Aqueous solvation dynamics studied by photon echo spectroscopy

Abstract

Three-pulse photon echo peak shift measurements were employed to study aqueous solvation dynamics. A new perspective of dielectric continuum theory [X. Song and D. Chandler, J. Chem. Phys. 108, 2594 (1998)] aided in characterizing the system-bath interactions of eosin in water. Application of this theory provides solvation energies, which were used within the spectral density representation ρ(ω), to calculate the experimental peak shift. Simulations with only solvation contributions to ρ(ω), where a substantial amplitude of the solvation occurs within ∼30 fs, are remarkably consistent with our data. Furthermore, simulations using this theoretical solvation spectral density and an experimentally determined intramolecular spectral density yield an excellent total simulation of the peak shift data over the entire dynamic range. Our experimental results substantiate predictions that interaction-induced polarizability effects, contributing via a ∼180 cm−1 band in the spectral density, influence the initial dynamics. Three-pulse photon echo peak shift measurements were employed to study aqueous solvation dynamics. A new perspective of dielectric continuum theory ͓X. Song and D. Chandler, J. Chem. Phys. 108, 2594 ͑1998͔͒ aided in characterizing the system-bath interactions of eosin in water. Application of this theory provides solvation energies, which were used within the spectral density representation ͑͒, to calculate the experimental peak shift. Simulations with only solvation contributions to ͑͒, where a substantial amplitude of the solvation occurs within ϳ30 fs, are remarkably consistent with our data. Furthermore, simulations using this theoretical solvation spectral density and an experimentally determined intramolecular spectral density yield an excellent total simulation of the peak shift data over the entire dynamic range. Our experimental results substantiate predictions that interaction-induced polarizability effects, contributing via a ϳ180 cm Ϫ1 band in the spectral density, influence the initial dynamics.

Real-time observation of the charge transfer to solvent dynamics

“…For such a distribution of t 2 times to arise, non-equivalent relaxed CTTS configurations must not interconvert within the decay time, because we observe the same rise of B60 fs at all emission energies below B30,000 cm À 1 . Consistent with this are the characteristic response times of water, which contain a comparatively slower (B1 ps) component 29,30 , associated to collective long-range solvent rearrangements. Thus, it is the distribution of solvent shell configurations around I À that govern the electron-ejection rate.…”

Towards a comprehensive model for the electronic and vibrational structure of the Creutz–Taube ion

“…Only 1% of the response occurs at intramolecular vibration frequencies, while the remaining 17% arises diffusely through the 100-800 cm K1 region. Interestingly, no response is found in the region of the water translational band near 180 cm K1 that is strongly coupled to the excitation in other chromophores (Lang et al 1999) but contributes only weakly to the intrinsic liquid density of states (Stern & Berne 2001), highlighting the danger in using known spectral responses of dissimilar compounds in modelling the Creutz-Taube ion. The calculated wavelength dependence of the solvation reorganization energy can be used to construct a realistic solvent model for use in non-adiabatic coupling simulations that recognize the tendency for the charge to delocalize in the Creutz-Taube ion.…”

“…However, in reality, this function must decay to zero at infinite frequency, and so it is common to assume either ur(u)Zexp(Ku/u c ) for some critical frequency u c or similar analytical forms (Gilmore & McKenzie 2007); these all result in multi-exponential-like forms for the Stokes function S(t). Results obtained for energy relaxation in liquid water indicate that these simple forms are inappropriate, however, as water displays (Lang et al 1999) a range of responses peaking at frequencies of approximately 2-10 cm K1 (diffusive motions), 60 cm K1 (hydrogen-bonding bend), 180 cm K1 (hindered translation) and 600-800 cm K1 (libration) and displays intramolecular bending and stretching responses. Also, the relative propensities of each of these relaxation phenomena are clearly related to the shape and nature of the chromophore.…”

“…We are, in particular, concerned instead with the reorganization-energy distribution function ur(u)Zj(u)/uZdl s /Zdu, where l s is the solvent reorganization energy (Passino et al 1997), In equation (2.3), the bar indicates thermal averaging over initial configurations of the solvent in equilibrium with the initial state of the chromophore, while excitation is assumed to occur as a result of application of a light pulse at time tZ0. The Stokes function, which by constraint takes on the limiting values Sð0ÞZ 1 and SðNÞZ 0, is typically dominated by multi-exponential decay functions depicting the dissipation of the optical energy into the solvent, although sinusoidal oscillations are also observed in S(t) reflecting long-lived energy transfer into specific intermolecular or solvent modes (Passino et al 1997;Lang et al 1999). The reorganization-energy distribution function ur(u) is finite at zero frequency, indicating that energy is transferred into very long-range motions in the solvent, and the simplest model of the effects of the fluid is the ohmic model, in which ur(u) is assumed to be a constant.…”

“…The dynamics of the interaction of a chromophore with its solvent is an intense field of research using theoretical (Zwanzig 1973;Calderia & Legget 1983;Mukamel 1995;Fleming & Cho 1996;Gilmore & McKenzie 2007), simulation (Maroncelli 1990;Carter & Hynes 1991;Song & Chandler 1998;Stern & Berne 2001) and femtosecond-spectroscopic means (Reid et al 1995;Passino et al 1997;Lang et al 1999;Hettich et al 2002;Kuhn et al 2002). The quantity of most interest is the thermally averaged solvent density of states r(u) weighted by the chromophore-solvent coupling as a function of frequency u.…”