Intersubband absorption linewidth in GaAs quantum wells due to scattering by interface roughness, phonons, alloy disorder, and impurities

TL;DR

This paper calculates the intersubband absorption linewidth in GaAs quantum wells considering various scattering mechanisms, revealing their different impacts and explaining experimental observations.

Contribution

It provides a detailed quantitative analysis of how interface roughness, phonons, alloy disorder, and impurities affect absorption linewidth and transport broadening in GaAs QWs, clarifying their distinct roles.

Findings

Interface roughness dominates linewidth over transport broadening.

LO phonons and impurities contribute less to linewidth than to transport broadening.

LA phonons and alloy disorder contribute similarly to linewidth and transport broadening.

Abstract

We calculate the intersubband absorption linewidth in quantum wells (QWs) due to scattering by interface roughness, LO phonons, LA phonons, alloy disorder, and ionized impurities, and compare it with the transport energy broadening that corresponds to the transport relaxation time related to electron mobility. Numerical calculations for GaAs QWs clarify the different contributions of each individual scattering mechanism to absorption linewidth and transport broadening. Interface roughness scattering contributes about an order of magnitude more to linewidth than to transport broadening, because the contribution from the intrasubband scattering in the first excited subband is much larger than that in the ground subband. On the other hand, LO phonon scattering (at room temperature) and ionized impurity scattering contribute much less to linewidth than to transport broadening. LA phonon scattering makes comparable contributions to linewidth and transport broadening, and so does alloy disorder scattering. The combination of these contributions with significantly different characteristics makes the absolute values of linewidth and transport broadening very different, and leads to the apparent lack of correlation between them when a parameter, such as temperature or alloy composition, is changed. Our numerical calculations can quantitatively explain the previously reported experimental results.