A critique of Boltzmann solution methods for calculating the longitudinal diffusion coefficient of electrons in gases

It is shown that, for the quasi-Lorentz gas model, there is no basis for doubting the correctness of the longitudinal diffusion coefficient D L , as obtained from the density gradient expanded solution of the Boltzmann equation. This conclusion is observed to hold true even for electrons in molecular gases where large inelastic processes are present. A corrected version of Lucas' (1970) derivation is shown to yield the same D L as that obtained from the density gradient expansion theories. Monte Carlo results for electrons in N 2 and model CH 4 are also presented and compared with existing simulation and Boltzmann solution results. The conclusions indicate that the calculations of D L in the Boltzmann code of Pitchford and Phelps (1982) are numerically in error.