Semiconductor Material and Device Simulations Involving Highly Mismatched Alloys
MetadataShow full metadata
In this thesis I investigate a specific highly mismatched alloy, BxGa1-xAs, over the full composition range using first principles DFT simulations with HSE06 hybrid functionals in VASP. I find that at low boron percentages the direct band gap decreases slightly, then increases towards the large minimum direct gap of BAs as more boron is added. My results show that the effect of isolated boron atoms on the band gap is small (<5%) at concentrations below 13%. I estimate that BGaAs transitions from direct to indirect band gap at around 17% boron content. I calculate the electron effective masses in the direct band gap region and investigate the effect of B-B pairs in nearest- neighbor group III sites on band gap, conduction band dispersion, and total free energy. I find that the lattice constant of BGaAs follows Vegard’s law and estimate that the boron concentration required to lattice match BGaAs to silicon is outside the direct gap regime. I then introduce TFETs as one possible application for highly mismatched alloys. Using the UCSD TFET model which I extended to include Kane’s non-parabolic dispersion relation I find the optimal combination of material and device properties that maximize I60. For low drain voltages (Vd = 0.1 V), the maximum I60 = 39 μA/μm occurs at moderate effective masses for both electrons and holes, while at larger drain voltages (Vd = 0.2, 0.5 V) the I60 continues to increase up to at least 1.3 m0, which makes highly mismatched alloys good candidates for TFET applications due to their increased effective masses.