Improving Polytype Identification of Silicon Carbide Using Dictionary Indexing

Author(s):

Jonathan Ligda

Abstract:

Two of silicon carbide’s most prominent polytypes (6H and 4H) are difficult to differentiate between experimentally. The typical techniques for identifying the polytypes are X-ray diffraction and Raman spectroscopy. While both are useful for this task, they lack the spatial resolution to analyze sub-micron features in detail. Utilizing electron backscatter diffraction improves the resolution but the polytype identification can be difficult. For this work two indexing methods are compared, the traditional Hough-style indexing and a new dictionary indexing. Increasing the number of bands used in the Hough method improves the confidence index but the dictionary method shows the best results.