Field-resolved measurement of reaction-induced spectral densities by polarizability response spectroscopy.

The experimental design and theoretical description of a novel five-pulse laser spectroscopy is presented with an application to a pyridinium charge transfer complex in acetonitrile and methanol. In field-resolved polarizability response spectroscopy (PORS), an electronically resonant laser pulse first excites a solvated chromophore (reactant) and off-resonant Raman spectra of the resulting nuclear motions are measured as a function of the reaction time. The present apparatus differs from our earlier design by performing the Raman probe measurement (with fixed pulse delays) in the frequency domain. In addition, the full electric fields of the signals are measured by spectral interferometry to separate nonresonant and Raman responses. Our theoretical model shows how the PORS signal arises from nuclear motions that are displaced/driven by the photoinduced reaction. The field-resolved off-resonant (of the solute's electronic transitions) probing favors detection of solvent (as opposed to solute) dynamics coupled to the reaction. The sign of the signal represents the relative strengths of polarization responses associated with the ground and photoexcited solutions. Signatures of nonresonant and PORS signal contributions to the experimental results are analyzed with numerical calculations based on a theoretical model we have developed for reaction-induced PORS. Our model identifies two mechanisms of PORS signal generation: (i) structural relaxation induced resonance; (ii) dephasing induced resonance. In the charge transfer reaction investigated, the solvent-dependent and time-evolving (solvent) polarizability spectral density (PSD) is readily obtained. The general trend of an initial broadband inertial nuclear response followed by a decrease in the linewidth of the PSD establishes that the measured PSD is inconsistent with the approximation of a linear response. Furthermore, the explicit time evolution of the PSD is important for properly describing solvent control of reactions that do not satisfy the time-scale separability inherent to nonadiabatic kinetic models.