A Nernst effect beyond the Sondheimer formula

TL;DR

This paper explores how electronic correlations might significantly alter the classical Sondheimer formula for the Nernst effect, proposing a theoretical scenario involving Majorana components and different relaxation times.

Contribution

It introduces a theoretical framework where correlations modify the Nernst effect beyond the classical Sondheimer formula using a Boltzmann approach with Majorana components.

Findings

Large Nernst signals can arise from correlation effects.

Classical Sondheimer formula may underestimate Nernst effect in correlated systems.

Proposes a phenomenological model with different relaxation times for Majorana components.

Abstract

The Nernst effect is known to be a sensitive probe of the scattering mechanisms. Large Nernst effects, a decade ago a topic mainly in vortex dynamics in superconductors, have recently attracted interest in correlated systems like cuprate metals and heavy fermion systems. However, it has been found that these ``giant" Nernst signals stay within one order of magnitude of the classical Sondheimer formula for a Fermi liquid. It therefore seems that a large Nernst effect neither relies on nor is sensitive to strong correlations. This might, however, be different in some recently studied supposedly correlated systems like FeSB_2. We do not address these systems specifically, but take them only as one inspiration for considering a theoretical scenario where correlations could actually sizably modify the Sondheimer formula. We point out that an enhanced Nernst signal would follow from the phenomenological Boltzmann approach introduced by Coleman, Schofield and Tsvelik proposed to capture electronic correlations by imposing different relaxation times on the charge conjugation even and odd Majorana components of the electron.