Exploring the plasma chemistry in microwave chemical vapor deposition of diamond from C/H/O gas mixtures.

Microwave (MW)-activated CH(4)/CO(2)/H(2) gas mixtures operating under conditions relevant to diamond chemical vapor deposition (i.e., X(C/Σ) = X(elem)(C)/(X(elem)(C) + X(elem)(O)) ≈ 0.5, H(2) mole fraction = 0.3, pressure, p = 150 Torr, and input power, P = 1 kW) have been explored in detail by a combination of spatially resolved absorption measurements (of CH, C(2)(a), and OH radicals and H(n = 2) atoms) within the hot plasma region and companion 2-dimensional modeling of the plasma. CO and H(2) are identified as the dominant species in the plasma core. The lower thermal conductivity of such a mixture (cf. the H(2)-rich plasmas used in most diamond chemical vapor deposition) accounts for the finding that CH(4)/CO(2)/H(2) plasmas can yield similar maximal gas temperatures and diamond growth rates at lower input powers than traditional CH(4)/H(2) plasmas. The plasma chemistry and composition is seen to switch upon changing from oxygen-rich (X(C/Σ) < 0.5) to carbon-rich (X(C/Σ) > 0.5) source gas mixtures and, by comparing CH(4)/CO(2)/H(2) (X(C/Σ) = 0.5) and CO/H(2) plasmas, to be sensitive to the choice of source gas (by virtue of the different prevailing gas activation mechanisms), in contrast to C/H process gas mixtures. CH(3) radicals are identified as the most abundant C(1)H(x) [x = 0-3] species near the growing diamond surface within the process window for successful diamond growth (X(C/Σ) ≈ 0.5-0.54) identified by Bachmann et al. (Diamond Relat. Mater.1991, 1, 1). This, and the findings of similar maximal gas temperatures (T(gas) ~2800-3000 K) and H atom mole fractions (X(H)~5-10%) to those found in MW-activated C/H plasmas, points to the prevalence of similar CH(3) radical based diamond growth mechanisms in both C/H and C/H/O plasmas.

Optical emission from microwave activated C/H/O gas mixtures for diamond chemical vapor deposition.

The spatial distributions and relative abundances of electronically excited H atoms, OH, CH, C(2) and C(3) radicals, and CO molecules in microwave (MW) activated CH(4)/CO(2)/H(2) and CO/H(2) gas mixtures operating under conditions appropriate for diamond growth by MW plasma enhanced chemical vapor deposition (CVD) have been investigated by optical emission spectroscopy (OES) as a function of process conditions (gas mixing ratio, incident MW power, and pressure) and rationalized by reference to extensive 2-dimensional plasma modeling. The OES measurements clearly reveal the switch in plasma chemistry and composition that occurs upon changing from oxygen-rich to carbon-rich source gas mixtures, complementing spatially resolved absorption measurements under identical plasma conditions (Kelly et al., companion article). Interpretation of OES data typically assumes that electron impact excitation (EIE) is the dominant route to forming the emitting species of interest. The present study identifies a number of factors that complicate the use of OES for monitoring C/H/O plasmas. The OH* emission from EIE of ground state OH(X) radicals can be enhanced by excitation energy transfer from metastable CO(a(3)Π) molecules. The CH* and C(2)* emissions can be boosted by chemiluminescent reactions between, for example, C(2)H radicals and O atoms, or C atoms and CH radicals. Additionally, the EIE efficiency of each of these radical species is sensitively dependent on any spatial mismatch between the regions of maximal radical and electron density, which itself is a sensitive function of elemental C/O ratio in the process gas mixture (particularly when close to 1:1, as required for diamond growth) and the H(2) mole fraction.

Spectroscopic and modeling investigations of the gas phase chemistry and composition in microwave plasma activated B2H6/CH4/Ar/H2 mixtures.

A comprehensive study of microwave (MW) activated B2H6/CH4/Ar/H2 plasmas used for the chemical vapor deposition of B-doped diamond is reported. Absolute column densities of ground state B atoms, electronically excited H(n = 2) atoms, and BH, CH, and C2 radicals have been determined by cavity ring down spectroscopy, as functions of height (z) above a molybdenum substrate and of the plasma process conditions (B2H6, CH4, and Ar partial pressures; total pressure, p; and supplied MW power, P). Optical emission spectroscopy has also been used to explore variations in the relative densities of electronically excited H atoms, H2 molecules, and BH, CH, and C2 radicals, as functions of the same process conditions. These experimental data are complemented by extensive 2D(r, z) modeling of the plasma chemistry, which result in substantial refinements to the existing B/C/H/O thermochemistry and chemical kinetics. Comparison with the results of analogous experimental/modeling studies of B2H6/Ar/H2 and CH4/Ar/H2 plasmas in the same MW reactor show that: (i) trace B2H6 additions have negligible effect on a pre-established CH4/Ar/H2 plasma; (ii) the spatial extent of the B and BH concentration profiles in a B2H6/CH4/Ar/H2 plasma is smaller than in a hydrocarbon-free B2H6/Ar/H2 plasma operating at the same p, P, etc.; (iii) B/C coupling reactions (probably supplemented by reactions involving trace O2 present as air impurity in the process gas mixture) play an important role in determining the local BHx (x = 0-3) radical densities; and (iv) gas phase B atoms are the most likely source of the boron that incorporates into the growing B-doped diamond film.

Spectroscopic and modeling investigations of the gas-phase chemistry and composition in microwave plasma activated B2H6/Ar/H2 mixtures.

This paper describes a three-pronged study of microwave (MW) activated B(2)H(6)/Ar/H(2) plasmas as a precursor to diagnosis of the B(2)H(6)/CH(4)/Ar/H(2) plasmas used for the chemical vapor deposition of B-doped diamond. Absolute column densities of B atoms and BH radicals have been determined by cavity ring-down spectroscopy as a function of height (z) above a molybdenum substrate and of the plasma process conditions (B(2)H(6) and Ar partial pressures, total pressure, and supplied MW power). Optical emission spectroscopy has been used to explore variations in the relative densities of electronically excited BH, H, and H(2) species as a function of the same process conditions and of time after introducing B(2)H(6) into a pre-existing Ar/H(2) plasma. The experimental measurements are complemented by extensive 2-D(r, z) modeling of the plasma chemistry, which results in refinements to the existing B/H chemistry and thermochemistry and demonstrates the potentially substantial loss of gas-phase BH(x) species through reaction with trace quantities of air/O(2) in the process gas mixture and heterogeneous processes occurring at the reactor wall.

On the role of carbon radical insertion reactions in the growth of diamond by chemical vapor deposition methods.

Potential energy profiles for the insertion of gas phase C atoms, and CH, CH(2), C(2), C(2)H, and C(3) radicals, into C-H and C-C bonds on a 2 x 1 reconstructed, H-terminated diamond {100} surface have been explored using both quantum mechanical (density functional theory) and hybrid quantum mechanical/molecular mechanical (QM/MM) methods. Both sets of calculations return minimum energy pathways for inserting a C atom, or a CH(X), C(2)(X), or CH(2)(a) radical into a surface C-H bond that are essentially barrierless, whereas the barriers to inserting any of the investigated species into a surface C-C bond are prohibitively large. Reactivity at the diamond surface thus parallels behavior noted previously with alkanes, whereby reactant species that present both a filled sigma orbital and an empty p(pi) orbital insert readily into C-H bonds. Most carbon atoms on the growing diamond surface under typical chemical vapor deposition conditions are H-terminated. The present calculations thus suggest that insertion reactions, particularly reactions involving C((3)P) atoms, could make a significant contribution to the renucleation and growth of ultrananocrystalline diamond (UNCD) films.

Simplified Monte Carlo simulations of chemical vapour deposition diamond growth.

A simple one-dimensional Monte Carlo model has been developed to simulate the chemical vapour deposition (CVD) of a diamond (100) surface. The model considers adsorption, etching/desorption, lattice incorporation, and surface migration along and across the dimer rows. The top of a step-edge is considered to have an infinite Ehrlich-Schwoebel potential barrier, so that mobile surface species cannot migrate off the edge. The reaction probabilities are taken from experimental or calculated literature values for standard CVD diamond conditions. The criterion used for the critical nucleus needed to form a new layer is considered to be two surface carbon species bonded together, which forms an immobile, unetchable step on the surface. This nucleus can arise from two migrating species meeting, or from direct adsorption of a carbon species next to a migrating species. The analysis includes film growth rate, surface roughness, and the evolving film morphology as a function of varying reaction probabilities. Using standard CVD diamond parameters, the simulations reveal that a smooth film is produced with apparent step-edge growth, with growth rates (∼1 µm h(-1)) consistent with experiment. The ß-scission reaction was incorporated into the model, but was found to have very little effect upon growth rates or film morphology. Renucleation events believed to be due to reactive adsorbates, such as C atoms or CN groups, were modelled by creating random surface defects which form another type of critical nucleus upon which to nucleate a new layer. These were found to increase the growth rate by a factor of ∼10 when the conditions were such that the rate-limiting step for growth was new layer formation. For other conditions these surface defects led to layered 'wedding cake' structures or to rough irregular surfaces resembling those seen experimentally during CVD of nanocrystalline diamond.