Volume 103, Issue 21, 01 December 1995
Index of content:

Hyperfine analysis of the mixed A ^{3}Π_{1} v=28 and X ^{1}Σ^{+} v=69 states of I^{35}Cl
View Description Hide DescriptionThe A ^{3}Π_{1} v=28 state of I^{35}Cl is known from previous studies to be strongly mixed with v=69 of the X ^{1}Σ^{+} state. Spectra of both these mixed states are taken at high resolution to allow accurate measurement of the hyperfine splittings due to the iodine nucleus. The hyperfine structure in both parity components of the A state is found to be perturbed, though the common rotational‐electronic induced mixing should only affect the e components. Detailed analysis shows that hyperfine terms in the Hamiltonian also contribute significantly to the rovibronic state mixing. A Hamiltonian for these states is described that accounts for the hyperfine structure in both the A and X states to within the observed linewidth. The hyperfine parameters of this Hamiltonian, including the perturbation parameters, can be accounted for semi‐quantitatively in terms of a separated atom model.

Free jet IR spectroscopy of (^{32}SF_{6})_{2} in the 10 μm region
View Description Hide DescriptionThe rotation‐vibration spectra of (^{32}SF_{6})_{2} have been studied near the ν_{3} band of the ^{32}SF_{6}monomer. The parallel band 14 cm^{−1} below the monomer band origin shows a well resolved J‐structure, while the perpendicular band 8 cm^{−1} above the origin exhibits several Q‐branch peaks as the only resolved strong lines. The structure of (^{32}SF_{6})_{2} is consistent with a D _{2d } symmetry from the intensity alternation and the existence of a first‐order Coriolis interaction observed in the perpendicular band. The energy difference between the two bands is very close to the value calculated by a dipole–dipole and dipole‐induced dipole interaction model, while the location of the two bands is blueshifted from the calculated values by about 2 cm^{−1}. The possible influence of internal rotation is discussed.

Isotope effects and proton hyperfine interactions in the lowest ^{3} nπ* state of substituted benzaldehydes
View Description Hide DescriptionThe zero‐field splittings, principal spin axes, kinetic parameters, and nuclear hyperfine interactions of the ^{3} nπ* state of p‐chloro‐ and p‐methylbenzaldehyde and several of their deuterated derivatives are investigated by zero‐ and low‐field optically detected magnetic resonance(ODMR) at 1.4 K in a p‐dimethoxybenzene host. The zero‐field splittings show large isotope effects. These are interpreted in terms of spin–orbit interaction with the nearby but higher lying ^{3}ππ* state, yielding the energy gap between the two states in both benzaldehyde derivatives. The locations of the spin axes are approximately along the local symmetry axes of the carbonyl group and are insensitive to isotope. But, the spin axis most nearly normal to the plane of a host molecule deviates from the normal by an angle of 7°–13°. The kinetic parameters of the ^{3} nπ* state also are relatively insensitive to isotope. The dominant hyperfine interactions are associated with the aldehyde hydrogen and indicate that the ^{3} nπ* state is largely localized on the aldehyde moiety. Various properties of the ^{3} nπ* and ^{3}ππ* states are compared.

Transient hole burning of s‐tetrazine in propylene carbonate: A comparison of mechanical and dielectric theories of solvation
View Description Hide DescriptionThe solvation dynamics of s‐tetrazine, a nonpolar solute, in propylene carbonate, a polar solvent, have been measured in the temperature range of 190–300 K and the time range of 1.5–300 ps by transient hole burning. A detailed model of the gas‐phase spectrum of s‐tetrazine is used to extract purely solvent‐induced effects from steady‐state and ultrafast spectra. Absolute measurements of the solvation response function are extracted from these spectra and are compared to dielectric and mechanical theories of solvation. Although the theories postulate very different solute–solvent interactions mechanisms, either theory can account for the available data.

Far‐infrared permanent and induced dipole absorption of diatomic molecules in rare‐gas fluids. I. Spectral theory
View Description Hide DescriptionWe present a spectral theory for the far‐infrared absorptionspectrum of a very diluted solution of diatomic molecules in a rare‐gas fluid, that includes permanent and induced contributions. The absorption coefficient is given as the convolution of a translational spectrum and a rotational spectrum. The former is described in terms of time correlation functions associated to the induced dipole moment. The latter is discussed on the basis of a model consisting of a quantum rigid rotor interacting with a thermal bath, making use of time correlation functions associated to the different anisotropic orders of the solute–solvent intermolecular potential. Non‐Markovian and line mixing effects are taken into account. Explicit expressions for the five leading contributions of the induced dipole moment are given.

Far‐infrared permanent and induced dipole absorption of diatomic molecules in rare‐gas fluids. II. Application to the CO–Ar system
View Description Hide DescriptionThe general theory of far‐infrared absorption of diatomic molecules in rare‐gas fluids reported in Paper I is applied to the case of the CO–Ar system. The experimental absorption profiles of CO in Ar at three different densities (low‐density gas, high‐density gas, and liquid) are theoretically reproduced by considering the permanent dipole contribution and the four leading electrostatic multipole‐induced dipole contributions. From these results an estimation of the quadrupole, ‖Θ‖, octupole, ‖Ω‖, and hexadecapole, ‖Φ‖, moments of CO is given.

Cavity ringdown laser absorption spectroscopy and time‐of‐flight mass spectroscopy of jet‐cooled gold silicides
View Description Hide DescriptionThe cavity ringdown technique has been employed for the spectroscopic characterization of the AuSi molecule, which is generated in a pulsed supersonic laser vaporization plasma reactor. Fifteen rovibronic bands have been measured between 340 nm–390 nm, 8 of which have been analyzed to yield molecular properties for the X and D ^{2}Σ states of AuSi. This assignment is in disagreement with previous emission studies of AuSi, which had assigned the ground electronic state as a ^{2}Π state. A time‐of‐flight mass spectrometer simultaneously monitors species produced in the molecular beam and has provided evidence for facile formation of polyatomic gold silicides. Comparison of AuSi with our recent results for CuSi and AgSi indicates regular bonding trends for the three coinage metal silicide diatoms.

Millimeter‐wave rotational spectroscopy of MgOD and CaOD (X ^{2}Σ^{+})
View Description Hide DescriptionPure rotational spectra of CaOD and MgOD have been recorded in the range 200–390 GHz using millimeter/sub‐mm direct absorption spectroscopy. Transitions arising from the (000), (010), (020), and (100) modes have been measured for the ^{2}Σ^{+}ground electronic states of these free radicals. The data were analyzed successfully using a linear ^{2}Σ^{+} model for CaOD; for MgOD, only the (000) and (010) states could be fit with this Hamiltonian. Moreover, the (010) data required the addition of a substantial p _{Π} term to account for contamination of excited ^{2}Π electronic states. For both species, the α_{2} vibration–rotation term was found to be negative, in contrast to MgOH and CaOH, suggesting a less anharmonic contribution to the bending potential in CaOD and MgOD. These measurements also indicate a shorter O–H bond in MgOH than the other alkaline earth hydroxide radicals, which likely results because this species is quasilinear.

Photoionization spectroscopy of Ag–rare gas van der Waals complexes
View Description Hide DescriptionPhotoionization electronic spectroscopy is reported for the van der Waals complexes Ag–Ar, Ag–Kr, and Ag–Xe. Two band systems are observed for each complex correlating to the ^{2} P _{1/2}←^{2} S and ^{2} P _{3/2}←^{2} S atomic asymptotes. An additional weaker band system correlating to the ^{2} D _{5/2}←^{2} S asymptote is also observed for Ag–Ar and Ag–Kr, but not for Ag–Xe. Extensive vibrational progressions are observed in each of these band systems indicating that there is a large change in bond distance between the ground and excited states. Isotopic analysis confirms that these spectra all have onsets at high vibrational quantum numbers, making it possible to probe the shape of the potentials near the dissociation limits. Hotbands are also observed providing ground state vibrational intervals. Vibrational constants and dissociation energies are obtained for the excited states and dissociation energies are obtained for the ground states of each complex. The excited states correlating to the ^{2} P asymptote are significantly more strongly bound than the ground state for each complex, while the states correlating to ^{2} D are extremely weakly bound with low vibrational frequencies. Dissociation energy trends are compared for the series of complexes and for corresponding spin–orbit states.

The dynamics of succinonitrile in the plastic and liquid phases from the depolarized Rayleigh spectra
View Description Hide DescriptionThe paper reports the determination of the times of the orientational relaxation of a succinonitrile trans isomer in the plastic and liquid phase in the temperature range from 292.3 to 348.9 K, by the depolarized Rayleigh light scattering method. The activation energy of the process of the trans isomer reorientation in the plastic phase was found to be 11.6±4.2 kJ/mol. The decay constants of the interaction‐induced processes in the liquid and plastic phase as well as the squares of the effective anisotropy of the optical polarizability of both trans and gauche isomers and the anisotropy due to the interaction‐induced processes were determined.

A classical trajectory study of Ar+Ar_{2} collisions: Phase space structures in three degrees of freedom
View Description Hide DescriptionA classical trajectory study of Ar + Ar_{2} collisions is described. The system provides a model chaotic scattering system in more than two degrees of freedom. The initial conditions that give rise to ejection of each of the three Ar atoms, and the corresponding collision mechanisms, are investigated. There are some large regions of the initial condition space in which the identity of the atom ejected does not change, and other regions in which it changes rapidly. Attention is focused on long‐lived trajectories, which lie at the boundaries between different product identities. The long‐lived trajectories are associated with sequences of periodic orbits. The different stability possibilities for periodic orbits in three degrees of freedom are discussed, and a sequence of periodic orbits responsible for dividing the initial condition space is identified. These periodic orbits are born at ‘‘avoided bifurcations,’’ at which a saddle‐center bifurcation occurs close to a parent periodic orbit. The generalization to systems with more than three degrees of freedom is discussed.

Photodissociation of HF in Ar_{ n }HF (n=1–14,54) van der Waals clusters: Effects of the solvent cluster size on the solute fragmentation dynamics
View Description Hide DescriptionA comprehensive study of the photodissociation of HF in Ar_{ n }HF van der Waals clusters, with n=1−14,54, for an ultrashort δ(t)‐pulse excitation, is presented. The emphasis is on the dependence of the photodissociationdynamics of the HF solute molecule on the size and geometry of the Ar_{ n }solventcluster. This cluster size range encompasses formation and closing of the first solvation shell, which occurs for n=12, the addition of the complete second solvent layer (n=54), as well as the change of the HF location in the cluster, from a surface site for n≤8 to the interior of a cage for n≥9 clusters. Evolution of the fragmentation dynamics is revealed by following how the H‐atom kinetic energy and angular distributions, the survival probability, and cluster fragmentation patterns change as a function of the cluster size and structure. Classical trajectories are used to simulate the photodissociationdynamics. The probability distributions of the initial coordinates and momenta of the H and F atom are defined by accurate quantum five‐dimensional eigenstates of the coupled, very anharmonic large amplitude intermolecular vibrations of HF in the cluster. All aspects of the dissociation process studied here are found to exhibit a strong dependence on the size and geometry of the Ar_{ n }HF clusters.

Competition between electron detachment and monomer evaporation in the thermal destruction of hydrated electron clusters
View Description Hide DescriptionWe have examined the competition between electron detachment and monomer evaporation in the thermal destruction (dissociation plus detachment) of hydrated electron clusters by monitoring the products in a selected ion flow tube apparatus as (H_{2}O)^{−} _{ n }clusters, 14≤n≤24, were heated over the temperature range 100 to 150 K. The destruction of the smaller clusters is dominated by electron detachment, and the detachment occurs over the narrow temperature range 120–145 K. The larger clusters initially undergo sequential evaporation of neutral monomer units, forming smaller and smaller ionic clusters. As the temperature increases, the electron detachment channel begins to compete with monomer evaporation, and the smaller ions eventually decay by electron detachment. Second‐order rate constants and activation energies were obtained for the thermal destruction of clusters 14≤n≤17 and 23≤n≤24. The activation energies for the destruction of the larger clusters,n≥17, are nearly constant at ∼0.34 eV, which is close to the energy required to evaporate a single water molecule from the clusters, ∼0.40 eV. The difference indicates we are in the low‐pressure limit of thermal dissociation. The activation energy for the smaller cluster sizes, n<16, is significantly smaller than the monomer evaporation energy, and since the primary thermal destruction channel for these clusters’ is electron detachment, the activation energies determined here are a measure of the clusters adiabatic electron affinity. The estimated electron affinities for n=14 and n=15 are 0.12 and 0.23 eV, respectively. The electron affinities are in accord with that predicted by the dielectric continuum model. A model reported by Klots considering the temperature‐ and size‐dependent kinetics for the evaporation of particles from van der Waals clusters is in accord with the experimentally observed competition between these two cluster thermal decay processes.

Nonequilibrium solvent effects on the S _{ N }2 reaction using a self‐consistent reaction field continuum model based on multipole expansions
View Description Hide DescriptionA simple model has been developed that allows analysis of nonequilibrium solvent effects on chemical processes. It is based on the use of a self‐consistent reaction field approach using a multipole development of the solvation energy and on the separation of the inertial and noninertial polarization of the solvent. The solute’s wave function is computed at the ab initio level. The main advantage with respect to previously reported models is that the inclusion of nonequilibrium or dynamic solvent effects are introduced through the definition of a single solvent coordinate which is related to the chemical system coordinates. Besides, inclusion of polarization effects is straightforward. Results are presented for the S _{ N }2 reaction F^{−}+CH_{3}F→FCH_{3}+F^{−}. The frozen‐solvent hypothesis and the role of solvent fluctuations are discussed. It is shown that the climb to the transition barrier must be preceded by a convenient fluctuation of the solvent so that its inertial polarization component is suitable to solvate the transition state. Other solvent fluctuations, energetically less favorable, could decrease or even suppress the transition barrier. Nonequilibrium solvation effects on the value of the transmission coefficient are discussed. The methodology proposed in this work may be extended to the study of other rapid processes in solution such as proton transfers or electronic excitations.

Hydrophobic hydration in methanol aqueous solutions
View Description Hide DescriptionIn this work we present a Monte Carlo study of a methanol–water mixture. The model potentials used includes polarization and nonadditive effects in the intermolecular interactions. The parameters were fitted to two‐ and three‐body energy surfaces computed ab initio with a basis set of 2ζ plus polarization quality. Correlation was included at the MP2 level, and the basis set superposition error was corrected with the counterpoise method. Very long runs, 20×10^{6} configurations, were used to assure that equilibrium was attained. Good agreement with experiment was found for the structural data; the carbon to water oxygen and the hydrogen to hydrogen radial distribution functions. We studied the hydrophobic hydration of the methyl group and the effect of the solute on the water structure. We found that there is hydration of the hydroxyl group and a caging of the methyl group; that the solute does indeed affect the water structure at close distance to the solute but that this effect is masked by the average over the whole system. The water structure is disrupted by a slight broadening of the first peak in the O–O rdf and a shift of the second peak toward larger distances, we have also found that there is some methanol–methanol association.

The potential energy function for a ligand substitution reaction of square‐planar platinum (II) complex in water: The important role of three‐body effect
View Description Hide DescriptionThe analytical potential energy function for interaction of [Pt(NH_{3})_{3}]^{2+}, Cl^{−}, and H_{2}O has been determined to describe the ligand substitution reaction: [Pt(NH_{3})_{3}(H_{2}O)]^{2+} +Cl^{−}→[Pt(NH_{3})_{3}Cl]^{+}+H_{2}O in solution. The Honda–Kitaura potential function is used as the two‐body potential function. Although the Honda–Kitaura potential reproduces the ab initio two‐body interaction energy very well, the potential function that assumes pairwise additivity cannot reproduce the potential energy of the entire three‐body complex because of a large repulsive three‐body interaction in the strongly interacting region. We analyzed the origin of three‐body energies, derived the physically meaningful potential functional forms from a perturbation theory, and fitted the ab initio three‐body energies into an analytical form. The full potential function, the sum of pairwise two‐body potential functions and the three‐body terms, can reproduce very well the ab initiointeraction energies in the entire geometrical space as well as the structures of the reaction intermediates and transition states. The results of the preliminary reaction coordinate and Monte Carlo calculation for the reaction in a cluster of water molecules are also presented.

Geometric phase effects and wave packet dynamics on intersecting potential energy surfaces
View Description Hide DescriptionThe impact of the geometric phase on the time evolution of quantum‐mechanical wave packets is studied theoretically. Two model systems of coupled electronic potential energy surfaces are compared. One of them, the well‐known E×e Jahn–Teller system, comprises two conically intersecting surfaces, and the dynamics is subject to the geometric phase. The other system, describing the (E+A)×e Pseudo‐Jahn–Teller effect, comprises three intersecting surfaces and the dynamics is not subject to a geometric phase. Apart from the geometric phase, the coupling to the upper surface is verified to be negligible for low‐energy wave packet motion. Still, the geometric phase leads to a pronounced difference of low‐energy wave packet dynamics in both systems. Most significant is the phenomenon of destructive self‐interference of the two parts of the wave packet that encircle the conical intersection on opposite sides. The importance of the resulting different shape of the wave packet for a fs pump‐probe spectrum is pointed out.

A coupled cluster study of the structures, spectroscopic properties, and isomerization path of NCS^{−} and CNS^{−}
View Description Hide DescriptionThree‐dimensional near‐equilibrium potential energy surfaces and dipole moment functions have been calculated for the X ^{1}Σ^{+}ground states of NCS^{−} and CNS^{−}, using the coupled cluster method with single and double substitutions augmented by a perturbative estimate of triple excitations [CCSD(T)] with a set of 154 contracted Gaussian‐type orbitals. The corresponding equilibrium bond lengths at their linear geometries are r _{ e }(NC)=1.1788 Å and r _{ e }(CS)=1.6737 Å for NCS^{−}, and r _{ e }(CN)=1.1805 Å and r _{ e }(NS)=1.6874 Å for CNS^{−}. The predicted equilibrium rotational constantsB _{ e } of NCS^{−} and CNS^{−} are 5918.2 and 6282.7 MHz, respectively. The former agrees very well with the known experimental value (5919.0 MHz). Full three‐dimensional variational calculations have also been carried out using the CCSD(T) potential energy and dipole moment functions to determine the rovibrational energy levels and dipole moment matrix elements for both NCS^{−} and CNS^{−}. The corresponding fundamental band origins (cm^{−1}) ν_{1}, ν_{2}, and ν_{3} and their absolute intensities (km/mol) at the CCSD(T) level are 2060.9/306.1, 451.5/2.2, and 707.5/12.8, respectively, for NCS^{−} and 2011.4/6.6, 343.7/2.3, and 624.9/0.2 for CNS^{−}. The calculated ν_{1} (CN stretching) value for NCS^{−} is in very good agreement with the experimental result, 2065.9 cm^{−1}. The calculated dipole moments of NCS^{−} and CNS^{−} in their ground vibrational states are 1.427 and 1.347 D, respectively. The transition state geometry (saddle point) for the isomerization of NCS^{−}→CNS^{−} is predicted at the CCSD(T) level to be r(NC)=1.2044 Å, R(CS)=1.9411 Å and θ(∠NCS)=86.8°. Its calculated energy is 62.6 and 26.5 kcal/mol above the minima of NCS^{−} and CNS^{−}, respectively, including zero‐point energy corrections. The structure of the NCS radical was also optimized at the same level of theory, yielding ion to neutral bond length shifts in excellent agreement with those derived from recent photoelectron spectroscopy experiments.

Incorporation of solvent effects into density functional calculations of molecular energies and geometries
View Description Hide DescriptionIn this paper, we present the implementation of the ‘‘conductorlike screening model’’ (COSMO) into the density functional program DMol. The electronic structure and geometry of the solute are described by a density functional method (DFT). The solute is placed into a cavity which has the shape of the solute molecule. Outside of the cavity, the solvent is represented by a homogeneous dielectric medium. The electrostatic interaction between solute and solvent is modeled through cavity surface charges induced by the solvent. The COSMO theory, based on the screening in conductors, allows for the direct determination of the surface charges within the SCF procedure using only the electrostatic potentials. This represents the major computational advantage over many of other reaction field methods. Since the DMol/COSMO energy is fully variational, accurate gradients with respect to the solute coordinates can be calculated for the first time, without any restriction on the shape of the cavity. The solvation energies and optimized molecular structures are calculated for several polar solutes. In addition, the trends in basicity of amines and the relative stabilities of molecular conformers are studied. Our results suggest that for neutral solutes, agreement between calculated and experimental solvation energies of better than about 2 kcal/mol can be achieved.

Relativistic effects on sixth group monohydrides
View Description Hide DescriptionDirac–Fock and Hartree–Fock calculations have been performed for the ground state of the HO, HS, HSe, HTe, and HPo molecules. Equilibrium geometries, atomization energies, and ionization potentials, with both methods, are evaluated, compared, and discussed. Calculations on the molecules H_{2}M (M=O, S, Se, Te, and Po) have been already published [L. Pisani and E. Clementi, J. Chem. Phys. 101, 3079 (1994)], therefore, the results of the two series of molecules are compared. The effects of electronic correlation have been estimated by using the k‐functional technique [L. Pisani, L. De Windt, and E. Clementi, Int. J. Quantum Chem. (in press)]. The agreement with the experimental data, available for low Z, is satisfactory.