ∼290°K. The present set of observed values of α0 also disagrees with the theoretical calculation of Monchick and Mason. Possible causes of this disagreement are: (1) inadequacy of the potential model, (2) neglect of inelastic collisions, and (3) formation of dimers. The self-diffusion coefficient of ammonia has also been studied in the temperature range of 233° to 353°K with a two-bulb apparatus. There is marked disagreement between observed and calculated values at low temperatures. This is also explained in terms of dimerization.

An extensive investigation of the isotopic thermal-diffusion factor α0 of argon in the mean temperature range 127° to 653°K has been carried out with a four-tube swing separator employed previously in similar work with helium and neon. Because of a linear variation of α0 with T in the low-temperature region, the mean temperature T̄ = (T2—T1)/ln(T2/T1 ) has been used instead of the arithmetic mean of T1 and T2. This temperature assignment has been verified by varying the temperature gradient significantly. The experimentally determined value of α0 increases from 0.068±0.015 at T̄ = 127°K to 0.485±0.035 at T̄ = 653°K, in fair agreement with earlier experiments. Just as for helium and neon, a Lennard-Jones 12:6 potential model fails to account for the observed temperature variation of argon α0 values. The best fit for an exp—six force model occurs for α = 17 and ε/k = 176°K. Using these force parameters, the best fit with experimental viscosity data yields rm = 3.535 Å. These parameters are then used to calculate the diffusion coefficient and second virial coefficient for argon. There is fair agreement with the experimental values of these other transport coefficients.

The isotopic thermal-diffusion factor ao for xenon has been determined over a mean temperature range of 245 0 to 543°K. An artificial mixture of xenon isotopes containing approximately 18% 129Xe and 15% 136Xe was subjected to thermal diffusion in a 10-tube swing separator; the isotopic thermal-diffusion factor was obtained from the change of concentrations of these two isotopes only. A best-fitting procedure yields potential parameters a=16 and ./k=257°K for the exponential-six interaction-potential model, whereas the Lennard-Jones (12:6) model fails to reproduce the experimental results. This exp-6 potential was then used to calculate the values of diffusion, viscosity, thermal conductivity, and second virial coefficient, and a reasonably good agreement was obtained with rm = 4.41 A.An intercomparison of the observed and the theoretically calculated isotopic thermal-diffusion factors for all the inert gases from T*=O.6 to 60 is also presented.