We survey impact-ionization coefficients for silicon, wide bandgap semiconductors (gallium nitride and 4H-silicon carbide), and ultrawide-bandgap semiconductors (aluminum gallium nitride, beta-gallium oxide, and diamond). We employ a custom genetic algorithm to fit those existing coefficients into a modified Thornber model. This fitting process leads to an estimation of material properties, such as ionization energy, optical-phonon energy, and mean free path. After evaluating electric-field profiles by solving the Poisson equation, we use the impact-ionization integral to fundamentally calculate the breakdown voltage and the critical field of various p-i-n structures. This work captures how the doping concentration and the thickness of the drift layer shape the breakdown voltage as well as the critical field achievable with a given material. It is observed that a wider bandgap is not the sole requirement for the achievement of a higher breakdown voltage. Impact-ionization coefficients and dielectric constants are additional material properties that influence the calculation of the breakdown voltage. The performance limits of diamond are found to be surprisingly poor for its large bandgap. AlxGa1–xN and $\beta $ -Ga2O3 have almost the same performance limits, but it is observed that aluminum gallium nitride can outperform gallium oxide at high voltage if its background doping (doping floor) can be reduced.

中文翻译：

---

从冲击电离系数研究宽禁带和超宽禁带半导体

我们调查了硅、宽带隙半导体（氮化镓和 4H-碳化硅）和超宽带隙半导体（氮化铝镓、β-氧化镓和金刚石）的碰撞电离系数。我们采用自定义遗传算法将这些现有系数拟合到修改后的 Thornber 模型中。这种拟合过程导致对材料特性的估计，例如电离能、光声子能和平均自由程。通过求解泊松方程评估电场分布后，我们使用碰撞电离积分从根本上计算了各种引脚结构的击穿电压和临界场。这项工作捕捉了掺杂浓度和漂移层的厚度如何影响击穿电压以及给定材料可实现的临界场。据观察，更宽的带隙并不是实现更高击穿电压的唯一要求。碰撞电离系数和介电常数是影响击穿电压计算的附加材料属性。发现金刚石的性能极限因其大的带隙而出奇地差。AlxGa1–xN 和 $\beta $ -Ga2O3 具有几乎相同的性能限制，但据观察，如果可以减少背景掺杂（掺杂层），氮化铝镓在高电压下的性能优于氧化镓。发现金刚石的性能极限因其大的带隙而出奇地差。AlxGa1–xN 和 $\beta $ -Ga2O3 具有几乎相同的性能限制，但据观察，如果可以减少背景掺杂（掺杂层），氮化铝镓在高电压下的性能优于氧化镓。发现金刚石的性能极限因其大的带隙而出奇地差。AlxGa1–xN 和 $\beta $ -Ga2O3 具有几乎相同的性能限制，但据观察，如果可以减少背景掺杂（掺杂层），氮化铝镓在高电压下的性能优于氧化镓。