Boosting the power factor with resonant states: a model study

TL;DR

This study uses a theoretical model to explore how resonant states and hybridization can enhance thermoelectric power factors, providing guidelines for optimizing material properties.

Contribution

It introduces a tight-binding model to analyze the impact of hybridization on thermoelectric properties, highlighting key parameters for power factor enhancement.

Findings

Hybridization boosts the Seebeck coefficient under certain conditions.

Conductivity is suppressed near the resonance energy.

Optimal power factor occurs with specific hybridization strength and Fermi level positioning.

Abstract

A particularly promising pathway to enhance the efficiency of thermoelectric materials lies in the use of resonant states, as suggested by experimentalists and theorists alike. In this paper, we go over the mechanisms used in the literature to explain how resonant levels affect the thermoelectric properties, and we suggest that the effects of hybridization are crucial yet ill-understood. In order to get a good grasp of the physical picture and to draw guidelines for thermoelectric enhancement, we use a tight-binding model containing a conduction band hybridized with a flat band. We find that the conductivity is suppressed in a wide energy range near the resonance, but that the Seebeck coefficient can be boosted for strong enough hybridization, thus allowing for a significant increase of the power factor. The Seebeck coefficient can also display a sign change as the Fermi level crosses the resonance. Our results suggest that in order to boost the power factor, the hybridization strength must not be too low, the resonant level must not be too close to the conduction (or valence) band edge, and the Fermi level must be located around, but not inside, the resonant peak.