Distributed Trap Levels and Hot-Electron Trapping in Power GaN HEMTs Characterization and Modeling

Gallium Nitride (GaN)-based high electron mobility transistors (HEMTs) are rapidly emerging as front-runners in high-power mm-wave circuit applications. Possible targets that would benefit of the advantages of GaN–based devices include efficient power supplies, DC/DC converters and AC/DC adapters, as well as the field of radars and telecommunications. Despite the recent commercial success of GaN-based devices, internal physical mechanisms are often not completely understood and still constitute a challenge to the development of a mature GaN-based technology. Besides the difficulties of growing a high–quality GaN material, point defects at interfaces play a major role in terms of reliability. In the present thesis, several aspect of the device instability have been investigated from both and experimental and theoretical point of view. Throughout this thesis, our goal is to build a quantitative and qualitative understanding of the main factors undermining the device stability under real application conditions. By comparing Hard and Soft switching turn-on commutations, we demonstrated how turn-on stress plays a major role in the on-resistance degradation, while off-state bias does not have a relevant influence on device properties. Furthermore, by repeating the experiment on several devices with different L_GD, we were able to demonstrate the important role of electric field in determining the R_ON increase and to rule out a significant contribution of self-heating. Then, in order to observe the full trapping and de-trapping kinetics of hot-electrons we focused on semi-ON stress analysis by means of a custom setup able to perform Drain Current Transient (DCT) analysis. Firstly, by focusing on the trapping phase, a physical understanding of the hot electron phenomena in GaN-based HEMTs is developed with a cross-comparison between theoretical analysis and experimental data. Linear dependency on the applied electric field and logarithmic dependency on the current density in determining the severity of current collapse are found. Results provide important information for the modeling of hot-electron trapping kinetics in GaN-based power transistors. The first 10 us of operation are critical in determining the current collapse during stress. Secondly, by focusing on the recovery phase, we propose a general methodology for mapping the properties (activation energy, cross sections) of a distribution of surface/interface states in GaN-based electronic devices. To prove the validity and usefulness of the model, the extracted map distributions are used as input for TCAD simulations. The results obtained by TCAD closely match the experimental transient curves, thus confirming the effectiveness of the developed technique. The theoretical knowledge built along this thesis allowed us to propose a new approach for compact modeling of the stretched exponential trapping/de-trapping kinetics of p-GaN HEMTs is proposed. Novel insight on the GaN HEMT dynamic performance degradation is given highlighting how the criticality of a trap is dependent on its location in the capture and emission time map. The duty cycle plays a key role in determining the performance degradation trajectory, while the frequency is related to the amplitude of the capture and emission process per cycle.