Excess electron mobility maxima in argon in the critical region

Abstract

Jahnke, Meyer, and Rice have reported anomalous maxima in the low-field mobility of excess electrons in fluid Ar near the liquid-gas critical temperature. Lekner has proposed a zero-scattering-length model to explain the maxima. We propose an alternative explanation which depends on structural changes induced in a compressible fluid during a mobility experiment rather than some dramatic change in the electron-atom scattering length with density. The basic idea is that inhomogeneities of density higher than that of the normal fluid may be established and sustained in the compressible fluid through the attractive electron-Ar interaction provided by a steady-state flux of excess electrons. Once established, these dense regions may provide contiguous channels of relatively high conductivity since condensed phase diffraction effects in dense, incompressible Ar more effectively reduce the excess electron scattering. Furthermore, excess electrons are preferably channeled into the denser regions since they provide a more attractive volume-averaged effective electron-Ar interaction (i.e., optical potential). Numerical calculations of the mobility are performed employing a simple two-phase model of the electronically perturbed fluid. The lower-density phase is regarded as the normal fluid with the usual equilibrium density at a given temperature and pressure. The volume fraction of the denser phase is assumed to grow as ${\xi }^3$, where $\xi$ is the Ornstein-Zernike long-range correlation length. Estimates of the density calculated from linear-response theory indicate the denser regions are solid-like except at the highest temperatures where the mobility was measured. Electron propagation is favored in the denser regions by an attractive optical potential which is assumed to be proportional to the density difference of the two phases. The mobilities calculated from this model are in reasonably good agreement with experiment in the critical region; the magnitudes and location of the mobility maxima are well predicted. The volume fraction of the denser domains required for the model to yield reasonable experimental agreement does not exceed ${10}^{-7\mathrm{}}$.