Physics-based trap modeling of GaN HEMTs

Gallium nitride (GaN) high-electron-mobility transistors (HEMTs) have emerged as the preferred choice for telecommunication system designs, owing to their superior power levels compared to their silicon counterparts, especially at high frequencies and in harsh environments. Despite these advantages, reliable GaN HEMTs have been impeded by challenges such as electron trapping, which significantly impairs transistor functionality. Accurately modeling these effects is essential for the successful deployment of GaN-based designs. Over the past 15 years, the focus has been on empirical models suitable for circuit design. These models often result in a disconnect between modeling and technological advancements, thereby limiting further optimization of device performance. This study presents a novel trap model for GaN HEMTs that combines accuracy with physical principles. The model is broadly applicable, enabling the simulation of current collapse and knee walkout under various bias conditions, and is adaptable to RF applications with exceptional accuracy, through different trap-circuit configurations. The first model configuration is effective in simulating small- and large-signal continuous-wave (CW) RF measurements. It accurately mimics the slow response of electrons in traps to voltage reductions, ensuring efficient performance in simulations where traps maintain a steady state. Its accuracy has been validated through multiple measurements, including pulsed IV, S-parameter measurements, RF power sweeps, and load pull contours. Its development was enhanced by integrating experimental data and theoretical insights, which improved its accuracy and facilitated the development of additional analytical tools, such as statistical and compact modeling for trap localization. The second configuration is advantageous for pulsed RF and switching applications, as well as high-temperature environments. This configuration enables precise characterization of the swift capture time constant, a capability previously unattainable with conventional measurement techniques. Its development followed a systematic approach, beginning with the characterization of trap levels within the band gap of the devices, followed by the application of TCAD simulations to locate the position of traps and assess their impact on device performance. Ultimately, the second trap model configuration incorporates Shockley-Read-Hall (SRH) statistics to model temperature- and bias-dependent trapping, aiming to simulate the slow gate-induced trapping observed in rugged GaN-based low-noise amplifiers (LNAs). The inclusion of SRH statistics addresses significant challenges in GaN HEMT trap modeling.