Weak antilocalization in two-dimensional systems with large Rashba splitting

TL;DR

This paper develops a comprehensive theory for quantum transport and magnetoconductivity in two-dimensional electron systems with large Rashba or Dresselhaus spin-orbit splitting, extending beyond the diffusion approximation.

Contribution

It provides an analytical expression for quantum conductivity correction considering arbitrary scattering events and large spin-orbit splitting, valid across diffusive and ballistic regimes.

Findings

Zero-field conductivity correction combines universal diffusive and spectrum-dependent ballistic terms.

Magnetoconductivity is negative across weak magnetic fields in both diffusive and ballistic regimes.

Magnetoconductivity behavior depends on the Fermi energy relative to the Dirac point.

Abstract

We develop the theory of quantum transport and magnetoconductivity for two-dimensional electrons with an arbitrary large (even exceeding the Fermi energy), linear-in-momentum Rashba or Dresselhaus spin-orbit splitting. For short-range disorder potential, we derive the analytical expression for the quantum conductivity correction, which accounts for interference processes with an arbitrary number of scattering events and is valid beyond the diffusion approximation. We demonstrate that the zero-field conductivity correction is given by the sum of the universal logarithmic "diffusive" term and a "ballistic" term. The latter is temperature independent and encodes information about spectrum properties. This information can be extracted experimentally by measuring the conductivity correction at different temperatures and electron concentrations. We calculate the quantum correction in the whole range of classically weak magnetic fields and find that the magnetoconductivity is negative both in the diffusive and in the ballistic regimes, for an arbitrary relation between the Fermi energy and the spin-orbit splitting. We also demonstrate that the magnetoconductivity changes with the Fermi energy when the Fermi level is above the "Dirac point" and does not depend on the Fermi energy when it goes below this point.