Fast switching FIN-JFET transistor design in GaN substrate and modified parameters with meta-heuristic algorithms
Year: 2025
Type: article
Abstract: Technology based on gallium nitride (GaN) is swiftly expanding into new domains, including power and radio frequency (RF) electronics, digital and quantum computing. In the power domain, GaN devices have been marketed in the 15–900 V classes. Here, a novel approach is offered to optimize the design and functionality of these devices. It is essential to use mathematically rigorous, less computationally demanding, nature-inspired, predictive modelling methods. This study aims to improve the design parameters using adaptive neural network (ANFIS) and meta-heuristic methods by presenting the static and dynamic power and RF parameter performance of a vertical GaN fin channel JFET on a GaN substrate. The initial step in optimizing the circuit is to model the parameters of the target device using the adaptive neural fuzzy system. Afterwards, the channel's dimensions and impurity were computed with the aid of two optimization methods, namely, grey wolf optimization (GWO) and Harris Hawks Optimization (HHO) algorithms. For ANFIS modelling and device optimization using meta-heuristic techniques, MATLAB 2019b software was utilized. Silvaco software was then used to verify the results. The device exhibits an on/off current ratio of ∼109, a specific on-resistance (RON) of 0.82 mΩ•cm2, and a threshold voltage (VTH) of more than 0.5 V, extracted at a drain current of 1 mA. Verified using Silvaco simulations and MATLAB models, the optimization methods' outputs are in good agreement with the physical device simulator's outputs (with a difference of less than 7 %), indicating that the design parameters have been successfully optimized.
Source: Heliyon
Authors: Mohammad Hassan Kashefi, Javad Karamdel, Hassan Ghalami Bavil Olyaee
Language: English
Subfield: Condensed Matter Physics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,270
OpenAlex ID: W4407852246