Saturated fluorescence measurements of the hydroxyl radical in laminar high-pressure flames

We have investigated theoretically and experimentally the efficacy of laser-saturated fluorescence (LSF) for OH concentration measurements in high-pressure flames. Using a numerical model describing the interaction of hydroxyl with nonuniform laser excitation, we have studied the effect of pressure on the validity of the balanced cross-rate model and the sensitivity of the depopulation of the laser-coupled levels to the ratio of rate coefficients describing (1) electronic quenching to $\sp2\Sigma\sp+$(V$\sp{\prime\prime} \u3e$ 0) and (2) vibrational relaxation from V$\sp{\prime\prime} \u3e$ 0 to v$\sp{\prime\prime}$ = 0. At sufficiently high pressures and near-saturated conditions, the total population of the laser-coupled levels reaches an asymptotic value, which is insensitive to the degree of saturation. When the ratio of electronic quenching to vibrational relaxation is small and the rate coefficients for rotational transfer in the ground and excited electronic states are nearly the same, the balanced cross-rate model remains a good approximation for all pressures. When the above ratio is large, depopulation of the laser-coupled levels becomes significant at high pressures, and thus the balanced cross-rate model no longer holds. Under these conditions, however, knowledge of the depletion of the laser-coupled levels can be used to correct the model. A combustion facility for operation up to 20 atm was developed to allow LSF measurements of OH in high pressure flames. Using this facility, we achieved partial saturation in laminar high-pressure ($\leq$12.3 atm) C$\sb2$H$\sb6$/O$\sb2$/N$\sb2$ flames. To evaluate the limits of the balanced cross-rate model, we compared absorption and calibrated LSF measurements at 3.1 and 6.1 atm. The fluorescence voltages were calibrated with absorption measurements in an atmospheric flame and corrected for their finite sensitivity to quenching with (1) estimated quenching rate coefficients and (2) an in situ measurement from a technique employing two fluorescence detection geometries. While the absorption and calibrated fluorescence measurements compare well at 3.1 atm, the OH fluorescence values are $\sim$25% below the absorption measurements at 6.1 atm, indicating an effective error of $\sim$25% in the balanced cross-rate model. We anticipate that with atmospheric-pressure fluorescence calibration and a measurement or a reasonable estimate of the quenching correction factor, one can measure OH concentrations within $\pm$50% at pressures up to 10 atm