Nonradiative recombination in 1.56 μm GaInNAsSb/GaNAs quantum-well lasers

Abstract

We have shown experimentally that in GaInNAsSb/GaAs quantum-well lasers there are significant nonradiative contributions to threshold current from the barriers and the well. By matching a simulation to the experiment we find that Auger recombination in the barriers is very weak, due to the low carrier density. Shockley–Read–Hall recombination is the dominant source of nonradiative current, with the barriers making the major contribution, possibly due to their higher defect density than the wells. This suggests that significant improvements could be made by optimizing growth conditions and layer design, with particular attention to the barrier.

Thermal dependence of the optical gain and threshold current density of GaInNAs/GaAs/AlGaAs quantum well lasers

“…In this laser, J mono contributes about 60%, J rad contributes about 10%, J Auger contributes for about 24%, and J leak about 6% of J th at T¼300 K. The contribution of J leak to J th is not significant, while J mono gives the major contribution to J th , which indicates that a nonradiative mechanism induced by N in GaInNAs material dominates the recombination process. [31][32][33][34] This implies that we can reduce the nonradiative monomolecular recombination in the GaInNAs QW by improving the crystalline quality of GaInNAs, and realize 1.3 lm high-performance GaInNAs QW lasers. [35][36][37][38][39] The Auger recombination is not the dominant recombination process while dominates the temperature sensitivity of the threshold current density.…”

Optical Gain in GaInNAs and GaInNAsSb Quantum Wells

“…Additional gain versus radiative current density measurements for the GaInNAsSb material were taken on a device with three quantum wells to provide results at a lower injection per well. Our previous work [12] showed that the wells in these three and one well samples were similar and that both the peak gain (1) with parameters given in Table I. and radiative current density scale with well number so the results in the lower plot of Fig. 1 are given as modal gain per well and radiative current density per well.…”

“…Where this non-radiative contribution is due to Shockley-Read-Hall (SRH) recombination it can in principle be eliminated, in which case the radiative current density indicates the intrinsic current density of the structure. In the case of GaInNAsSb wells our previous work [12] suggests that the intrinsic Auger recombination current density in the well may be significantly larger than the radiative current density, and simulations also show a substantial nonradiative current due to carriers thermally excited to the adjacent barrier material. We have estimated that if SRH recombination could be eliminated a threshold current density of 320 A cm −2 (Radiative + Auger) could be achieved in a similar structure for a 1 mm long device [12].…”

“…In the case of GaInNAsSb wells our previous work [12] suggests that the intrinsic Auger recombination current density in the well may be significantly larger than the radiative current density, and simulations also show a substantial nonradiative current due to carriers thermally excited to the adjacent barrier material. We have estimated that if SRH recombination could be eliminated a threshold current density of 320 A cm −2 (Radiative + Auger) could be achieved in a similar structure for a 1 mm long device [12]. The experimental gain data of Thranhardt et al [15] also shows that a higher drive current is required to produce similar peak gain from GaInN 0.0043 As 0.9957 and GaInAs wells (by a factor of about 2).…”