Combined Spatially Resolved Optical Emission Imaging and Modeling Studies of Microwave-Activated H₂/Ar and H₂/Kr Plasmas Operating at Powers and Pressures Relevant for Diamond Chemical Vapor Deposition

Abstract

Microwave (MW) activated H2/Ar (and H2/Kr) plasmas operating under powers and pressures relevant to diamond chemical vapor deposition have been investigated experimentally and by 2-D modeling. The experiments return spatially and wavelength resolved optical emission spectra of electronically excited H2 molecules and H and Ar(/Kr) atoms for a range of H2/noble gas mixing ratios. The self-consistent 2-D(r, z) modeling of different H2/Ar gas mixtures includes calculations of the MW electromagnetic fields, the plasma chemistry and electron kinetics, heat and species transfer and gas−surface interactions. Comparison with the trends revealed by the spatially resolved optical emission measurements and their variations with changes in process conditions help guide identification and refinement of the dominant plasma (and plasma emission) generation mechanisms and the more important ArH, ArH2 and HH2 coupling reactions. Noble gas addition is shown to encourage radial expansion of the plasma, and thus to improve the uniformity of the H atom concentration and the gas temperature just above the substrate. Noble gas addition in the current experiments is also found to enhance (unwanted) sputtering of the copper base plate of the reactor; the experimentally observed increase in gas phase Cu* emission is shown to correlate with the near substrate ArH + (and KrH +) ion concentrations returned by the modelling, rather than with the relatively more abundant H3 + (and H3O +) ions.

Self-consistent modeling of microwave activated N2/CH4/H2 (and N2/H2) plasmas relevant to diamond chemical vapor deposition

“…Tendencies towards {CN}max saturation are observed, however, in the limits of high N2 additions (X0(N2) >> X0(CH4), reflecting a shortage of carbon species, rows 4 and 5) but not at high CH4 additions (X0(CH4) >> X0(N2), rows 6, 7 and 9 in Table 3 and Figure S2). The rates of the critical reactions (38) and (−38) are highly equilibrated (Table 2) and the calculated pressure dependence {CN}max ≈ 6.5×10 11 [21]). A similar analyses of the current model outputs return the following P-dependence {CN}max ~ 6.5×10 11 ×(P/1500) 1.85 (with the input power P in W), cf.…”

“…By way of illustration, the calculated total numbers of ions are 2.06×10 13 and 1.73×10 13 for the 0.06%N2/H2 and 6%N2/H2 mixtures, respectively. N2 addition also leads to some spatial redistribution of the averaged reduced electric fields |E|/(N×a) [38] (with the factor a = (1+ω 2 /veff 2 ) 0.5 for the MW frequency ω = 2π×2.45 GHz) in the near substrate region (Figure 3(a)), of the absorbed power density, PD, (in favour of the near substrate region and to a lesser extent the plasma core, Figure 3(a))…”

“…As before [37,38] Unfortunately, there is as yet little reliable data for such reactions (especially at high gas temperatures) and it is important to try and ensure that this does not lead to the neglect of potentially significant processes. As an example, the possible importance of H atom reactions with electronically excited N2* molecules could be a significant N2 decomposition route (and A c c e p t e d M a n u s c r i p t 9 a major source of NHx radicals) under the prevailing conditions, as argued in the next paragraph.…”

“…Note that the local mole fractions are not preserved at the initial X0(N2) values, however; as with other minority heavy species (e.g. hydrocarbons [62], Ar [38], etc) in a light gas (H2), thermodiffusive transfer and, to a lesser extent, H2 dissociation ensure that X(N2) in the hot plasma core is less than 40% the X0(N2) value.…”

“…To reproduce the measured {CN(v=0)}(z) column densities in the present 2-D modeling (Figure 6), it is necessary to increase the rate coefficient k−38 by reducing the activation energy from Ea ~0.92 eV (derived using the recommended value for k38 [67] and the relevant thermochemical data) to ~0.77 eV, i.e. by an amount less than the energy of the nascent vibrationally excited HCN(v1,0,v3)) products of the forward reaction (38). Some fraction of these product molecules will relax via vibrational-translational (V-T) energy transfer, but a comparable fraction will have a chance to undergo the fast reverse reaction (−38).…”