Time‐resolved femtosecond CARS from 10 to 50 Bar: collisional sensitivity

Abstract

Femtosecond time‐resolved coherent anti‐Stokes Raman scattering (CARS) measurements are performed on neat N2 and O2 as well as on N2 in the presence of various collisional partners to examine the effects of collisions on the observed signal at pressures as high as 50 bar. An exponential‐gap energy‐corrected‐sudden (ECS‐E) scaling law is used to model the rotational energy transfer of pure N2 at these elevated pressures. After accounting for line‐mixing effects, the long‐time collisional decay (20–60 picoseconds behavior observed here is consistent with the behavior observed in previous time‐resolved CARS studies over the pressure range 0–5 bar. Despite the observation of significant pressure dependence on long‐delay timescales, the experimental results demonstrate negligible dependence on pressure and identity of colliding partner during the first 1–3 picoseconds, a time range that has been shown previously to be highly sensitive to temperature. These results suggest that a variety of recently developed time‐resolved CARS approaches that allow accurate thermometric measurements to be made over this short 1–3 picosecond range can be used at pressures up to 50 bar without the need to employ models accounting for collisional decay. Copyright © 2013 John Wiley & Sons, Ltd.

Time- and frequency-dependent model of time-resolved coherent anti-Stokes Raman scattering (CARS) with a picosecond-duration probe pulse

“…(3)] to simulate the time-dependent CARS signal. 23,25,53,76 As noted in Sec. III, NR contributions to the observed CARS signal are modeled via a rapidly decaying term in the relevant third-order response function; in particular, a large decay constant (five orders of magnitude larger than the corresponding resonant molecular-transition decay constants) is assumed.…”

“…Thus, these 3-ps-delay data represent an optimal compromise that allows detection of spectra essentially free of NR background while exhibiting no collision-dependent behavior at pressures as high as 20 bars. 9,10,53 Still, the changes in spectral profiles observed in the 39-ps-and 68-ps-delay data emphasize that much of the spectral information observed within the hybrid fs/ps CARS signal is contained in the time-dependent nature of this signal. Whereas these longer delay spectra contain coarse peaks associated with the same vibrational transitions observed in the 3-ps data, a unique fine structure is observed within the spectra at each of these delays, with the 39-ps data emphasizing mid-range rotational Q-branch frequencies and the appearance of bimodal rotational structure with peaks near J ≈ 22 and J ≈ 28.…”

“…Thus, in cases where the molecular decay timescale, τ ca = (2π c ca ) −1 , in which the multiplicative variable c is the speed of light, is long compared to the shapedlaser-pulse decay time constant, the observed spectral lineshapes will be dominated by the linewidth of the probe pulse; this will be true at atmospheric pressure in most gas-phase experiments, where collisional dephasing occurs on timescales of tens to hundreds of ps. 25,59 However, in cases in which the molecular decay timescale is fast compared to the probe decay time, as would be the case in most condensed-phase and potentially in high-pressure gas-phase CARS measurements, 53 the spectral linewidths measured in the outgoing CARS signal will be dominated by these natural Raman linewidths.…”

“…Under conditions where line-mixing effects are negligible (≤1 atm in the case of the high-temperature N 2 simulations described here), 23,25,53,76 these diagonal elements can be used directly as the rotational-state-dependent Q-branch decay constants, ca , within the third-order response function [Eq. (3)], where states |a and |c correspond, respectively, to J levels within the lower and upper vibrational manifolds associated with v = +1 vibrational transitions, and vibrational dephasing is neglected.…”

“…(3)], where states |a and |c correspond, respectively, to J levels within the lower and upper vibrational manifolds associated with v = +1 vibrational transitions, and vibrational dephasing is neglected. We note that, under highpressure conditions where it becomes necessary to account for line-mixing effects, the G-matrix formalism 23,25,53,76 can be readily applied. Here, G represents a block-diagonal matrix containing square submatrices corresponding to each vibrational manifold in the form…”