SBIR Phase I: A-Plane Silicon Carbide Wafers

This Small Business Innovative Research Phase I project aims to investigate the feasibility of producing low-defect-density wafers of a-axis oriented SiC. Aymont will grow boules of SiC in the a-axis direction by physical vapor transport (PVT) to produce these wafers. When available, wafers are normally fabricated out of boules grown in the c-axis direction. This project will model thermal gradients during PVT growth to promote growth in the a-axis direction as well as moderate crystal expansion in the {0001} direction. Gas-phase precursors will be used in addition to solid SiC to counteract Si depletion during PVT growth and control stoichiometry. Methods used in epitaxial growth of SiC will be utilized in the beginning stages of bulk growth in order to promote step flow growth in the a-axis direction. Implementation of these innovative methods will enable low-defect-density a-axis wafers by the end of Phase I. Silicon carbide is well-established as a substrate material for high-power devices, microwave devices and GaN-based emitters. To date the orientation primarily utilized for these applications is {0001), which is largely a result of the relative ease of crystal growth in this orientation rather than advantages in device properties. An analogy can be drawn to silicon, where the (111) orientation is easiest for crystal growth. Metal-oxide-semiconductor (MOS)-based devices in Si ended up being widely adopted on (100) rather than (111) because of the reduced surface-state density in this orientation. Advantages of a-axis material with respect to {0001}include a lower surface state density, lack of polarization charge, and smoother surfaces after SiC epitaxial growth. Epitaxial growth occurs successfully over a wider range of growth conditions on a-axis substrates than {0001}. Polarization-induced charge in GaN-based devices should be eliminated if GaN is grown on a-axis SiC. This may lead to more rapid development of GaN-based power switching devices. Wider acceptance of a-axis substrates requires that they be available for purchase for industrial and academic research and development at prices lower than for (0001) SiC substrates today. The proposed work should make this possible.