Skip to main navigation Skip to search Skip to main content

Intervalley scattering in GaAs: Ab initio calculation of the effective parameters for Monte Carlo simulations

  • Laboratoire des Solides Irradiés
  • Tomsk State Pedagogical University - TSPU

Research output: Contribution to journalArticlepeer-review

Abstract

Interactions between excited electrons and short-wavelength (intervalley) phonons in GaAs are studied using density functional theory for the conduction bands, and density functional perturbation theory for phonon frequencies and matrix elements of the electron-phonon interaction. We have calculated the deformation potentials (DPs) and the average intervalley scattering time 〈τ〉. The integration of the scattering probabilities over all possible final states in the Brillouin zone has been performed without any ad hoc assumption about the behavior of the electron-phonon matrix elements nor the topology of the conduction band. For transitions from the L point to Γ valley (within the first conduction band), we find 〈τ〉L to be 1.5 ps at 300 K, in good agreement with time-resolved photoluminescence experiment. We discuss the difference between our calculated DPs, and effective parameters used in Monte Carlo simulations of optical and transport properties of semiconductors. The latter are based on Conwell's model, in which electron-phonon interaction is described by one single constant and a parabolic model is used for conduction bands. We deduce the effective DP from our 〈τ〉, and compare it to our calculated DPs. We conclude that only effective DPs obtained from a full calculation of 〈τ〉 are relevant parameters for Monte Carlo simulations.

Original languageEnglish
Pages (from-to)301-307
Number of pages7
JournalApplied Physics A: Materials Science and Processing
Volume86
Issue number3
DOIs
Publication statusPublished - 1 Mar 2007

Fingerprint

Dive into the research topics of 'Intervalley scattering in GaAs: Ab initio calculation of the effective parameters for Monte Carlo simulations'. Together they form a unique fingerprint.

Cite this