Computational Discovery of Ultra-Wide Band Gap Semiconductors for Radio Frequency Applications

The extreme radio frequency (RF) spectrum (50-1000 GHz) holds potential for future information transmitting and sensing electronics, including beyond-5G wireless communications. These applications will push the frequency, agility, bandwidth, and size-weight-power limits of today’s high-power high-frequency amplifier devices. Ultra-wide band gap (UWBG) semiconductors (E_g &gt; 4eV) are key enablers to increasing power density over current state-of-the-art materials like gallium nitride (GaN), yet only a handful of the many UWBG semiconductors have been explored for RF applications. Many of those proposed UWBG semiconductors like BN, AlN, and diamond, face growth, quality, and scaling challenges. In this work, we use first-principles calculations as outlined in Goerai <i>et al.</i>^[1] to screen through candidates from the Inorganic Crystal Structures Database and determine new UWBG candidates with the potential for improved ambipolar RF performance The candidates are ranked using the calculated Johnson figure of merit. This figure of merit measures signal amplification performance based upon a semiconductor’s attainable voltage and current which are governed by the critical breakdown field (E_b) and the charge carrier saturation velocity (v_s) respectively. These inputs are calculated using density functional theory (DFT) and previously validated semi-empirical models. Of the 1400 oxides, nitrides, carbides, borides, and sulfides screened, we find 175 with better n-type and p-type figure of merit than GaN. About 10% of these have better predicted performance than current UWBG semiconductors like AlN. Among the top candidates are a set of IA-IB-C₂ carbides and wurtzite structure II-IV-N₂ nitrides. This presentation reports these candidates, their structure-performance predicted trends, and our findings.<br/>[1] <i>Energy Environ. Sci., </i>2-19, <b>12</b>, 3338