From Atoms to Devices: Performance Bottlenecks of Nitride Semiconductors Using Multi-Scale Quantum Methods

Abstract

III-nitride semiconductors have revolutionized modern electronics by enabling high-power radio-frequency and lighting technologies. These materials hold immense potential for new technologies such as miniaturized displays, ultraviolet sterilization, and fast electric-vehicle charging. However, there are still performance bottlenecks that need to be addressed. In this thesis, I investigate the microscopic mechanisms that limit the performance of nitride semiconductors in power-conversion and lighting applications, and propose new solutions using quantum-mechanical methods that connect the microscale physics to macroscale device phenomena.

First, I examine the limitations of III-nitride semiconductors in power-conversion applications. To increase the breakdown voltage of GaN, which is critical for higher power devices, it is promising to alloy it with Al. However, this approach decreases the electron mobility due to alloy scattering. In this thesis, I develop a novel approach to calculate the low-field mobility of semiconductor alloys from first principles. I find that the mobility of AlGaN decreases by a factor of $sim$7 compared to GaN. Consequently, Al compositions above 75% are required to achieve even a two-fold increase in the Baliga Figure of Merit (BFOM) compared to GaN, at which point impurity doping becomes increasingly difficult. To address this issue, I propose using atomically thin superlattices of AlN and GaN that are free of disorder. My calculations indicate that these nanostructures exhibit a 4.8 eV band gap and a mobility $3-4times$ higher than random AlGaN. By accounting for the dopant ionization fraction in the BFOM, I show that the superlattices exhibit the highest modified BFOM among prominent competing semiconductors.

Second, I investigate the mechanism that causes InGaN light-emitting diodes (LEDs) to suffer from an emission blueshift and linewidth broadening when operated at high currents, leading to a degradation of their color properties. By systematically considering the effects of polarization-field screening, phase-space filling, and many-body plasma renormalization, I comprehensively explain the current-dependent spectral characteristics of polar III-nitride quantum wells. My analysis overturns the prevailing hypothesis that the emission blueshift is primarily due to the screening of internal polarization fields, as this explanation neglects the contribution of plasma renormalization, which is nearly equal but opposite in magnitude. In contrast, the blueshift is explained only by accounting for a complex interplay of polarization-field screening, plasma renormalization, and phase-space filling. On the other hand, the spectral broadening occurs mainly due to phase-space filling. My analysis suggests that the key to improving the spectral characteristics of InGaN LEDs is to accelerate carrier recombination and transport and reduce the carrier density required to operate them at high power density.

Finally, I investigate the concept of defect tolerance in InGaN emitters. Recent experiments have challenged the widely accepted hypothesis that carrier localization suppresses diffusion and enhances the tolerance of InGaN emitters to defects. By examining the competition between radiative and Shockley-Read-Hall recombination in InGaN alloys using a formalism that I recently developed, I propose that carrier localization and polarization fields enhance the quantum efficiency at low current densities, without invoking the suppression of carrier diffusion. Decreasing the oscillator strength or increasing the quantum-well thickness may improve the quantum efficiency of LEDs for low-power applications but it exacerbates efficiency droop and impair color purity control at high operating powers.

Overall, this thesis paves the way for the continued development of III-nitride technology, and its approach can be generalized to other emerging semiconductors.