Heteroepitaxial diamond film formed on Si(001) wafer

摘要

Single-crystal diamond films that are necessary for electronic device applications have only been grown homoepitaxially on diamond substrates. Although the heteroepitaxial growth of single-crystal diamond films is still a challenge for today's scientists, it is the most hopeful way to success. There is some evidence that diamond can be grown epitaxially on mu n-sized cubic boron nitride (c-BN) crystals. However, Si is the most hopeful material for heteroepitaxial growth. The high lattice mismatch between silicon and diamond (the lattice parameter of Si is a=0.543 run; the surface energy of Si (111) plane is 1.5 J centre dot m~(-2) while the lattice parameter of diamond is a=0.356nm; the surface energy of diamond is 6J centre dot m~(-2)) has caused considerable controversy about the possible heteroepitaxial diamond growth on silicon substrates. Therefore, pretrea-tment of the substrate is commonly applied to obtaining diamond films on Si wafer. Textured diamond growth on beta-SiC and the growth of oriented diamond films a few mu m thick on silicon substrates have been obtained using bias-enhanced nucleation of diamond on silicon. Based on the experimental results mentioned above, Stoner et al. came to a conclusion that any observed local epitaxy ofdiamond on the Si substrate is most likely a result of limited diamond epitaxial growth on the interfacial [3-SiC layer. However, in this paper we submit the direct evidence of heteroepitaxial diamond directly formed on a mirror-polished silicon wafer observed by the high-resolution cross-sectional transmission electron microscopy (HREM). There is an angle of about 7.3° between Si (100) and D(100) heteroepitaxial crystalline planes. This is a rather encouraging finding and deserves careful theoretical and experimental attention.