Calculated electrical and thermal resistivities of Nb and Pd

摘要: The electrical and thermal resistivities ($\ensuremath{\rho}$ $W$) of pure Nb Pd are calculated from nearly first principles. Realistic Korringa-Kohn-Rostoker energy bands wave functions, experimental phonon frequencies Born-von K\'arm\'an eigenvectors, rigid muffin-tin electron-phonon potentials used to generate the velocities scattering probabilities in Bloch-Boltzmann equation, at a mesh 48 000 points on Fermi surface. Solutions for $\ensuremath{\rho}$ $W$ exhibited three levels accuracy: (1) lowest-order variational approximation (LOVA) where surface displaces rigidly; (2) $N$-sheet different sheets displace independently; (3) fully inelastic calculation is distribution function allowed arbitrary variations with (normal surface) reflect inelasticity scattering. Above $T=100$ K, corrections LOVA order 1%, but below both give large results. These results also compared Bloch-Gr\"uneisen formulas fitted $T\ensuremath{\sim}{\ensuremath{\Theta}}_{D}$. In range $100 \mathrm{K}\ensuremath{\lesssim}T\ensuremath{\lesssim}300 \mathrm{K}$, calculations exceed by \ensuremath{\sim} 10%. Good agreement persists into $10 \mathrm{K}\ensuremath{\lesssim}T\ensuremath{\lesssim}100 except that theory underestimates experiment significantly lower-temperature end, suggesting possible error models small $\stackrel{\ensuremath{\rightarrow}}{\mathrm{Q}}$ interpretation complicated Coulomb effects. Below $T=10$ finite size prevents reliable calculations. Simple such as inadequate account data. Mott's (1936) "$s\ensuremath{-}d$" picture shown be qualitatively correct Pd. Extension this was suggested subsequently various authors, present does not support this.