Answers are grounded in this paper and cite their sources.
In classical liquids differing only in their isotopic mass M, with a total potential energy function independent of mass, the van Hove function S(k, ω, M, T) is shown to have the form M 1/2(k, ω 1/2, T), where f is the same function for different isotopes. This result applies also to the self part Ss (kω) and, combined with the Kubo formula for the coefficient of self-diffusion D, it follows that D is rigorously proportional to M −1/2 in this theory. From the corresponding correlation function expressions for the shear and bulk viscosities, τ and ω respectively, it is found that these are both proportional to M 1/2, a result obtained earlier on dimensional grounds by Rowlinson. It is argued in liquid metals, from the machine calculations of Paskin and Rahman, that the frequency spectrum has a rather well defined range, essentially up to the Debye frequency ω d . First-order formulae, consistent with the isotopic mass dependence discussed above, are then obtained for diffusion and viscosity of liquid metals at the melting point. An interesting feature of the formula for viscosity is its explicit dependence on the long wavelength limit of the liquid structure factor. A more general argument is also presented, which leads to similar results for diffusion and viscosity even when the Debye edge is badly blurred, as in argon. Shear viscosity measurements for Li6 and Li7 deviate appreciably from the predictions of the classical theory. Measurements of the frequency spectra of light isotopes may help in clarifying the situation. Another possibility might be to study neutron scattering from a binary isotopic mixture, the theory of which is discussed in an appendix.
No takes yet. Share an insight, caveat, or question.
Brown et al. (1968) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: