Fundamental Studies of the Silicon Carbide MOS Interface
Climate change has placed a spotlight on renewable energy. Power electronics are essential to minimize energy loss when electricity is converted to a form used on the power grid. With silicon devices now approaching performance limits, SiC MOSFET can deliver power electronics to greater heights. However, the power capability of SiC MOSFETs is constrained by having low interface carrier mobility. It was coincidentally discovered that MOSFETs with oxide grown in alumina tubes have significantly higher mobility. We believe that the large surface potential fluctuations in SiC MOS interface results in percolation transport, and sodium ions from the alumina tubes reduces these percolative effects. Fabrication of SiC MOSFETs with different oxide thickness can vary the surface potential fluctuations and is used to verify the impact of percolation transport on SiC interface mobility. Characterization techniques on SiC devices are adopted from their silicon counterparts. Many characterization techniques are not tailored to the specification of SiC materials and hence, result in conflicting results during comparison of data among different research groups. The later chapters discussed the inaccuracies in the MOS AC conductance technique caused by the non-linear surface potential – gate voltage relationship and an energy-dependent interface state density. Using an exact model, we quantify errors in the extraction of interface state density, capture cross section, and position of the surface Fermi level when analyzed using the standard Nicollian-Goetzberger equations. We show that the exponential dependence of capture cross section on energy near the band edges is an artifact of the data analysis.
Cooper, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our