Investigation of stagnation-flow diamond-forming flames using advanced laser diagnostics

Gas phase temperature and atomic-hydrogen concentration profiles were measured near the deposition substrate in atmospheric-pressure, stagnation-flow diamond-forming flames. In these flames, a rich acetylene/oxygen/hydrogen mixture accelerates through a nozzle and impinges on a water-cooled molybdenum substrate, stabilizing a flat-flame approximately 1 mm below the substrate. A thin, polycrystalline diamond film is deposited on the substrate under appropriate conditions of flame stoichiometry and substrate temperature. Scanning electron microscopy and surface Raman spectroscopy verified the quality and uniformity of the diamond films deposited with this flame. Coherent anti-Stokes Raman scattering (CARS) spectroscopy of hydrogen was used to measure temperature profiles between the substrate and the incoming premixed jet. Hydrogen atom concentration profiles were measured using three-photon-excitation laser-induced fluorescence (LIF). The CARS measurements showed peak flame temperatures of 3300 K, approximately 200 K above the adiabatic equilibrium temperature for these flames. Temperatures were measured to within approximately 50 {dollar}mu{dollar}m of the molybdenum substrate, sufficient to capture most of the steep temperature gradient near the substrate. The LIF measurements showed peak atomic hydrogen concentrations on the order of 6%, well below computed adiabatic equilibrium concentrations. The superadiabatic flame temperatures are believed to occur because of insufficient time for complete dissociation of unburned acetylene in these rich flames. Since atomic hydrogen is one of the dissociation products of acetylene, the observation of subequilibrium atomic hydrogen concentrations is to be expected. Measured temperature and hydrogen atom concentration profiles are in good agreement with a numerical flame model developed by Meeks and coworkers at Sandia National Laboratories, except that the measured distance between the substrate and the reaction zone is much less than predicted. The measurements show that the flame stand-off distance is approximately 0.6 to 0.8 mm along the stagnation streamline, while the model predicts a flame stand-off distance of 1.2 mm. The difference between the measured and predicted stand-off is probably due to differences between the actual and modeled velocity flowfield.