In a recent paper, we used induced fluorescence of O2 excited by a broad-band ArF laser to measure the temperature in a CO2 laser beam. In this preliminary work, only the peak around 235 nm of the hot O2 laser-induced fluorescence spectrum was analyzed; it was calibrated by comparison with the same feature observed in the absence of the heating laser beam. The procedure, based on a comparison of experimental and simulated spectra, was very simple and efficient, but it suffered from several limitations. On one hand, it did not take advantage of the full experimental data set, the observed spectrum consisting of 5 peaks extending from 220 to 260 nm. More questionable was the calibration, since the two compared spectra recorded in the presence and in the absence of the CO2 laser beam were observed independently; therefore, the experimental conditions—in particular, the geometry of the absorbing zone—were not identical. On the other hand, the rotational variation in the predissociation rate in the B state of O2, which is, indeed, fairly large (in particular, for uB = 6, 7, 9), was neglected. In the present paper, we describe a more refined spectral analysis of the 230–260 nm region of the LIF spectrum. A comparison with the 300 K LIF spectrum is unnecessary when the spectral simulation analysis is extended to several peaks; the shape of the different features and the relative intensity provide internal consistency. The earlier neglect of the rotational variation of the B quantum yield resulted in an underestimation of the population of the higher vibrational levels (ux = 2, 3) and, therefore, of the temperature. Eventually, the actual temperature proved to be notably higher than our previous estimation (700 K).
Read full abstract