Mechanisms for the control of two-mode transient stimulated Raman scattering in liquids

Abstract

Recent adaptive feedback control experiments demonstrated control of transient (i.e. nonimpulsive) Stokes emission from two closely spaced Raman-active modes in liquid methanol [e.g., B. J. Pearson et al., Phys. Rev. A 63, 063412 (2001)]. Optimally shaped pulses were found that selectively excited one of the two Stokes lines alone, optimized emission from both modes together, or completely suppressed all Stokes emission. Here, two general control mechanisms capable of affecting the ratio of intensities of the Stokes lines are identified. The first is operational when the duration of the pump pulse $({t}_{p})$ is on the order of the collisional dephasing time $({t}_{d})$. The ratio of the peak heights of the two Stokes lines can then be controlled by simply varying the duration and/or intensity of the pump pulse. The second operates when $1∕{t}_{p}$ is on the order of the energy separation of the two Raman modes, and hence when the two Raman modes are coupled due to overlapping nonlinear polarizations that drive the stimulated Raman scattering. In this regime, asymmetry in the spectral amplitudes within the pump pulse can control the asymmetry in the peak heights of the Stokes emission. Both these mechanisms have the same clear physical interpretation: shaping the pump pulse controls the nonlinear optical response of the medium, which in turn controls the stimulated Stokes emission, itself a ${\ensuremath{\chi}}^{(3)}$ nonlinear effect. In neither mechanism does the ratio of peak heights in the Stokes spectrum reflect directly the ratio of excited-state populations associated with the two Raman modes, as was assumed in the experiments, nor does the control involve coherent quantum interference effects.