Key issues for mid–infrared emission

Abstract

Semiconductor emitters in the mid–infrared (mid–IR) fall broadly into two categories; interband and intersubband devices. The former are based on transitions from conduction to valence bands, whereas the latter use one–dimensionally quantum confined states within the conduction band. The problems which limit light–emitting diode (LED) and laser operation by keeping efficiencies low and laser thresholds high are quite different in these two cases, but both stem from high competing transition rates due to non–radiative processes. In interband devices Auger scattering dominates, and in intersubband devices, though Auger scattering is still fast, phonon emission is even faster. Working intersubband devices are only made possible by the fact that the phonon scattering is strongly resonant in energy, so three–level laser schemes may be made with a dynamic equilibrium that is inverted. We shall review our own work on suppression of these non–radiative processes by band–structure engineering, and we present new results on Auger suppression in 11 μm bandgap InSbN alloys and predictions for phonon suppression in narrow–gap intersubband devices.