Ground state properties of 3d metals from self-consistent GW approach

TL;DR

This study applies the self-consistent GW method to 3d transition metals to evaluate ground state properties, finding it more accurate than LDA but less so than GGA, with systematic deviations indicating potential for further theoretical improvements.

Contribution

The paper demonstrates the application of self-consistent GW to 3d metals and compares its accuracy with LDA, GGA, and RPA, highlighting its systematic errors and potential for refinement.

Findings

scGW underestimates Wigner-Seitz radius by about 1%

scGW overestimates bulk modulus by about 25%

scGW is more accurate than LDA but less accurate than GGA

Abstract

Self consistent GW approach (scGW) has been applied to calculate the ground state properties (equilibrium Wigner-Seitz radius $S_{WZ}$ and bulk modulus $B$) of 3d transition metals Sc, Ti, V, Fe, Co, Ni, and Cu. The approach systematically underestimates $S_{WZ}$ with average relative deviation from the experimental data about 1% and it overestimates the calculated bulk modulus with relative error about 25%. It is shown that scGW is superior in accuracy as compared to the local density approximation (LDA) but it is less accurate than the generalized gradient approach (GGA) for the materials studied. If compared to the random phase approximation (RPA), scGW is slightly less accurate, but its error for the 3d metals looks more systematic. The systematic nature of the deviation from the experimental data suggests that the next order of the perturbation theory should allow one to reduce the error.