Saturation effects in coherent anti-Stokes Raman scattering spectroscopy of hydrogen

Abstract

Saturation of coherent anti-Stokes Raman scattering (CARS) spectra of the Q(1) line of the hydrogen (1, 0) vibrational transition was investigated experimentally by using high-resolution lasers and theoretically by solving the time-dependent density matrix equations. The saturation behavior of hydrogen is complicated by the large Doppler width of the resonance and the high rate of velocity-changing collisions relative to dephasing collisions. Experimentally, GARS line shapes and saturation curves were measured in pure hydrogen at pressures of 100 and 3050 Torr. Surprisingly, the measured saturation intensity was found to be less at 3050 Torr than at 100 Torr. The time-dependent density matrix equations were numerically integrated to obtain CARS saturation curves and line shapes. Excellent agreement between calculated and experimental line shapes was obtained at both 100 and 3050 Torr, and the predicted saturation intensity was less at 3050 Torr than at 100 Torr. Based on the good agreement between theory and experiment obtained at 100 and 3050 Torr, the theoretical results were extended over a much wider pressure range, from 0.1 to 100,000 Torr. Below 1 Torr the saturation behavior is independent of pressure because collision times are long compared with times associated with laser excitation of the resonance, and the molecular response is completely transient. Between 1 and a few hundred Torr the saturation intensity increases as the rate of velocity-changing collisions increases. Above a few hundred Torr, however, the saturation intensity begins to decrease because the high rate of velocity-changing collisions ensures that all molecules couple effectively with the Raman-pumping lasers. Calculations show a minimum in saturation intensity at 2000 Torr. For even higher pressures, saturation is controlled by dephasing collisions, and the saturation intensity increases rapidly with pressure.