Physico-Chemical Aspects of Surface Treatments for Diamond Nucleation

Abstract

Diamond has been shown to nucleate and grow on a wide variety of materials.^1,2 It is known that nucleation density can be increased by several orders of magnitude when the substrate is roughened by abrasive powders (diamond, silicon carbide, cubic boron nitride, alumina).^2,3 Several alternative reasons for this effect have been suggested. One is the argument that diamond, diamondlike carbon or other carbonaceous residues from polishing or abrading with diamond powder are left adherent to, or embedded in, the polished surface and that these supply nucleation sites for subsequent diamond growth.⁴ Alternatively, it can be argued that highly disordered surface sites or microscopic crater edges in the polished surface create de facto high energy sites which can be preferred nucleation sites for diamond.⁵ Interestingly, Yarbrough² reported that when a diamond polished silicon surface was subsequently polished with cubic boron nitride, the observed nucleation density was reduced to well below that obtained using only a diamond polish. This in spite of the fact that it was also shown that diamond readily nucleates on c-BN, including the particles used to polish the samples. We report in this paper an Auger Electron Spectroscopy (AES) investigation of the residue left in the scratches after polishing silicon substrates with diamond, and cBN powders as well as graphite surfaces. Diamond films were subsequently grown using a conventional hot filament CVD process.