Efficiency and Forward Voltage of Blue and Green Lateral LEDs with V-shaped Defects and Random Alloy Fluctuation in Quantum Wells (original) (raw)

2022, Physical Review Applied

For nitride-based blue and green light-emitting diodes (LEDs), the forward voltage V for is larger than expected, especially for green LEDs. This is mainly due to the large barriers to vertical carrier transport caused by the total polarization discontinuity at multiple quantum well and quantum barrier interfaces. The natural random alloy fluctuation in QWs has proven to be an important factor reducing V for. However, this does not suffice in the case of green LEDs because of their larger polarization-induced barrier. V-defects have been proposed as another key factor in reducing V for to allow laterally injection into multiple quantum wells (MQWs), thus bypassing the multiple energy barriers incurred by vertical transport. In this paper, to model carrier transport in the whole LED, we consider both random-alloy and V-defect effects. A fully two-dimensional drift-diffusion charge-control solver is used to model both effects. The results indicate that the turn-on voltages for blue and green LEDs are both affected by random alloy fluctuations and V-defect density. For green LEDs, V for decreases more due to V-defects, where the smaller polarization barrier at the V-defect sidewall is the major path for lateral carrier injection. Finally, we discuss how V-defect density and size affects the results. Given that V-defects play an important role in carrier injection, it is important to understand the role of random alloy fluctuations and of V-defects in carrier injection. The V-defect and random alloy effects may be simulated separately in three-dimensional simulations. 8,15 On the one hand, V-defects range from a few hundred nanometers in size to the µm scale, allowing for large mesh sizes. On the other hand, random alloy fluctuations occur on a scale of a few nanometers, requiring mesh elements to be as small as possible to account for such small-scale fluctuations. Including both effects in a three-dimensional simulation requires unreasonable computer memory (> a few TB) and computing times. However, to understand the influence of random alloy fluctuation