On conversion of luminescence into absorption and the van Roosbroeck-Shockley relation

TL;DR

This paper critically examines the conversion of luminescence spectra into absorption cross sections, highlighting the inaccuracies of the van Roosbroeck-Shockley relation for semiconductors and proposing a more physically consistent method.

Contribution

It introduces a new, physically consistent method for converting luminescence spectra into absorption, addressing limitations of the traditional van Roosbroeck-Shockley relation.

Findings

Traditional van Roosbroeck-Shockley relation is incorrect for semiconductors.

Proposes a practical prescription for spectrum conversion.

Discusses the relation to nonequilibrium generalizations.

Abstract

The problem of conversion of experimentally measured luminescence spectrum into the absorption cross section is revisited. The common practice of using the van Roosbroeck-Shockley (or Kubo-Martin-Schwinger or Kennard-Stepanov) relation in this context is incorrect because luminescence from semiconductors is essentially all due to the spontaneous emission component of the recombination of carriers distributed far-from-equilibrium. A simple, physically consistent, and practical prescription for converting the luminescence spectra into absorption is presented and its relation to the so-called nonequilibrium generalization of the van Roosbroeck-Shockley relationship is discussed.