An ab-initio study of the electron-phonon coupling within a Cr(001)-surface

TL;DR

This study uses ab-initio density functional perturbation theory to analyze electron-phonon coupling at the Cr(001) surface, finding it insufficient to explain the observed resonance, which is likely due to electron-electron correlations.

Contribution

It provides the first ab-initio calculations of electron-phonon coupling on the Cr(001) surface, challenging previous interpretations of the surface resonance.

Findings

Calculated electron-phonon mass-enhancement factors are 0.19 and 0.16.

Experimental data shows a mass-enhancement factor of about 0.70.

Resonance is attributed to electron-electron correlations, not electron-phonon effects.

Abstract

It is experimentally well established that the Cr(001)-surface exhibits a sharp resonance around the Fermi level. However, there is no consensus about its physical origin. It is proposed to be either due to a single particle dz2 surface state renormalised by electron-phonon coupling or the orbital Kondo effect involving the degenerate dxz/dyz states. In this work we examine the electron-phonon coupling of the Cr(001)-surface by means of ab-initio calculations in the form of density functional perturbation theory. More precisely, the electron-phonon mass-enhancement factor of the surface layer is investigated for the 3d states. For the majority and minority spin dz2 surface states we find values of 0.19 and 0.16. We show that these calculated electron-phonon mass-enhancement factors are not in agreement with the experimental data even if we use realistic values for the temperature range and surface Debye frequency for the fit of the experimental data. More precisely, then experimentally an electron-phonon mass-enhancement factor of 0.70~0.10 is obtained, which is not in agreement with our calculated values of 0.19 and 0.16. Therefore, we conclude that the experimentally observed resonance at the Cr(001)-surface is not due to polaronic effects, but due to electron-electron correlation effects.