a lack of knowledge of the likely mechanisms (for example, ionic or molecular) of Molecular dynamics simulations were used to study the acid ionization of hydrochloric HC1 reactions. In the first colnputational acid (HCI) at the basal plane surface of ice at 190 kelvin, as a model for the acid ionization study of HC1 adsorption on PSC ice (14), it process in Antarctic polar stratospheric clouds (PSCs). Initial conditions for HCI placewas assumed that HC1 neither is incorpoment within the top bilayer of the ice lattice were selected on the basis of relevant dynamic rated into the ice lattice nor ionizes; the equilibrium adsorption-desorption conditions. Free energy changes calculated for the first resulting calculated surface coverages were step in the stepwise acid ionization mechanism ranged from -5.8 to -6.7 kilocalories per orders of magnitude too small as compared mole for various likely initial conditions. These results indicate that acid ionization is with experimental results (7, 9, 10). Althermodynamically favorable and that this process has important implications for ozone though there is some indirect experimental depletion mechanisms involving PSCs.evidence that ionization may occur (7,9,15), there is no direct experimental evidence for ionization. One phenomenological model assumes ionization (6), assisted by T h e mechanism of catalytic gas-phase Oj quasi-liquid surface layer (5) is not required a proposed quasi-liquid layer (5).depletion (1 ) during the Antarctic spring to induce ionization. In earlier theoretical work on HC1 acid requires the heterogeneous formation of acHanson and Ravishankara (7) observed ionization in liquid water, Ando and Hynes tive chlorine species such as C12 on PSCs that HC1 uptake at an ice surface was lim-(1 6) found a stepwise proton-transfer mechfrom inert chlorine reservoir species such as ited to about a monolayer under stratoanism: a first proton transfer from the HC1 HC1 and CIONOz (1-7). For these reserspherically relevant conditions, and diff~l-to a first H 2 0 , producing the contact ion voir species to react readily, they must be in sion into the bulk was ruled out. Upon pair (CIP), and a subsequent proton transfer ample supply at or near the surface of the reexposure to HCI after a 5-to 30-min from the first H,O to a second H 2 0 , proPSCs (3). Two primary types of PSCs exist cessation, the HCI uptake was limited to ducing the solvent-separated ion pair. Be-(4) in the Antarctic stratosphere: type I, about 40% of that of the unexposed surface, cause the aqueous reaction mechanism (1 6) believed to be primarily nitric acid trihyindicating that not all the HCI desorbed showed a dominant, thermodynamically fadrate (NAT), and type 11, primarily ordiwhen the substrate was no longer exposed vorable first step to produce the CIP, we nary ice. It is known that HCI is involved in to HCI. We interpret these observations as used only the first step as our model (1 7, the reactions of many of the other reservoir indicators of two forms of HCI at the sur-18). In the calculat...

Molecular dynamics (MD) simulations of the model SN2 reaction Cl−+CH3Cl→ClCH3+Cl− in water, and variants thereof, are presented. The resulting transmission coefficients κ, that measure the deviations of the rates from the transition state theory (TST) rate predictions due to solvent-induced recrossings, are used to assess the validity of the generalized Langevin equation (GLE)-based Grote–Hynes (GH) theory. The GH predictions are found to agree with the MD results to within the error bars of the calculations for each of the 12 cases examined. This agreement extends from the nonadiabatic regime, where solvent molecule motions are unimportant and κ is determined by static solvent configurations at the transition state, into the polarization caging regime, where solvent motion is critical in determining κ. In contrast, the Kramers theory predictions for κ fall well below the simulation results. The friction kernel in the GLE used to evaluate the GH κ values is determined, from MD simulation, by a fixed-particle time correlation function of the force at the transition state. When this is expressed as a (Fourier) friction spectrum in frequency, marked similarities to the pure solvent spectrum are observed, and are used to identify the water solvent motions that determine the transmission coefficient κ. The deviations of κ from unity, the TST value, are dominated by solvent motions (translational and reorientational) which on the time scale of the recrossings are essentially static configurations. The deviations from the frozen solvent, nonadiabatic limit values κNA are dominated by the hinderd rotations (librations). Finally, the underlying assumptions of the GLE and the GH theory are discussed within the context of the simulation results.

Molecular dynamics are computed for a model SN2 reaction Cl−+CH3Cl→ClCH3+Cl− in water and are found to be strongly dependent on the instantaneous local configuration of the solvent at the transition state barrier. There are significant deviations from the simple picture of passage over a free energy barrier in the reaction coordinate, and thus, a marked departure from transition state theory occurs in the form of barrier recrossings. Factors controlling the dynamics are discussed, and, in particular, the rate of change of atomic charge distribution along the reaction coordinate is found to have a major effect on the dynamics. A simple frozen solvent theory involving nonadiabatic solvation is presented which can predict the outcome of a particular reaction trajectory by considering only the interaction with the solvent of the reaction system at the gas-phase transition barrier. The frozen solvent theory also gives the transmission coefficient κ needed to make the transition state theory rate agree with the outcome of the molecular dynamics trajectories. This theoretical κ value, which is the implementation for the SN2 reaction of the van der Zwan–Hynes nonadiabatic solvation transmission coefficient, is in good agreement with the trajectory results. In contrast, a Kramers theory description fails dramatically.

No abstract

No abstract