Improved model for the thermal conductivity of binary metallic systems

TL;DR

This paper refines Mott's two-band model for binary metal alloys, incorporating first-principles calculations and high-throughput measurements to accurately predict their thermal and electrical conductivities across compositions.

Contribution

It introduces a corrected and extended Mott model that accounts for element-specific orbital contributions, validated by experiments and first-principles calculations.

Findings

Excellent agreement between model and experimental data

Orbital contributions vary with alloy composition

Model can be integrated into CALPHAD framework

Abstract

We extended and corrected Mott's two-band model for the composition-dependence of thermal and electrical conductivity in binary metal alloys based on high-throughput time-domain thermoreflectance (TDTR) measurements on diffusion multiples and scatterer-density calculations from first principles. Examining PdAg, PtRh, AuAg, AuCu, PdCu, PdPt, and NiRh binary alloys, we found that the nature of the two dominant scatterer-bands considered in the Mott model changes with the alloys, and should be interpreted as a combination of the dominant element-specific s- and/or d-orbitals. Using calculated orbital and element-resolved density-of-states values calculated with density functional theory as input, we determined the correct orbital mix that dominates electron scattering for all examined alloys and find excellent agreement between fitted models and experiments. The proposed description of the composition dependence of the resistivity can be readily implemented into the CALPHAD framework.