UC Santa Barbara

Engineering of semiconductor material properties such as band gaps, lattice constants, and polarization charge densities is achieved largely through forming high quality alloys from similar materials. Extreme alloys formed from constituent materials with extreme differences in their base properties enables material property engineering across a larger range of values than traditional alloys. Such extreme alloys represent a scientific opportunity but also a challenge because they are more difficult to grow with high crystal quality than more conventional alloys. This dissertation will present research findings for thin films of the extreme III-nitride alloys InAlN, BGaN, and BAlN grown via the technique of plasma assisted molecular beam epitaxy (MBE).

The InAlN alloy system has band gap energies that span the entire visible range and is a promising candidate for band gap engineering. To facilitate band gap engineering, the band gap bowing for the entire composition range must be well characterized. In rich InxAl1-xN (x > 0.60) films were grown via plasma-assisted MBE on freestanding GaN substrates. The InxAl1-xN film compositions were determined using high resolution X-ray diffraction (HRXRD). On-axis 0002 omega two-theta scans were used to determine the c spacing of the films, reciprocal space maps of the 1 ̅015 peaks were used to determine that the films were completely relaxed, and the film compositions were calculated from that information assuming Vegard’s Law. The band gaps were independently measured using absorbance spectroscopy and fitting to the linear region of a Tauc plot (dependence of α2 hν on hν) as determined by interpolating the second derivative of the data and selecting a region of low curvature. To compare to literature, the data was fit to a composition-independent band gap bowing model and a bowing parameter of b = 4.0 ± 0.2 eV was calculated, which is consistent with previous results.

Incorporating B into the III-nitride system is predicted to enable higher density polarization charges than is currently possible using AlGaN. However, due to the significant difference between BN and the rest of the III-nitrides, growing high crystal quality films has proven difficult. Additionally B is a difficult material to use in a MBE reactor due to its high melting point as a pure compound and consequently, there is interest in exploring alternative sources for B in MBE.

The construction and operation of a novel BBr3 gas injection system for a plasma-assisted MBE reactor is detailed. Data from BGaN, BAlN, and hexagonal BN films are presented as a proof of concept for the source and significant additional characterization is presented for the BGaN system. We report the growth of high crystal quality, random alloy BxGa1-xN thin films with x up to 3.04% and thicknesses up to 280 nm grown on (0001) Ga-face GaN on sapphire. HRXRD was used to measure both the c plane spacing and the strain state of the films. It was determined that the films were fully coherent to the GaN substrate. Elastic stress-strain relations and Vegard’s Law were used to calculate the composition. Atom probe tomography (APT) was used to confirm that the BxGa1-xN films were random alloys. APT and secondary ion mass spectroscopy of a representative B0.03Ga0.97N film showed a high level of Br impurity on the order of 1 x 1019, atoms/cm3 and atmospheric contamination consistent with a low purity source. BBr3 is successful as a B source for high crystal quality BGaN films, however the bromine incorporation from the source limit the applications for BBr3 as a B source in molecular beam epitaxy.