Pattern Recognition Analysis of Lattice-match and Thermo-match for Sample Collection Containing ZnO Semiconductor Material

Some pattern recognition methods including dynamic clustering analysis, minimal spanning tree and principal component analysis have been used to recognize the extents of both lattice-match and thermo-matchat300KforthesamplescontainingZnOsemiconductormaterial. Fromthepatternrecognitionanal-ysis, it is concluded that there exists a lower lattice and thermal misfitsbetween ZnOand threeⅢ-Ⅴgroup nitrides. The results alsoshow that pattern recognition analysisseems to be asuitable tool toselect substrate or bufferlayerforepitaxygrowth, aswellastodesignoptoelectronicdevicesofblueandvioletlightsemicon-ductor materials, such as ZnO and GaN.