Multiscale simulations of carrier transport in multi-quantum well (Al,Ga)N-based deep-UV LEDs: Alloy disorder, acceptor activation, and injection efficiency

Abstract

Abstract Aluminium gallium nitride ((Al,Ga)N) has become a key material system for deep-UV light emitters, particularly in the UV-C range (<280 nm). However, device efficiencies remain far below their visible-range counterparts, motivating the need for detailed theoretical insight. Accurate modelling of (Al,Ga)N LEDs requires capturing alloy-induced fluctuations, valence band mixing, and incomplete acceptor ionisation. In this work, we employ a quantum corrected multiscale simulation framework that builds on atomistic electronic-structure theory to establish a hybrid band model that is then combined with drift-diffusion modelling to investigate carrier transport and recombination in a deep-UV multi-quantum well LED. We show that p-doping the electron-blocking layer substantially enhances the current-injection efficiency, even for high acceptor activation energies. Comparing random alloy and virtual crystal descriptions reveals that alloy fluctuations drive a strongly asymmetric hole distribution toward the n-side of the device. Electron densities remain uniform across the wells, but barrier region alloy disorder opens percolative pathways that significantly increase electron leakage from the active region, which affects the current injection efficiency and ideality factors. Overall, our work demonstrates that multiscale simulation is essential for disentangling the intertwined physical mechanisms that govern the performance of a deep-UV LED, allowing us to clearly identify the individual roles of alloy disorder, carrier transport, and recombination processes played in this context.