On the Monte Carlo Description of Hot Carrier Effects and Device Characteristics of III‐N LEDs

Abstract

of LED operation remain poorly understood, presenting an obvious bottleneck for further LED development and optimization. [5] For example, recent measurements have shown that III-N LEDs exhibit notable hot carrier distributions already close to the peak efficiency, [6][7][8][9] suggesting that hot carriers might have a more profound effect on the device characteristics than previously anticipated.To date, conventional LED simulations have mainly relied on the drift-diffusion (DD) model (see, e.g., refs. [10-13]), which cannot genuinely account for hot carriers and generally also produces larger turnon voltages for MQW LEDs than observed experimentally. [14] Many reasons have been suggested to explain this discrepancy, including charge transport through indium fluctuations, [15] V-shaped pits, [16] tunneling through traps [17] or hot carrier transport. [18] At present it is, however, unclear if the discrepancy is of a physical origin or if it arises simply because the very foundations of the widely adopted DD model make it incapable of predicting the most essential device-level characteristics of LEDs.The Monte Carlo (MC) method has been established as the standard tool to simulate hot-carrier effects in unipolar semiconductor devices such as transistors. Recently the MC method has also been expanded by Bertazzi et al. to simulate hot electron effects in GaN barriers of LEDs using full-band MC methods, [19] and ourselves, as we have recently developed a coupled Monte Carlo-drift-diffusion (MCDD) simulation method to investigate hot-electron effects in full III-N MQW LED structures. [20][21][22] More recently, we have also introduced a full MC model, [23] which can be used to carry out full MC simulations of both electrons and holes in a complete LED device. Nevertheless, the previous works have not yet provided a full analysis of the pertinent physics and differences between MC and DD. To provide more insight into the physics of MQW devices and to facilitate their further development, we now deploy our full MC simulator to answer the following questions:

Influence of photo-generated carriers on current spreading in double diode structures for electroluminescent cooling

“…[21] and an optical coupling factor of 90%. Further details about the full 3D simulation model can be found elsewhere [14,16,22].…”

Elimination of Lateral Resistance and Current Crowding in Large‐Area LEDs by Composition Grading and Diffusion‐Driven Charge Transport

“…While considerable uncertainty generally surrounds the quantitative description of optical and electrical properties of III-N QWs (see, e.g., refs. [18][19][20]), such uncertainties do not significantly affect the main conclusions of this paper due to the comparative nature of the simulations and since our results mainly focus on the charge transport in the bulk layers. For the surface recombination at the edges of the DHJ active region we employ the simple surface recombination model based on ref.…”

Diffusion-driven GaInP/GaAs light-emitting diodes enhanced by modulation doping

“…Also, J p ( J n ) is the hole (electron) current density, p ( n ) is the hole (electron) mobility, and R is the net recombination rate per unit volume. More detailed description of the model can be found in Kivisaari et al (2015), Kivisaari et al (2017b), Sadi et al (2014). The recombination rate R, consisting of Shockley-Read-Hall (SRH) R SRH , radiative R rad , and Auger recombination R Aug , is modeled using the wellknown parameterized formula given by Kivisaari et al (2015) where n i is the intrinsic carrier density, and A, B and C are recombination constants for SRH, radiative and Auger processes, respectively.…”