Diamond/c-BN HEMTs for power applications: A theoretical feasibility analysis

Abstract

Diamond is a promising material for high-power electronic applications in both the dc and rf domains. However, the predicted advantages are yet to be realized for a number of technical challenges. In particular, n-type devices have not been feasible due to the large ionization energies and low thermodynamic solubility limits of n-dopants. Motivated by the recent advances in nonequilibrium processing, we propose and theoretically examine a diamond/c-BN HEMT that can circumvent the critical limitations. A first-principles calculation suggests the desired type-I alignment at the heterojunction of these two nearly lattice matched semiconductors. The investigation also illustrates that a large sheet carrier density in excess of $5\times10^{12}~\mathrm{cm^{-2}}$ can be induced in the undoped diamond channel by the gate bias. A subsequent analysis of a simple prototype design indicates that the proposed device can achieve large current drive (~10 A/cm), low $R_{on}$ ($\sim 0.05 ~ \mathrm{m}\Omega \cdot \mathrm{cm}^2$), and high $f_T$ (~300 GHz) simultaneously.