Electronic energy bands calculated from local-density-functional (LDF) theory are used to predict anisotropic transport properties of the oxide superconductors ${\mathrm{La}}_{2\mathrm{-}\mathrm{x}}$${\mathrm{Sr}}_{\mathrm{x}}$${\mathrm{CuO}}_4$ and ${\mathrm{YBa}}_2$${\mathrm{Cu}}_3$${\mathrm{O}}_7$. Calculations of the magnitude of the resistivity tensor ${\rho }_{\alpha \beta }$ (assuming only electron-phonon scattering) lead to results smaller than given by current experiments, suggesting that LDF theory overestimates the Drude plasma frequency (or Fermi velocity), or else that another scattering mechanism dominates. Determinations of the Fermi velocity from critical field measurements suggest however that it is close to the LDF value. Quantities which are independent of the scattering are calculated with more success: the resistivity anisotropy ${\rho }_{\mathrm{zz}}$/${\rho }_{\mathrm{xx}}$, the Hall tensor $R_{\alpha \beta \gamma \mathcal{’}}^H$ and the thermopower $S_{\alpha \beta }$. It is predicted that $R^H$ should appear ‘‘holelike’’ when electrons orbit in metallic planes but S should appear ‘‘electronlike’’ for thermal gradients in the plane. When fields are tipped 90°, all coefficients change sign. There is not very much single-crystal data to compare with, but what exists is qualitatively consistent, including anomalous signs of the Hall tensor. Polycrystalline data for $R^H$ seem completely at variance with the LDF results, so it is surprising how well single-crystal results agree.

• Received 30 November 1987

DOI:https://doi.org/10.1103/PhysRevB.37.7482