Raman spectroscopy of nanocrystalline diamond: An ab initio approach

The use of Raman spectroscopy to detect nano-sized diamond crystals is controversial; the origins of peaks at ∼1150 cm−1 in chemical vapor deposition nanodiamond films and ∼500 cm−1 in nanodiamond particles, which have both been suggested as evidence for nanophase material, remain uncertain. Many studies have produced evidence showing that the ∼1150 cm−1 peak is in fact due to polyacetylenelike structures at grain boundaries and interfaces, but little work has been done to confirm the assignment of the ∼500 cm−1 peak. In this paper we approach the problem from the molecular level, using Hartree-Fock theory to calculate the Raman spectra of diamond hydrocarbons, and observe the variation of the spectra with molecular size. Molecules with Td symmetry are studied, varying in size from adamantane to C84H64, an octahedral 1 nm-sized diamond crystallite. For comparison with nanodiamond thin films, the mass of the terminal hydrogen atoms were artificially increased to 100 amu, approximating the effects of matrix isolation. The calculated spectra are discussed in terms of the signals commonly observed in the Raman spectra of nanocrystalline diamond samples. This study finds no evidence for Raman active vibrations of diamond nanocrystals at either ∼1150 cm−1 or ∼500 cm−1, whether hydrogen terminated or confined in a matrix. Further, it appears that the only signals produced by a nanodiamond crystal are the broadened zone-center (1332 cm−1) mode and low frequency (<100 cm−1) deformations/Lamb-type vibrations. This suggests any other peaks observed in the Raman spectra of nanocrystalline diamond are due to defects, surface structures, amorphous material, or any other nondiamond material in the sample, and should not be taken as definitive evidence of nanocrystalline diamond.