Library

Notes: Molecular-dynamics simulations are used to elucidate the molecular basis for the solvent effects on the isolated C–H stretching bands observed in the Raman spectrum of cyclohexane-d11. The main focus is on modeling the density dependence of the spectrum in supercritical CO2 recently reported by Pan, McDonald, and MacPhail [J. Chem. Phys. 110, 1677 (1999)], but several liquid solvents (CCl4, CS2, and CH3CN) have also been examined. The frequency shifts and line shapes of the Raman spectrum are simulated using a rigid solute and standard line shape theory in the limit of pure dephasing. Three models for the vibration–solvent coupling are considered. The simplest model, which is based on ground-state forces alone, provides a surprisingly good representation of the density dependence of the linewidths–line shapes but predicts the wrong sign for the gas-to-solution frequency shifts. This failure is due to the neglect of changes in bond polarizability upon vibrational excitation. Allowing for this polarizability difference via a semiempirical approach provides an accurate description of both the linewidths and frequency shifts with a physically reasonable vibrational difference potential. Interpretation of the instantaneous frequency shifts simulated with this model leads to the following general conclusions concerning the solvent effect on these spectra: (i) The relatively small gas-to-solution frequency shifts observed in experiment are the result of the near cancellation of much larger positive and negative contributions from repulsive and attractive interactions. (ii) Fluctuations in the instantaneous frequency are sufficiently fast (correlation times ∼100 fs) that the spectra are homogeneously broadened in all solvents examined. (iii) The dynamics of the solvent–solute interactions that determine the Raman line shapes are quite well described by an isolated binary collision ("IBC") type picture. (iv) The simplicity of the dynamics, and the success of this IBC description, is due at least in part to the special, localized character of these isolated C–H stretching modes. (v) The linear density dependence of the linewidths observed in supercritical CO2 reflects the modest extent of local density augmentation in the cyclohexane–CO2 system. © 1999 American Institute of Physics.

Notes: This paper describes results of simulations of solvation dynamics of a variety of solutes in two reference solvents, acetonitrile and methanol. Part of these studies involve attempts to realistically model the solvation dynamics observed experimentally with the fluorescence probe coumarin 153 (C153). After showing that linear response simulations afford a reliable route to the dynamics of interest, experimental and simulation results for C153 are compared. Agreement between the observed and calculated dynamics is found to be satisfactory in the case of acetonitrile but poor in the case of methanol. The latter failure is traced to a lack of realism in the dielectric properties of the methanol model employed. A number of further simulations are then reported for solvation of a number of atomic, diatomic, and benzenelike solutes which are used to elucidate what features of the solute are important for determining the time dependence of the solvation response. As far as large polyatomic solutes like C153 are concerned, the solute attribute of foremost importance is shown to be the "effective moment'' of its charge distribution (actually the difference between the S1 and S0 charge distributions). This effective moment, determined from consideration of continuum electrostatics, provides a simple measure of how rapidly the solute's electric field varies spatially in the important regions of the solvent. Simulations of fictitious excitations in a benzene solute show that this single quantity is able to correlate the dynamics observed in widely different solutes.Also explored is the effect of solute motion on its solvation dynamics. While of minor relevance for large solutes like C153, in small solutes of the size of benzene, solute motion can dramatically enhance the rate of solvation. A model based on independent solvent dynamics and solute rotational motion is able to account for the bulk of the observed effects. Finally, the influence of solute polarizability on solvation dynamics is considered. Simulations of diatomic molecules with a classical polarizability show that the rate of solvation decreases roughly in proportion to the polarizability of the solute. This dynamical effect can be understood in terms of the change that polarizability produces on the solvation force constant. These simulations indicate that the magnitude of the effect should be relatively small (10%–25%) in real systems, at least in the linear response limit. © 1995 American Institute of Physics.

Notes: Small hydrogen bonded complexes of HCN and DCN have been examined in the gas phase using FTIR and photoacoustic Raman spectroscopy (PARS). Four distinct bands are seen and assigned to CN stretching modes of the HCN dimer (D) and trimer (T). Previous microwave studies have shown the HCN dimer to be linear and analysis of the IR trimer band shape unambiguously demonstrates that the HCN trimer is also linear. The pressure dependent intensity of the observed transitions yields IR and Raman cross sections relative to the monomer of: (σD/σM)IR =30±10, (σT‘/σM)IR =500±200, (σD/σM)Ram =0.5±0.1, (σT/σM)Ram =3.3±0.7. The dramatic enhancements seen in the IR cross sections are due to small but significant (∼2×) changes in the CH bond dipole derivative. CARS spectroscopy has been used to examine the complexes formed in supersonic expansions of pure HCN (DCN) and HCN diluted in Ar and He carriers. Due to the cooling and collision-free conditions in the jet, this method yields higher effective resolution than is possible with equilibrium samples. The added ability to "tune'' the size of complexes by varying expansion conditions has allowed us to observe a number of other modes in HCN and DCN dimer and trimer as well as in higher polymers. These data have been used to partially refine a harmonic valence force field for the stretching modes of the dimer and trimer complexes.

Notes: The distributions of conformational defects that exist in the high-temperature phase II (also referred to as the hexagonal or rotator phase) of the crystalline n-alkanes C21 and C29 have been measured by an infrared CD2-substitution technique and have been accounted for in terms of a lattice model that provides freedom for longitudinal displacement of the chains. The defects consist almost entirely of gtg' kinks distributed nonuniformly along the chain. The uneven distribution is indicated in the variation in the concentration of gauche bonds measured at various sites along the chain. The highest concentration is at the chain ends, and the concentrations at interior sites decrease exponentially in going toward the middle. To explain the distribution we used a modification of a lattice model that had been successfully applied to the lipid bilayer. Comparison of observed distributions with those computed from the model indicates that the factors that determine the shape of the distribution are quite different in the n-alkane and bilayer cases. For the bilayer, the dominant factor is the variation in the lateral density of chains; for the n-alkane, the dominant factor is associated with longitudinal displacement of the chains.

Notes: Papazyan and Maroncelli [J. Chem. Phys. 95, 9219 (1991)] recently reported computer simulations of solvation dynamics of an ion in a Brownian dipole lattice solvent. In the present article we compare these results to predictions of a number of theories of solvation dynamics in the diffusive limit. The frequency-dependent dielectric response functions needed as input to many of the theories are derived from further simulations of the lattice solvent [H. X. Zhou and B. Bagchi, J. Chem. Phys. 97, 3610 (1992)]. When properly applied, all of the currently popular molecular theories yield reasonable predictions for the time scale of the solvation response. The dynamical MSA model [P. G. Wolynes, J. Chem. Phys. 86, 5133 (1987)] and the memory function theory of Fried and Mukamel [J. Chem. Phys. 93, 932 (1990)] both provide nearly quantitative agreement with all aspects of the solvation dynamics observed in these simulations.

Notes: Simulations of a simplified model system are used to test analytical theories of dielectric friction and explore its connection to dipole solvation dynamics. The simulation model consists of a point dipole solute interacting with a finite collection of dipolar solvent molecules, all situated on a simple cubic lattice and undergoing rotational Brownian motion in the pure diffusion limit. An extensive set of simulations are reported in which four model properties, the solute dipole moment and charge, and the solvent polarity and relaxation time, have been systematically varied. Static and dynamic aspects of dipole solvation observed in these systems are compared to the predictions of the simple continuum and dynamical mean spherical approximation (MSA) theories. Within the linear solvation regime the MSA theory is found to yield essentially quantitative predictions for both static and dynamic solvation properties. The simple continuum model, on the other hand, provides a poor description of either the static or the dynamic behavior. Solute rotational correlation functions of various rank and the dielectric friction functions calculated from them are compared to a variety of theories of rotational dielectric friction. Since all of the analytical theories examined rely on simple continuum descriptions of dipole solvation, they all fail to yield quantitatively accurate results. However, the more sophisticated theories do generally provide useful guides for understanding the trends observed in the data. The one instance where all of the theories fail in a qualitative manner is in predicting the rotational dynamics in the slow solvent limit. Reasons for this failure are discussed and a semiempirical approach for understanding the actual behavior in this limit is presented. © 1995 American Institute of Physics.