Nonequillibrium Green’s function analysis of interwell transport and scattering in monopolar lasers

Abstract

Listen

Translate article

Transport and scattering is examined in a simple coupled double-well model of infrared monopolar lasers via a nonequilibrium Green's function (NEGF) based analysis. Roughly speaking, in such lasers a more or less three level system is formed where electrons are injected into the first excited state subband of a leading well, decay via photon emission to intermediate ideally resonant-state subbands resulting from coupling of the ground state subband of the leading well to the first excited state subband of the trailing well, and subsequently, preferably quickly to allow population inversion between the initial and intermediate state, decay via phonon emission to the ground state subband of the trailing well. Golden Rule based analysis is widely used to model scattering including in this system. Implicit in its use is a random-phase approximation among the final states. However when scattering processes appear to produce not only changes in energy states but also real-space transport as between the wells here, this approximation can become suspect. In this work the affects of this approximation on scattering-induced population and depopulation of intermediate level(s) are addressed and overcome using a NEGF technique that allows consideration of transport and scattering absent the Golden Rule and associated random phase approxmiations. It is found that as the barrier becomes thick, the Golden Rule approximation can overestimate the depopulation rate of the intermediate levels. Through changes in the homogeneous broadening of the photon transition associated with changes in the depopulation rates of the intermediate level(s), the variations in barrier thickness could also have additional effects on population inversion and gain not apparent through Golden Rule calculations. Accordingly, barrier thickness is found to also be a potentially critical parameter for optimizing device performance.