Investigation of photochemical effects in flame diagnostics with picosecond photofragmentation laser-induced fluorescence

Abstract

Photofragmentation laser-induced fluorescence (PFLIF) is for the first time performed based on picosecond laser pulses for detection of hydroperoxyl radicals (HO2) in a stoichiometric laminar methane/air flame. Photofragmentation is performed with a pump laser pulse of 80ps duration and a wavelength of 266nm, whereupon the produced OH photofragments are detected by a second picosecond probe laser pulse, inducing fluorescence via excitation in the A2Σ+(v=1) ← X2Π(v=0) band of OH near 283nm. Excitation spectra of the OH photofragments formed in the reaction zone were recorded for pump-probe delays ranging from 0 to 5ns. The spectra suggest that the population distribution of the nascent OH fragments is rotationally cold and that it takes on the order of 5ns for the nascent non-equilibrium rotational distribution to relax into a thermal distribution. The radial OH-fragment distribution was extracted from spectral images (radial position versus emission wavelength) recorded at six different pump-probe delays. Photochemical OH production was observed both in the reaction zone and the product zone. Comparison with a kinetic model for OH production suggests that more than 20% of the oxygen fragments produced by photolysis in the reaction zone are formed in the excited 1D state, explaining a very rapid initial signal growth. The OH-production model was also compared with previous reaction-zone data, acquired with nanosecond laser pulses in the same flame, indicating that no O(1D), but only O(3P), is formed. A plausible explanation of the discrepancy between the two results is that the picosecond pulses, having more than two-orders of magnitude higher irradiance than the nanosecond pulses used in the previous study, might cause 2-photon photodissociation, allowing production of O(1D). In terms of flame diagnostics with PFLIF, it is concluded that a setup based on nanosecond laser pulses, rather than picosecond pulses, appears preferable since photochemical OH production in the reaction zone can be avoided while for short delay times the ratio between the photofragment signal and the photochemical interference in the product zone, stemming from CO2 photolysis, is sufficiently large to clearly visualize the photofragments.