The practical utility and unique properties of diamond films have produced a great deal of interest in the synthesis of diamond films by chemical vapor deposition. In most cases growth is initiated by the dissociation of a gaseous mixture of H₂ and a simple hydrocarbon precursor such as CH₄. Recently, extremely smooth (~30 nm rms roughness) diamond films have been grown in experiments at Argonne National Laboratory involving chemical vapor deposition following fragmentation of buckyball, C₆₀, in a microwave discharge, both with and without the addition of hydrogen. The C₂ molecule, produced in large amounts in the fragmentation of C₆₀, has been proposed as the principal growth species, with diamond growth occurring by insertion of C₂ into the C-H bonds of the hydrogen-terminated diamond surface.

A possible mechanism for growth of diamond thin-films based on C₂ as the growth species using computational quantum chemical methods has been carried out. This study has employed large cluster models for the diamond surface. The reaction energies and energy barriers for postulated steps in a mechanism, based on addition of C₂ to adjacent sites on a diamond (110) surface, were investigated. The model of the surface used is illustrated in the figure at the right. The addition of a C₂ to the trough in the hydrogen terminated surface is energetically very favorable, with energy lowerings of about 160 kcal per mole of C₂. The energy barriers for addition of C₂ to the surface are small. Adjacent C₂ moieties on the surface, adsorbed in ethylene-like arrangements, can be connected via a radical mechanism involving initiation by hydrogen atom addition to the double bond of one ethylene-like group or by a mechanism involving no hydrogen. Our results suggest that there is little or no energy barrier for either reaction. The completion of a new surface layer via formation of single bonds between the radical structure and ethylene-like adsorbates results in an energy lowering of about 40-50 kcal per mole of C-C bonds. This growth mechanism is unique in that it does not require hydrogen and may be responsible for the extremely smooth diamond films produced from buckyballs.

In contradistinction to diamond films grown by the traditional method involving CH4/H2 mixtures, where micron-size crystallites typically constitute the microstructure of the films, diamond grown from argon plasmas containing carbon is characterized by a microstructure consisting of crystallites with an average size of 3-10 nm. This nanocrystalline diamond can be grown to thicknesses of at least 30 µm from argon/fullerene or argon/methane microwave plasmas. A major advance was achieved recently at Argonne National Laboratory when it was discovered that diamond crystallite size can be controlled continuously and reproducibly from the micron to the nanometer size by the addition of Ar to CH4/H2 microwave plasmas.

In nanocrystalline diamond several percent of the atoms are located at the interface between grains. Therefore, electronic and mechanical properties of the diamond material are largely controlled by grain boundaries. A periodic density functional study of the high-energy p-bonded (100) stacking fault in diamond that can serve as a prototype of a twist grain boundary has been carried out. These calculations adequately reproduce the properties of (100) twist grain boundaries. The formation energy of the p-bonded planar defect is 4.2 J/m².

The electronic structure of diamond with defects is characterized by the presence of p-states in the forbidden gap that are localized on the interfacial atoms. This is illustrated in the figure below. These states produce several broad peaks in the gap and play a key role in the grain boundary conductivity. The defect region is very thin, having a thickness of one to two atomic layers. Formation of larger graphitic regions normal to the interface is shown to be energetically unfavorable due to a large lattice mismatch between graphite and diamond.  

1. "A Theoretical Study of the Energetics of Insertion of Dicarbon (C₂) and Vinylidene into Methane C-H Bonds" D. A. Horner, L. A. Curtiss, and D. M. Gruen, Chemical Physics Letters 233, 243 (1995).

2. "Theoretical Studies of Growth of Diamond (110) from Dicarbon." P. Redfern, D. A. Horner, L. A. Curtiss, and D. M. Gruen, Journal of Physical Chemistry 100, 11654 (1996).

3." Theoretical Studies on Nanocrystalline Diamond: Nucleation by Dicarbon and Electronic Structure of Planar Defects," D. M. Gruen, P. C. Redfern, D. A. Horner, P. Zapol, and L. A. Curtiss, Journal of Physical Chemistry, in press.