ALBERT

Notes: A theory is presented for the phonon inelastic scattering of light atoms and molecules from surfaces. Both the gas species and the thermal fluctuations of the solid are treated in a fully quantum fashion. A self-consistent field method is used to reduce the evolution of the reduced density matrix to the propagation of a single wave function and a set of coefficients describing phonon excitation and annihilation. The method allows one to extend recent time dependent molecule–surface scattering theories to finite temperature, with only a small increase in computer time. Agreement is found with experimental data for the thermal attenuation of diffraction peaks for He scattered from Cu. Energy transfer is found to be sensitive to the steepness of the repulsive potential, the molecular kinetic energy, and the angle of incidence, and only weakly dependent on the well depth. The "Beeby correction'' is examined and shown to be invalid, except at very low beam energies where there is a small correlation between well depth and inelastic scattering. For this model, energy transfer does not scale with the normal component of the beam energy.

Notes: Sub-Doppler resolution infrared spectra have been obtained for the ν1 bands of N2 –HCN and OC–HCN using the opto-thermal detection method, from which accurate ground and excited state molecular constants are determined. Vibrational predissociation lifetimes are estimated from the homogeneous broadening of the observed transitions, giving lifetimes of 80 and 2.6 ns for N2 –HCN and OC–HCN, respectively. These lifetimes are considerably longer than those obtained previously for N2 –HF and OC–HF. This difference can be understood in terms of the strength of the coupling between the intra- and intermolecular motions, which is also reflected in the vibrational frequency shift associated with complex formation.

Notes: The far infrared spectrum of gaseous chloroacetyl fluoride, CH2ClC(O)F, has been recorded at a resolution of 0.10 cm−1 in the 350 to 35 cm−1 region. The fundamental asymmetric torsional frequencies of the more stable trans (two halogen atoms oriented trans to one another) and high energy gauche (Cl–C–C=O torsional dihedral angle of 122°) have been observed at 86.5 and 48.8 cm−1, respectively, each with excited states falling to lower frequency. From these data the asymmetric torsional potential function governing internal rotation about the C–C bond has been determined. This potential function is consistent with torsional potential coefficients of: V1=350±12, V2=306±6, V3=420±1, V4=44±1, and V6=2±1 cm−1. The trans to gauche, gauche to gauche, and gauche to trans barriers have been determined to be 796, 245, and 271 cm−1, respectively, with an energy difference between the conformations of 525±24 cm−1 (1.50±0.07 kcal/mol). From studies of the Raman spectrum at variable temperatures the conformational energy difference has been determined to be 445±80 (1.27±0.2 kcal/mol) and 534±68 cm−1 (1.53±0.2 kcal/mol) for the gaseous and liquid phases, respectively. A complete assignment of the vibrational fundamentals observed from the infrared (3500 to 50 cm−1) spectra of the gaseous and solid states and Raman (3200 to 10 cm−1) spectra of the gaseous, liquid, and solid states is proposed. All of these data are compared to the corresponding quantities obtained from ab initio Hartree–Fock gradient calculations employing both the 3-21G* and 6-31G* basis sets. Additionally, complete equilibrium geometries have been determined for both rotamers. The results are discussed and compared with the corresponding quantities obtained for some similar molecules.

Notes: A detailed analysis of a wide range of magic angle sample spinning (MASS)/multiple-pulse NMR experiments is presented. The analysis depends upon the calculation of the magnetization trajectories of individual crystallites, and thus is valid for inhomogeneously broadened systems. Using this model we examine a variety of experiments used to control and manipulate rotational sidebands in MASS spectra. This model provides a straightforward analysis of both the two and four π pulse experiments used for total suppression of sidebands (TOSS), and allows us to derive new timings for both sequences. We are able to provide the first proof that the TOSS sequence exactly eliminates sidebands, and that the two-pulse sequence eliminates first-order sidebands. We also demonstrate that the TOSS cycle can actually lead to inverted spectra, and we have accurately quantified intensity losses which occur in both the two- and four-pulse versions. The effect of pulse errors on sideband suppression experiments is also briefly discussed, as well as one possible means to reintroduce information about shift anisotropies into TOSS-type experiments.

Notes: The atom–atom pair correlation functions and thermodynamics of the central force model of water, introduced by Lemberg, Stillinger, and Rahman, have been calculated accurately by an integral equation method which incorporates two new developments. First, a rapid new scheme has been used to solve the Ornstein–Zernike equation. This scheme combines the renormalization methods of Allnatt, and Rossky and Friedman with an extension of the trigonometric basis-set solution of Labik and co-workers. Second, by adding approximate "bridge'' functions to the hypernetted-chain (HNC) integral equation, we have obtained predictions for liquid water in which the hydrogen bond length and number are in good agreement with "exact'' computer simulations of the same model force laws. In addition, for dilute ionic solutions, the ion–oxygen and ion–hydrogen coordination numbers display both the physically correct stoichiometry and good agreement with earlier simulations. These results represent a measurable improvement over both a previous HNC solution of the central force model and the ex-RISM integral equation solutions for the TIPS and other rigid molecule models of water.

Notes: The adiabatic compressibility of ethanol–water mixtures has been measured as a function of temperature and alcohol mole fraction; the measurements were carried out at 5 °C intervals over the 10–40 °C range of temperature and over the entire range of cosolvent concentration. The experimental results are compared with other physicochemical measurements available for these mixtures and discussed in terms of possible mechanisms of molecular aggregation in the various regions of ethanol concentration.

Notes: We show that by means of a linear transformation of the trial functions in the second approximation to the distribution function, a simple expression for the thermal conductivity of polyatomic molecules can be obtained. This is made possible using our results of state-selective measurements of the nonequilibrium distribution in a heat conducting gas. Also a direct relationship between the pure gas transport coefficients can be deduced, which gives an accurate description of both monatomic and polyatomic gases. This is illustrated by a comparison between calculated and experimental data of the thermal conductivity.

Notes: The method recently developed to calculate the configurational density of states (or combinatory factor) on the regular solutions alloy model, from Monte Carlo simulation is applied to binary systems having bcc and fcc structures. This method enables one to calculate all the thermodynamic quantities of mixing of the model, including entropy. The results clearly show that the quasichemical approximation, although correct at high temperatures, does not yield a correct temperature dependence of the thermodynamic properties. In bcc alloys with ordering tendency, the Monte Carlo calculations yield configurational entropy curves that, up to now, have been interpreted in terms of the "associated model.'' We show that an exact statistical treatment of the regular solutions model is sufficient to reproduce the strong local order observed at low temperatures. The densities of states themselves are studied; They are Gaussian near the center of the distribution, satisfying the central limit theorem. In the tails of the distribution, the density of states decreases less rapidly with E ; in a certain region of energies, the density of states is roughly exponential, giving rise to a phase transition behavior. At the far extreme of the tails, the density of states decreases faster with energy.

Notes: The heat capacities of KOH and its deuterated analog KOD have been measured over the temperature range 17〈T〈342 K and 32〈T〈295 K, respectively. The thermodynamics of the phase transformations to the low-temperature antiferroelectric phases have been characterized: Ttr (KOH)=226.7±0.2 K; Ttr (KOD)=253.1±0.1 K; ΔStr (KOH)=0.121±0.002 R; ΔStr (KOD)=0.1258±0.0007 R. The thermodynamic changes are consistent with tunneling of the hydroxide (deuteroxide) ions between two minima in a double-well potential.

Notes: The Percus–Yevick equation for hard spherocylinders has been numerically solved using a recent algorithm to calculate the shortest distances between rods proposed by ourselves. The equation is solved for four different reduced densities to η=0.3879 and length-to-breadth ratio L*=1.0. Comparison with available Monte Carlo results for some selected orientations shows moderate agreement. Dependence of the pair correlation function and direct correlation function on mutual orientations and density are analyzed for these orientations. We found that none of the current approximations used in perturbation theories or in simplified integral equations can give even a qualitative description of the behavior of correlation functions for the above elongation L*.

Notes: An extension of the Poisson–Boltzmann theory of the electric double layer is applied to a symmetric electrolyte between two charged planar surfaces. This analytic treatment includes the effects of images and of ion correlations via the Debye–Hückel closure for the direct correlation function. It is demonstrated that at large separations the corrections to Poisson–Boltzmann theory appear as an effective surface charge. A means of correcting the apparent ion binding inferred from fitting experimental force data is also given. The extended theory shows good agreement with accurate numerical calculations for systems with low coupling. The generalization of the Onsager–Samaras limiting result to higher concentrations and to the case of the surface free energy of a charged dielectric interface is also presented.

Notes: A simple three-component lattice model employed previously to describe ternary mixtures with ionic surfactants is extended in order to describe mixtures with nonionic surfactants. This entails replacing the orientation-dependent hydrogen-bonding interaction between water and surfactant with an effective interaction which is isotropic but temperature dependent. We obtain the progression from two-phase to three-phase to two-phase coexistence with temperature which is typical of these systems. The behavior of characteristic temperatures with varying system parameters, such as the alkane number, is qualitatively reproduced.

Notes: We confirm the existence of a 15 A(ring) period in iodine-doped polyacetylene and provide a new interpretation for this key feature as part of a general model for structural changes during iodine doping. The observed diffraction intensities for different samples suggest the existence of structures with two different types of dopant-containing layers: layers obtained by complete replacement of polyacetylene chains by iodine columns (F layers) and layers obtained by replacement of every other polyacetylene chain by an iodine column (P layers). The F layers in the heavily doped complex alternate with dopant-free layers of polyacetylene chains (U layers), corresponding to a (UF)n stacking sequence. The phase obtained at a lower dopant concentration, which provides the 15 A(ring) spacing, is attributed to a (UPUF)n stacking sequence. At still lower dopant concentrations, one obtains a (UP)n stacking sequence. This model, along with published Raman, Mössbauer, and photoelectron spectroscopy data, suggests that the ratio of I−5 to I−3 increases in going from P layers to F layers. Intense and monotonically decreasing, diffuse x-ray scattering suggests that vacancies of size ∼3 A(ring) are present, probably in iodine columns. A diffuse reflection at 3.1 A(ring), observed in all iodine-doped samples, is due to an average iodine–iodine distance in disordered columnar arrays. On the other hand, ordered arrays of iodine columns in oriented samples give rise to sharp meridional reflections. All ten observed reflections (down to 1.17 A(ring)) in one sample could be indexed based on a 33.8 A(ring) repeat corresponding to (–I−3–I−5–I−3–)n arrays. The observed diffraction pattern was calculated from this model without using any freely adjustable parameters.

Notes: The definition for the surface scattering cross section of an object (defect, adatom, etc.) on a periodic surface in three dimensions is given. The analog of the Optical Theorem for gas phase scattering is derived. Calculations (made using semiclassical wave packets) are reported and compared to experiments for CO on Pt(111).

Notes: Our calculations of proton–anion distance distribution functions for HPTS dissociation recombination in the excited state are extended to the nanosecond time regime. When these functions are used for averaging the ground-state recombination probability, good agreement with experimental ground-state recombination yields is obtained. The relation with the theory of "delayed'' geminate recombination is briefly discussed.

Notes: Accurate values of the electronic transition moment function for the first positive (A 3∑+u−B 3Πg) and infrared afterglow (B 3Πg −B'3∑−u) systems of N2 were obtained through multiconfigurational linear response (also known as the multiconfigurational time-dependent Hartree–Fock, MCLR/MCTDHF). MCLR/MCTDHF furnished both excitation energies and transition moments. MCLR/MCTDHF calculations on the first positive system demonstrate the importance of configurations involving the 2σg orbital. To obtain accurate values, configurations were chosen with a complete active space of all valence orbitals (2σg, 2σu, 1πu, 3σg, 1πg, 3σu). This is the first MCLR/MCTDHF calculation for molecules as large as first row diatomics in which a complete active space including the entire valence shell is employed. The results are in substantial agreement with accurate theoretical calculations by others and, hence, disagree in their overall behavior with the previously published empirical values. One of our conclusions is that the experimental values for the first positive system far from the equilibrium region are in error.

Notes: A comprehensive theoretical investigation of the third-order susceptibilities for the nonlinear optical processes dc Kerr and dc electric-field-induced second-harmonic generation is reported for H2 and D2. Accurate wave functions of the James–Coolidge-type, which account for electronic correlation, are utilized and particular attention is given to the effects of vibration. Results for both the parallel and perpendicular components of the susceptibilities are given for λ=∞, 10 640, 6943, 6328, 5900, 5145, and 4880 A(ring). These are sufficient to enable a direct comparison with certain experimental data and, in general, the agreement is within the expected error bounds.

Notes: Optical measurements were used to detect structural anisotropy in concentrated dispersions over a range of Peclet numbers. Silica spheres of 49 and 130 nm radii with grafted octadecyl chains were dispersed in cyclohexane at volume fractions from 0.1 to 0.4. The apparatus consisted of a Couette cell with the dispersion in the annulus probed by a HeNe laser beam parallel to the axis of rotation. The dichroism and birefringence of the transmitted beam varied linearly with shear rate at low Peclet numbers with an orientation coincident with the principle direction of shear. Increasing the Peclet number, by increasing the particle size, produced a nonlinear response with the orientation tending to align in the direction of flow. A theory coupling the nonequilibrium microstructure under shear to the optical properties of the suspension enables direct interpretation of the dichroism. Comparison of sample dichroism calculations for two different forms of the theory demonstrates that the optical technique can be used to discriminate between theories which predict the microstructure.

Notes: New experimental and computational methods make possible measurement or simulation of the thermodynamic properties of fluids over wide ranges of temperature and density. We use a model for fluid thermodynamics based on the (1/r)n repulsive potential to show qualitatively what behavior one might expect for measurements of Grüneisen's gamma and acoustic velocity for compressed and expanded fluid metals and molecular fluids. We find limited regions of validity for such empirical rules as ργG =constant and linear density dependence of sound velocity. These results are compared with experimental data for fluid metals, and with N2 simulations presented in the accompanying paper.

Notes: (CH3)3 NHNO3 belongs to the monoclinic system with space group P21/c (C52h ) with four molecules per unit cell. Raman and infrared spectra (10–5000 cm−1) of single crystals and polycrystalline samples of (CH3)3 NHNO3 and of its N-deuterated derivative were investigated in the temperature range 90–415 K. Assignments of the measured frequencies to the internal vibrational modes were obtained. A systematic temperature dependent study of frequencies, bandwidths and integrated intensities was performed. The results indicate the existence of phase transitions at 359 and 407 K and both phase transitions show thermal hysteresis, characteristic of first order. In the high temperature phases II and I, the cations and anions achieve reorientational motion.

Notes: The Törring–Ernst–Kindt (TEK) model for the alkaline-earth monohalide molecules was extended to predict molecular polarizabilities and quadrupole moments in addition to dipoles. Calculations were carried out for the 19 molecules for which experimental bond lengths and vibrational frequencies are known. It was found that for many of these molecules the TEK model predicts negative polarizability anisotropies, α⊥〉α(parallel). This is in disagreement with the dipole-induced dipole, or Silberstein, formula, but in agreement with an empirical correlation between the anisotropy and the molecular asymmetry noted by Winicur. The TEK polarizabilities give rise to a much smaller second-order Legendre anisotropy in the CaCl–Ar interaction potential than was found in earlier work using the Rittner model. The D-shell model applied earlier to the alkali halides was also applied to the 19 alkaline-earth monohalides. With a slight modification it was found to describe both classes of ionic compounds successfully, although less successfully for the alkaline-earth compounds. The D-shell model was shown to be a generalization of the TEK model which includes the effect of the shell overlap on the polarizing fields at the ions. Nevertheless, the TEK model predicts better dipole moments, unless the shell charge is treated as an additional, adjustable parameter in the D-shell model.

Notes: Monomer concentration fluctuations in high molecular weight polymer blends are investigated. The fluctuations reduce the value of the χ interaction term in the Flory–Huggins free energy by a factor of (1-6/π2) and add a concentration and molecular weight contribution which decreases phase stability for off-critical compositions. Except at the critical point, the spinodal values are smaller than those predicted from a Flory–Huggins-type free energy. For an incompressible blend with components of equal size N, the corrections to the spinodal were found to vary as ||1−2φ0||4/3N−1/3, where φ0 is the mean concentration.

Notes: A model Hamiltonian approach to chemisorption is described which is a development of the model described previously for the cases of larger overlaps SA between metal wave functions and adorbitals. Inconsistencies due to basis set overcompleteness are cured. Nonorthogonality between adsorbate projected metal wave functions on different adorbitals is accounted for by explicitly evaluating the overlap integrals between them as well as the hopping terms in the metal part of the Hamiltonian due to coupling of metal wave functions through the adorbitals. Physically adequate hopping between adatom and metal through the core potentials is introduced for high overlap situations. The core hopping terms merge smoothly into the Wolfsberg–Helmholz approximation to core hopping, established earlier to lead to a reasonable description of chemisorption interactions at small overlaps. For the limiting case SA =1, due to completeness of the metal basis set, the adorbitals will be represented by the metal wave functions. k–k state scattering between metal electrons in the overlap region is introduced explictly. Electron repulsions between metal electrons in the overlap region are estimated subject to conditions of hybridizational and rotational invariance and requirements imposed by overcompleteness. The model Hamiltonian is applied to Li adsorption on the (100) surface of copper. The numerical results obtained are discussed extensively and compared to other theoretical work and to experiment.

Notes: The experimentally measured rotational diffusion times of various neutral solute molecules are compared with hydrodynamic and molecular models for microscopic friction. New viscosity and temperature dependent results for the rotational time of a large neutral solute (26 A(ring) long) indicate nearly perfect agreement with stick hydrodynamic predictions in both alkane and alcohol solvents. For smaller solute molecules, the results of previous studies show an increasingly large deviation from hydrodynamic predictions as the solute size decreases. This solute size dependent effect is compared with the theoretical predictions of Zwanzig's hydrodynamic "bumpy cylinder'' model, Dote–Kivelson-Schwartz's molecular free volume model, and a new generalized hydrodynamic model for the effects of frequency dependent viscosity on molecular rotational diffusion. Although hydrodynamic predictions agree with experiment in some respects, it is found that molecular aspects of the solvation process must be introduced in order to get quantitative agreement with the observed solute size dependence.

Notes: The rate of entropy production due to chemical reaction is calculated for a variety of parameter values in the reversible Oregonator model. The average values over cycles of oscillation are compared to those in the coexisting stationary states. The present work corrects an earlier calculation [A. K. Dutt, J. Chem. Phys. 86, 3959 (1987)]. Contrary to previous impressions and claims, there is no consistent relationship between the magnitudes of the entropy production in coexisting stationary and oscillatory states, in this case.

Notes: Rescigno and Schneider(ref.2), have suggested a simplification in the S−matrix version of Kohn variational method. It is pointed out that their idea works not only for electron−atom/molecule scattering, but also for heavy particle reactive scattering. Applications to atomic hydrogen and fluorine reactions with molecular hydrogen are considered.(AIP)

Notes: A Brownian dynamics simulation of a lipid chain is used to model the motional properties of a dipalmitoyl phosphatidylcholine bilayer. The effects of the bilayer environment on the chain are represented by a mean field derived from an extension of the Marcelja model. The simulation was run 44 million steps, the equivalent of approximately 0.66 μs for a viscosity of 2.2 cp. The results are compared with those of a 30 million step simulation of the chain in the absence of the mean field. Deuterium order parameters for the methylene groups along the chain and the average chain length calculated from the mean field trajectory are shown to converge to the experimentally determined values for DPPC with an appropriate choice of parameters. An analysis of the torsional dynamics of the chain, including transition rates and kink probabilities, is carried out. It is demonstrated that kink formation is sometimes, though not always, concerted. A comparison of the membrane and free chain simulations implies that the internal dynamics of a hydrocarbon chain in a bilayer is very similar to that of a neat alkane. Thus, the significant limits on the orientational freedom of the chain, as expected by the nonzero order parameters, are induced by a potential that has only a small effect on the local motions.

Notes: The ratio of the mean-square radius of gyration to mean-square end-to-end distance for linear chains and the mean-square radius of gyration ratio of ring to linear chains are investigated for two- , three- , four- , and five-dimensional polymers, with and without excluded volume, via Brownian dynamics. Our studies, for the Lennard-Jones repulsive interaction, indicate the universality of these ratios.

Notes: The Grassberger–Procaccia method has been employed to study the transitions which occur as a classical Ar3 cluster, modeled by pairwise Lennard-Jones potentials, passes from a rigid, solid-like form to a nonrigid, liquid-like form with increasing energy. Power spectra and lower bounds on the fractal dimensions and K entropies are presented at several energies along the caloric curve for the Ar3 cluster. In addition, the full spectrum of Liapunov exponents has been computed at these same energies to get an accurate value of the K entropy. Chaotic behavior, though relatively small, is observed even at low energies where the power spectrum displays largely normal-mode structure. The degree of chaotic behavior increases with energy at energies where some degree of regularity is observed in the spectrum. However, at energies that just allow the system to pass into and across saddle regions separating local potential minima, the phase space appears to be separable into a region within the equilateral triangle potential well where the behavior is highly chaotic, and a region of lower dimensions and less chaos around the saddle of the linear configuration. Dimensions from approximately three to eight are observed. A clear separability of time scales for establishment of different extents of ergodicity permits the determination of fractal dimensions of the manifold on which the phase points moves, for time scales of physical, i.e., observable significance. We believe this to be the first evaluation of the dimensionality of the space on which the phase point moves, for a Hamiltonian system displaying this range of dimensions.

Notes: Neutron and x-ray diffraction from vapor deposited krypton is reported. Strong diffuse intensity between the positions of the (111) and (200) reflections and between the (311) and (222) reflections of the fcc lattice is observed and ascribed to long range perturbations of the lattice.

Notes: Doppler-free polarization spectra of the A 1Σ+u (v=8, J=0–15) and b 3Π0u (v=14, J=0–15) states were observed. By analyzing the energy shifts, the spin–orbit interaction between the A 1Σ+u (v=8) and b 3Π0u (v=14) states was studied. The magnitudes of the hyperfine splittings were different in corresponding lines of the two states. In addition to the energy shifts, the line intensities and the line shapes were found to change remarkably near the perturbation. These are useful to study the mixing of the perturbing states.

Notes: The microwave spectrum of trimethylene sulphone has been studied in the ground and five vibrational excited states of the ring-puckering vibration. Very large deviations from the rigid rotor behavior were observed for the lowest pair of states. These perturbations resulted from a strong Coriolis coupling between them and have been analyzed using a reduced axis system Hamiltonian. The energy separation between the v=0 and v=1 levels has been determined from this analysis to be of 0.9106±0.0004 cm−1. Small deviations from rigid rotor spectrum were observed for the v=2 to v=5 ring-puckering states that were found to fit semirigid rotor theory where the quartic centrifugal distortion constants account for the vibration–rotation interaction effects. From the variation of the rotational constants and the vibrational separation between v=0 and v=1 states, a reduced ring-puckering potential function has been determined to be V(X)=6.62(X4−9.22X2), which gives a barrier to the planar configuration of 140±35 cm−1. With this potential function the vibration–rotation interaction effects observed for states v=2 to v=5 have been interpreted. Considerations about the ring conformation and the dynamics of the ring-puckering motion have also been made. The μa component of the electric dipole moment has been determined to be 4.8 D.

Notes: Two-quanta anharmonic spectra, namely overtones and combination bands with degenerate vibrations, in uniaxial gyrotropic crystals from point group of symmetry C3, D3, C4, D4, are investigated, with due account taken of the time inversion. The symmetry of two-phonon states is studied. A general model of the anharmonicity and the influence of the crystal chiral structure on intermolecular exchange of the circularly polarized degenerate vibrations is elaborated. The contribution of overtones and of the combination bands to dielectric permittivity tensor, and to the gyration tensor is studied too. The paper also deals with the IR and vibrational circular dichroism (VCD) spectra near the frequencies of the unbound and of the bound two-phonon states as well as the general features of two-phonon circularly polarized states appearing in the linear and in nonlinear optical processes in gyrotropic crystals.

Notes: The Fock transformation reduces the momentum-space Schrödinger equation for one-electron diatomics to a homogeneous Fredholm-type integral equation with a degenerate kernel. Two nonvariational linear combination of atomic orbitals (LCAO) solutions to this integral equation are rederived in a systematic manner and their mutual relation is clarified. The upper boundness of these nonvariational solutions is proved explicitly. Relative superiority of the two solutions is discussed theoretically based on their convergence rates. The momentum-space LCAO method is shown to reproduce the exact electronic energy for the hydrogen molecular ion at various internuclear distances.

Notes: The Doppler-free high-resolution spectrum of the B 1 Π–X 1Σ+ transition of NaK was measured by the technique of laser polarization spectroscopy. The molecular constants of the B 1 Π state, which reproduced the observed 831 unperturbed line positions (v=0–6, J=1–94) with a standard deviation of 0.002 cm−1 , were determined. Many perturbed lines, which were attributed to the perturbation between the B 1 Π and c(2) 3 Σ+ states, were observed. By analyzing the energy shifts of the B 1 Π(v=4,J) levels around J=13, we estimated the rotational constant Bv of the c(2) 3 Σ+ state to be 0.048 cm−1, and the matrix element of the spin-orbit interaction 〈c(2) 3 Σ+ vN=JJM||Hso|| B 1Πv=4JM〉 to be 0.14 cm−1 . We found that the strongly perturbed lines split into four lines, and we identified them as a hyperfine splitting caused by a mixing of the c(2) 3 Σ+ state. The splitting into four lines is explained by the magnetic dipole interaction due to a nucleus of I=3/2.

Notes: New time and energy resolved data on vibrationally relaxed and unrelaxed emissions from the valence a 4 Π(v=0), B 2 Π(v=0,3,4,5), and the Rydberg A 2 Σ+ (v=0,1,2) states of NO in Ne matrices are reported. Rydberg ↔ valence and valence ↔ valence nonradiative transitions are identified. The Rydberg → valence transitions are seen to occur after lattice relaxation accommodating the Rydberg orbital. The branching ratios for intramolecular relaxation and the measured lifetimes are described in terms of a model which combines the intramolecular spin–orbit matrix elements and Franck–Condon factors with the spectroscopically determined phonon Franck–Condon factors. For the levels B 2 Π(v=5,6), a Förster–Dexter-type energy transfer between NO molecules is also invoked in the description of the relaxation cascade.

Notes: Transient changes in polarizability during collisions between atoms and molecules give rise to interaction-induced rototranslational Raman scattering: the scalar component of the collision-induced polarizability Δα00 accounts for isotropic scattering, while the second-rank component ΔαM2 accounts for collision-induced depolarized scattering. We have evaluated the changes in electronic polarizability due to interactions between an atom and a molecule of D∞h symmetry in fixed configurations, with nonoverlapping charge distributions. We have cast the resulting expressions into the symmetry-adapted form used in spectroscopic line shape analyses. Our results are complete to order R−6 in the atom–molecule separation R. To this order, the collision-induced change in polarizability of an atom and a D∞h molecule reflects not only dipole-induced–dipole (DID) interactions, but also molecular polarization due to the nonuniformity of the local field, polarization of the atom in the field due to higher multipoles induced in the molecule, hyperpolarization of the atom by the applied field and the quadrupolar field of the molecule, and dispersion. We have analyzed the dispersion contributions to the atom–molecule polarizability within our reaction-field model, which yields accurate integral expressions for the polarizability coefficients. For numerical work, we have also developed approximations in terms of static polarizabilities, γ hyperpolarizabilities, and dispersion energy coefficients. Estimated polarizability coefficients are tabulated for H, He, Ne, and Ar atoms interacting with H2 or N2 molecules. The mean change in polarizability Δα¯, averaged over the orientations of the molecular axis and the vector between atomic and molecular centers, is determined by second-order DID interactions and dispersion. For the lighter pairs, dispersion terms are larger than second-order DID terms in Δα¯. In both Δα00 and ΔαM2, first-order DID interactions dominate at long range; other interaction effects are smaller, but detectable. At long range, the largest deviations from the first-order DID results for Δα00 areproduced by dispersion terms for lighter species considered here and by second-order DID terms for the heavier species; in ΔαM2, the largest deviations from first-order DID results stem from the effects of field nonuniformity and higher multipole induction, for atoms interacting with N2.

Notes: Ultrasensitive infrared laser absorption spectroscopy in a slit supersonic expansion is used to obtain the spectrum of the HF stretching fundamental of D2HF. Both a Π←Π band due to para-D2HF and a ∑←∑ band due to ortho-D2HF are observed, in contrast to the H2HF spectrum which consists of the Π←Π band alone. Analysis of the spectrum indicates that the D2HF Π states are more strongly bound than the ∑ states. Doublet splittings in the Π←Π band are analyzed to determine barriers to internal rotation of D2 within the complex. The vibrationa1 predissociation rate of D2HF is approximately 25 times faster than that of H2HF, suggesting the opening of a channel which results in vibrational excitation of the D2 fragment.

Notes: Charged particle production in thermal-energy K(nd)–CS2 collisions is investigated for intermediate values of n. The data show that collisions result in the formation of relatively long-lived CS−2 ions and of free electrons. A fraction of the CS−2 ions is observed to undergo rapid electric-field-induced detachment in fields of only a few kilovolts per centimeter and this novel phenomenon is discussed.

Notes: Subexcitation electrons lose their kinetic energy through vibrational excitation, rotational excitation, and elastic collisions in molecular gases. Initial yields of vibrationally and rotationally excited states of nitrogen molecules are calculated by using the Spencer–Fano equation (SFE) and its simplification, the continuous-slowing-down approximation (CSDA), both in time-independent and time-dependent representations. One focus of the present study is a close comparison of the CSDA with the rigorous treatment of the SFE in the subexcitation domain. The present result reveals for the first time distinct energy regions in which either vibrational excitation or rotational excitation dominates. This recognition explains the different time dependence of the yields of vibrational and rotational excitation.

Notes: We have calculated variationally highly excited vibrational (J=0) levels of the water molecule up to ∼27 000 cm−1 (relative to the minimum of the potential surface), for a global Sorbie–Murrell-type potential surface. The calculation has been performed in Radau coordinates, using the recently developed DVR-DGB variational approach [Z. Bacic and J. C. Light, J. Chem. Phys. 85, 4594 (1986); 86, 3065 (1987)]. 110 symmetric and 77 antisymmetric vibrational levels have been determined accurately, requiring diagonalization of relatively small Hamiltonian matrices of dimension ∼600. Many of the calculated levels correspond to large amplitude bending vibrations. Nearest neighbor level spacing statistics for the calculated levels above 18 000–20 000 cm−1 conform closely to a Wigner distribution, suggesting classically chaotic behavior in this energy range. Convergence rates of these variational calculations for H2O are comparable to those seen earlier for LiCN/LiNC and HCN/HNC. The DVR-based vibrationally adiabatic approach introduced by Light and Bacic [J. Chem. Phys. 87, 4008 (1987)] has also been tested here. Perturbative inclusion of the nonadiabatic corrections has allowed reliable identification of vibrational (J=0) levels of H2O up to 18 000–20 000 cm−1. With this model potential energy surface, reasonable agreement (∼1%) is obtained with experimentally known vibrational states to ∼20 000 cm−1.

Notes: The two-photon first ionization cross section was computed for H2O in the vertical transition approximation. A Stieltjes imaging technique was used to obtain from the results of a L2 basis set calculation, the two-photon oscillator strength density in the continuum. The calculations were performed in the independent particle static-exchange approximation for the final states. The ground-intermediate state transition matrix elements and excitation energies were computed by adopting the single channel static-exchange approximation or the multichannel random phase approximation; in the last case the intermediate-final state transition matrix elements were obtained by the equation of motion method. The calculated cross sections refer to the case of single beam of linearly or circularly polarized photons.

Notes: We present a new method for numerical integration of the radial electronic Schrödinger equation with these characteristics: (i) it uses a quantity directly related to the logarithmic derivative of the wave function, thereby facilitating the matching of solutions obtained for different radial regions; (ii) it avoids difficulty from the singularity of the logarithmic derivative at the nodes of the wave function; and (iii) it takes appropriate cognizance of the asymptotic form of the wave function at infinite radius. Examples are presented showing that eigenvalues can be obtained by the new method by outward integration alone, but that a combination of inward and outward integration leads to efficiencies which compare favorably with those achievable by the most popular previously existent method, that of Numerov.

Notes: SCF–MO and CI calculations are carried out for AuFq−6 (q=0, 1, 2, and 3) using a model potential method in which the major relativistic effects are incorporated. The calculated electron affinity of AuF6 is 9.56 eV, which is very close to Bartlett's estimation of 10±0.5 eV. The calculated values for the equilibrium Au–F distance and the totally symmetric vibrational frequency of AuF−6 agree well with experimental ones. The AuF−6 anion has the lowest minimum energy in the AuFq−6 (q=0, 1, 2, and 3) sequence, and the AuF3−6 anion is higher in energy by ∼4 eV than AuF−6 and AuF2−6 at their equilibrium positions. The effects of adding a d-type polarization function on F and an f-type polarization function on Au are discussed.

Notes: Two dynamic elements of the Willamowski–Rössler network are identified: one of which is a Lotka–Volterra oscillator involving two autocatalytic species X and Y, while the other is a switch between X and a third autocatalytic species Z. These two dynamic elements are coupled via X. Nonperiodic oscillations arise only if X autocatalyzes faster than Z. The chaotic nature of the oscillations is confirmed using the Shil'nikov theorem which requires the existence of a homoclinic orbit doubly asymptotic to a steady state of the saddle-focus type. Under chaotic conditions, two steady states coexist inside the positive orthant of concentration space and both satisfy the conditions of the Shil'nikov theorem. Chaos is further shown by the existence of several unstable period-3 fixed points of first-return Poincaré maps. A chaotic attractor is found and its changing structure under various sets of parameters is established.

Notes: The electronic structure of a metal heated above its critical temperature (an "expanded liquid metal'') shows dramatic changes as the density is increased. There is some experimental evidence that as one proceeds from insulating to metallic behavior the substance can even go through two separate phase transitions: from an ordinary insulator to a so-called excitonic insulator, and then from an excitonic insulator to a metal. In an effort to study how the metallic phase is approached, we have used discretized-path-integral methods to look at the statistical mechanics and the electronic structure of a model liquid. In the gas phase, the atoms in the model have a single valence electron constrained to occupy one of the s or p orbitals of the valence shell, but at higher densities, the orbitals hybridize, leading to instantaneous dipole–dipole interaction. We show that formulating this electronic structure in terms of occupation numbers allows us to monitor the hybridization evolution via an imaginary-time correlation function—which we calculate from an analytical solution to the mean-spherical approximation for the model. The numerical results strongly suggest that the model has a sudden hybridization transition, suggesting, in turn, that it might be profitable to think of the excitonic insulator transition in this language.

Notes: We investigate various properties of Blum's solution of the MSA for a mixture of dipolar hard spheres and equal-diameter charged hard spheres, fully rederived in a former publication. Numerical results for thermodynamic quantities, phase equilibria, and osmotic pressure are presented together with results for the ion–ion and ion–dipole correlation functions obtained via a simple integral-equation procedure. A generalization of the MSA results to a system of polarizable dipolar hard spheres and charged hard spheres is also given.

Notes: We have applied the polarizable electropole model of water and of water–ion interaction to molecular dynamics simulations of aqueous microclusters of alkali metal cations and halogen anions in the temperature range 150–400 K. This model effectively correlates observed enthalpies of formation; it is especially reliable for the larger alkali metal cations. Our simulations show unambiguous evidence of hydrogen atom inequivalency in the anion mono- and dihydrates over the temperature range studied. For the analogous cation hydrates, the structures are symmetric. We find two solvation shells in the larger microclusters; the largest ions have the fewest water molecules in their first solvation shells. The iodine and cesium inner solvation shells appear to contain only three and four water molecules, respectively. The structure of the anion hydrates strongly suggests that their stability arises in part from water–water binding.

Notes: A generalized Taylor dispersion analysis is presented of the combined diffusion and sedimentation of a Brownian particle subjected to a spatially periodic potential plus a uniform force. At long times this transport process is macroscopically characterized by a uniform mean drift velocity vector of the Brownian particle, and by a dispersion dyadic that quantifies stochastic "spreading'' about the (traveling) mean position. A general explicit integral formula is derived for the dispersion coefficient in the one-dimensional case, which supplements known results for the mean velocity. Illustrative calculations show that, for sufficiently large values of the drift force, the dispersivity exceeds the molecular diffusivity, which characterizes dispersion in the absence of any potential barriers. These calculations also lead to the surprising conclusion that an increase in the height of the potential barriers can sometimes lead to an increase in the dispersivity. A general asymptotic analysis is developed for the limiting case where the drift force tends to infinity. Although in one dimension the dominance of the drift force over the potential leads to a spatially uniform, steady-state, periodic distribution, the same behavior does not generally obtain in higher-dimensional cases. Asymptotic expressions are derived for the mean velocity and dispersivity characterizing one-dimensional systems and a class of two-dimensional systems. The analysis is subsequently applied to the related problem of sedimentation and dispersion of a particle within a medium for which its molecular diffusivity is a periodic function of position (in the absence of any potential). Although a periodic potential and a periodic diffusivity lead to equivalent mathematical problems in the absence of a sedimentation force, the presence of drift leads to important qualitative differences.

Notes: Vibrational spectra of CO physisorbed onto well defined NaCl(100) surfaces were studied using a Fourier transform infrared interferometer. Structures of CO starting from the monolayer to multilayers were explored. At 31.5 K and a CO pressure of 1×10−6 mbar only the monolayer is formed. Polarization measurements confirm our earlier study that the monolayer CO molecules are aligned perpendicular to the NaCl(100) surface. Increasing the CO pressure to 7×10−6 mbar produces multilayer adsorption. The multilayer spectra closely resemble that of α-CO absorption previously reported. The near perfect match of crystal structures and lattice constants of α-CO and NaCl is reasoned to force the epitaxial growth of single crystal multilayers in our experiments. At 22 K the monolayer absorption is at 2155.01 cm−1 with a bandwidth (FWHH) of 0.26 cm−1. The two prominent features in the multilayer spectra at 22 K are assigned to the longitudinal optical (LO) mode at 2142.54 cm−1 and the transverse optical (TO) mode at 2138.51 cm−1. Their frequency separation is a consequence of the lowering of the cubic symmetry of the bulk α-CO crystal by the shape, in the form of thin slabs, of our multilayer samples. Their bandwidths depend on the thickness of the sample and are characterized by a bandwidth parameter of 0.25 cm−1 for the LO mode and 0.85 cm−1 for the TO mode. The relative absorbances of these modes depend on the polarization of the infrared radiation. Theoretical formalism to account for the band splitting and absorption profiles of the infrared absorption is reviewed and applied to our measurements. While many features of our data can be explained by the present theory, further work is required to account for all the experimental results.

Notes: Chemisorption of hexafluoroazomethane on Si(111) was studied with synchrotron photoemission as a model system for studies of halocarbon-based etching. C 1s, N 1s, and F 1s core level spectra were collected for various room temperature exposures of clean silicon to hexafluoroazomethane, and after mild annealing. The initial reaction of hexafluoroazomethane with silicon is shown to be quite complex, resulting in the formation of many adsorbed dissociation products. Annealing resulted in the desorption and decomposition of these species. These results demonstrate the buildup of a passivating carbon layer on the silicon samples, which may be responsible for the inability of CFx radicals to etch silicon effectively.

Notes: Thirteen vibronic bands assigned to the two-photon 1 1Δu←A 1Πu transition, ν00=49 097±1 cm−1, have been observed by (2+2) resonance-enhanced multiphoton (REMPI) spectroscopy for laser wavelengths ranging from 370–480 nm. The measured term value Te=57 720±2 cm−1 of the previously unobserved 1 1Δu state is in good agreement with upper limits given by ab initio calculations. Additional spectroscopic constants measured for the 1 1Δu state are: D0=12 990 cm−1, ωe=1150±1 cm−1, ωexe=21.3±0.5 cm−1, Be=1.361±0.002 cm−1, αe=0.026±0.002 cm−1, and re=1.437±0.002 A(ring).

Notes: Transient and stationary fluorescence properties of rhodamine 6G in liquid ethanol have been investigated in a wide temperature range between 110 and 300 K. Various phenomena concerning the interaction between the dye molecule and the surroundings such as dynamic Stokes shift, stepwise temperature dependence of the fluorescence peak energy, and excitation-wavelength dependent fluorescence peak shift, have been observed. To explain these phenomena in a unified manner, a stochastic theory based on the Onsager cavity model has been developed. It is shown that the fluorescence as well as the absorption characteristics are explained well by this treatment. Various parameters characterizing the optical properties of the molecule, i.e., the dipole moments in the ground and excited states, transition energy in vacuum, temperature-dependent relaxation time of the dynamic Stokes shift and so on, can be derived from the comparison between the theory and experiment. Finally, the origin of the relaxation processes found in the present work is discussed in connection with the dielectric properties of the solvent and also with the resonance Raman scattering process.

Notes: We have performed a study of the water dimer interaction using larger basis sets and higher levels of theory than have been previously applied to this system. For the minimum geometry we have used spdf basis sets containing up to 212 orbitals. Our most accurate SCF interaction energy for the minimum is −3.73±0.05 kcal/mol. We have shown that this energy can be reproduced to within 0.1 kcal/mol using much smaller basis sets containing proper (diffuse) exponents. Accounting for the basis set superposition error is shown to be essential. We computed the dispersion energy with neglect of the intramolecular correlation using basis sets of various sizes. The best value obtained in a large spdf basis set with exponents which optimize this quantity is −1.93 kcal/mol and it is expected to be accurate to 0.1 kcal/mol or better. Using some of these basis sets we have performed supermolecular many-body perturbation theory (MBPT) and coupled-cluster (CC) calculations including triple excitations. We have shown that if the correlated supermolecular interaction energies are not corrected for the basis set superposition error, the correlated part of the interaction energy varies widely with the basis set. In contrast, the corrected values converge smoothly. On the basis of the dispersion energy and the MBPT/CC results we predict the correlated part of the interaction energy to be −1.0±0.3 kcal/mol, which leads to the total interaction energy of −4.7±0.35 kcal/mol. Thus, our work favors the lower limit of the experimentally predicted interaction energy of −5.4±0.7 kcal/mol.

Notes: The potential energy surfaces of all trans hexatriene and octatetraene are investigated within the harmonic approximation in the diabatic and adiabatic representations for the 1A−g, 2A−g, and 1B+u electronic states by an extended Pople–Pariser–Parr (PPP/CI) model. The effect of excitation and of vibronic coupling on the molecular force fields of the three states is examined. While electronic excitation affects only diagonal force constants of local oscillators, vibronic coupling changes drastically the couplings between local oscillators. The calculations reproduce well the observed increase of the frequency of the in-phase ag C=C stretch upon excitation to the 2A−g state.

Notes: At constant temperature and pressure, the differential of the Gibbs energy can be written in terms of species, components, extents of reaction, isomer groups, and homologous series groups under specified conditions. The applicability of each of these options in deriving expressions for equilibrium constants and devising other methods for calculating equilibrium compositions is discussed. If one or more reactants are available to a system at a constant chemical potential, their terms can be removed from the summation by use of a Legendre transform to obtain a new thermodynamic potential that is a minimum at equilibrium. This pemits the definition of homologous series and other groups that can be used as single terms in the fundamental equation for the purpose of equilibrium calculations. By use of these techniques, calculations can be made on organic systems with indefinitely large numbers of species.

Notes: We have measured the viscosity of four binary mixtures near their consolute points: (1) methanol+cyclohexane, (2) isobutyric acid+water, (3) nitroethane+3-methylpentane, and (4) 2-butoxyethanol+water. The viscosity data are consistent with the power-law divergence: η∼||T−Tc||−y, with an apparent viscosity exponent in the range 0.0404〈y〈0.0444. Recent theoretical estimates for y are near 0.032, which is outside the experimental range. The value of y is independent of whether the critical point is an upper or a lower consolute point and of whether the approach toward Tc is at constant pressure or at constant volume. Our torsion oscillator viscometer is unique in its simultaneous low frequency (∼1 Hz) and low shear rate (∼0.1 s−1), allowing its use close to the critical point before encountering non-Newtonian fluid behavior associated with critical slowing down. Nevertheless, we find quantitative evidence for viscoelasticity near the critical point.

Notes: We study the long-time self-diffusion coefficient and the zero-frequency effective viscosity for a suspension of spherical Brownian particles. The correction to first order in the volume fraction to the self-diffusion coefficient and the correction to second order in the volume fraction to the effective viscosity are given by integral expressions derived by Batchelor. Using exact series expansions of the hydrodynamic interaction functions in powers of the inverse distance between centers of a pair of particles we obtain accurate values for the correction terms. We consider hard spheres with mixed slip-stick boundary conditions as well as spherical liquid droplets with a viscosity different from the bulk.

Notes: Vibrational predissociation linewidths have been resolved in the two H–F stretching bands of the HF dimer using an optothermal (bolometer-detected) molecular-beam color-center laser spectrometer. In addition to the strong vibrational mode dependence reported earlier by several groups, we observe a substantial K-rotational and tunneling dependence to the longer-lived mode ν1, which is associated with the "free-H'' stretch. The predissociation linewidths (FWHM in MHz) for this vibration are 6.4(5) for K=0+, 9.5(5) for K=0−, 10.2(5) for K=1+, and 11.8(5) for K=1−, where the +/− superscripts refer to the symmetric/antisymmetric tunneling states. The J dependence (at low J) is negligible compared to the K dependence. The K=0 levels of the "bound-H'' stretch have tunneling-independent widths of 330(30) MHz. Extraneous broadening due to saturation effects was observed and corrected for in these measurements.

Notes: An approach is presented and demonstrated for determining predissociation lifetimes of small molecules in the vapor phase at room temperature through measurement of resonance Raman depolarization ratios. Depolarization ratios for low-pressure CS2 vapor have been measured using ten excitation wavelengths between 1995 and 2001 A(ring), spanning one of the strongest vibrational features of the S3←S0 electronic transition. The excitation wavelengths are obtained by anti-Stokes Raman shifting in H2 the fourth harmonic of a Nd:YAG laser and the second harmonic of a YAG-pumped dye laser. The Raman depolarization ratio ρ(I⊥/I(parallel)) varies from less than 0.4 with excitation on the high-frequency side of the vibronic band to nearly 0.6 on the low-frequency side. Numerical simulation of the depolarization dispersion curve using literature values for the rotational constants shows that this behavior can be attributed straightforwardly to partially selective excitation of initial J states for which the S3 predissociation lifetime is either long or short relative to rotational periods. The best fit between the experimental and calculated depolarization dispersion curves is obtained with a homogeneous linewidth of 6 cm−1 (excited state lifetime 0.9 ps). The relevance of this result to measurements of dissociation times through photofragment anisotropy analysis is discussed.

Notes: The collision induced dissociation (CID) and charge transfer (CT) cross sections have been determined for H+2 and D+2 colliding with H2 using a surface hopping trajectory method. Approximately 40 000 trajectories have been analyzed for collisions at 4.0, 6.0, and 8.0 eV (center of mass) and for H+2 (D+2 ) in vibrational states from 0 to 10. Our results are consistent with the recent experiments of Guyon, Baer, Cole, and Govers [Chem. Phys. 119, 145 (1988)]. However we have come to a different understanding of the mechanism for dissociation. We find that there are two pathways for CID: (1) formation of a H+3 intermediate followed by dissociation and (2) direct dissociation of a H+4 transition state via vibrational excitation. The H+3 intermediate pathway predominates at low collisional and low H+2 (D+2 ) vibrational energies.

Notes: Comparisons of three variational principles commonly used in scattering problems, namely those due to Kohn (KVP), Schwinger (SVP), and Newton (NVP), are presented. These comparisons are conducted by computing K-matrix elements for elastic scattering from nine different interaction potentials. We represent the KVP trial functions as expansions containing two non-L2 terms that represent the asymptotic free wave, and a set of L2 functions, while the SVP and the NVP trial functions are expansions containing only the L2 terms. Three different sets of L2 functions are used to examine the effect of changing the basis on the convergence characteristics of the three methods. We find that the rates of convergence for the Kohn, Schwinger, and Newton methods are strongly dependent on the nature of the potential and the basis set used. We also find that purely repulsive potentials are, in general, easier to converge than purely attractive potentials.

Notes: Resonance transition probabilities are obtained without the need to compute eigenvectors of the complex scaled Hamiltonian matrix. The new computational approach is a combination of the complex Lanczos recursion method, the complex coordinate method, and the extension of the Wyatt and Scott modified QL algorithm to complex matrices. To illustrate the method, the coupling strength between a resonance state and different scattering channels of HD from a flat Ag surface are calculated.

Notes: Extensive configuration interaction (CI) and multiconfiguration self-consistent-field (MCSCF) calculations on the A–X transitions of CH and CH+ are reported, which show how the transition dipole matrix element between two electronic states of a molecule can be determined accurately and systematically by a sequence of ab initio calculations. Two different computational approaches are investigated: one uses different orbitals for the two states involved and the other uses the same orbitals for both states. Both methods are shown to converge rapidly when the active orbital space is increased in such a way that all orbitals of similar importance are included at the same time. It is also shown that natural orbital occupation numbers from second-order CI calculations using a valence active orbital space provide an useful guide to the relative importance of natural orbitals; all orbitals of comparable occupation numbers must be included together.

Notes: The ionization potentials of the molecules PuF5, PuF4, and PuF3 have been measured using electron impact ionization. The values are, respectively, 14.5±0.1, 11.8±0.4, and 8.6±0.5 eV. The parent ion PuF+6 was not detected. From the ionization potentials and the fragment ion appearance potentials the bond dissociation energies of the molecules F5Pu–F, F4Pu–F, F3Pu–F, and F2Pu–F have been calculated to be 53±2, 108±12, 113±5, and 141±16 kcal/mol, respectively. Upon comparison with data available on the plutonium fluorides the bond energies of F3Pu–F and F2Pu–F are accurate to within the stated uncertainty. The values of the bond energies of F5Pu–F and F4Pu–F disagree with the reported enthalpy of formation of PuF6(s). Analysis of available data indicates that the enthalpy of formation of PuF6(s) is most likely −484±10 kcal/mol.

Notes: A new theory is introduced to model the lipid membrane structure and stability both above and below the gel to liquid phase transition temperature. Recently, we elaborated a self-consistent-field (SCF) theory, in which the full set of conformations was generated in a rotational isomeric state scheme and Boltzmann statistics was used to determine the statistical weight per conformation. In the present paper we also take into account that the anisotropic distribution of the molecules on the lattice induce a self-consistent anisotropic molecular field. This field, which is a function of the bond orientations, leads to an extra factor in the statistical weight of each conformation and is based on a generalization of Flory's and Di Marzio's analysis of systems with rigid rods. This elegant refinement follows from elementary statistics, is free of new adjustable parameters, and significantly improves details of the structure of the model membranes. To examine the properties of the SCAF (self-consistent anisotropic field) theory we use a model membrane built up by lecithin-like molecules composed of apolar and polar segments. The model has three nearest-neighbor interaction parameters of the Flory–Huggins type, namely for the interaction between apolar segments and water, that between polar segments and water, and that between polar and apolar segments. A fourth parameter is the dihedral trans/gauche energy difference. The theory predicts a first order gel to liquid phase transition for the model membranes. Depending on the membrane concentration, both an intercalated (in the dilute regime) and a nonintercalated (in the concentrated regime) gel phase are observed. Detailed information on the various membrane phases is obtained. Order parameter and segment density profiles are given.

Notes: The inelastic scattering of NO molecules from graphite surfaces is studied by classical trajectory methods. The experimental results from Frenkel et al. (1982), Segner et al. (1983), and Häger and Walther (1984) are analyzed. A model using a small isolated part of the graphite surface in interaction with the NO molecule gives results in good agreement with experiment. The parameter values in the model are fixed at the values previously found to reproduce the angular distributions well [Nyman and Pettersson (1987)]. For this system, the experimental results give a "rotational cooling'' such that the rotational temperature of the inelastically scattered molecules becomes smaller than the surface temperature. This effect is reproduced accurately by the calculations, giving a rotational temperature of 250 K, independent of the surface temperature above 300 K. The main factor controlling this inelastic rotational cooling is the low initial value of the normal component of the total angular momentum. A "rotational rainbow'' structure is found in the calculations in many cases, primarily at high surface temperatures. The final energy distributions are shown to be mainly statistical by application of a unimolecular decomposition picture, similar to the common RRK type model used for gas phase reactions.

Notes: We have included the proper quantum effects into the theory for thermal rate constants between any two sites of a multisite system by Voter and Doll [J. Chem. Phys. 82, 80 (1985)]. The thermal rate constants are expressed in terms of quantum flux correlation functions and we apply the formalism to surface self-diffusion. Using a potential energy function that previously has been used in order to describe hydrogen adsorbed on a copper surface, we determine the quantum corrections to the classical transition state theory, both those associated with bound degrees of freedom and those associated with tunneling and nonclassical reflexion. Our results agree quite well with previous calculations. To evaluate the temperature and time propagators necessary for the quantum dynamics we use a fast Fourier transform technique and we demonstrate that it is feasible to study the probability of correlated jumps in the case of surface diffusion. By solving for the three-dimensional and classical motion for the hydrogen atom it is shown that it is crucial to go beyond a one-dimensional treatment in order to study the possibility of correlated jumps. In this preliminary study we have neglected the motion of the substrate degrees of freedom, but the main aim with the presented approach is to be able to treat quantitatively the multidimensional case.

Notes: A theoretical approach to the problem of the marked irreproducibility of certain chemical reactions studied by Epstein's group at Brandeis University is presented. The model is based on the use of a Langevin equation with a multiplicative noise. Stirring and finite volume effects are incorporated in our approach. Analytical results are provided for the mean first passage time (MFPT), which is the appropriate quantity we associate here with the reaction time.

Notes: The low-density expansions of the concentration of monomers, dimers, trimers, and the mean cluster size are computed exactly, up through three-body cluster integrals, for a continuum–percolation model of spheres characterized by both exclusion-volume repulsions and short-range attractions. The compactness of trimers for our model is studied in the dilute limit. Interestingly, the [1,1] Padé approximant of the mean cluster size yields percolation thresholds which exhibit the proper qualitative dependence on the strengths of the repulsive and attractive interactions. The predicted thresholds agree qualitatively well with the trends reported in a Monte Carlo study by Bug et al.

Notes: A sub-Doppler infrared spectrum of (HCCH)2 has been obtained in the region of the acetylene C–H stretching fundamental using an optothermal molecular-beam color-center laser spectrometer. Microwave spectra were obtained for the ground vibrational state using a pulsed-nozzle Fourier transform microwave spectrometer. In the infrared spectrum, both a parallel and perpendicular band are observed with the parallel band being previously assigned to a T-shaped C2v complex by Prichard, Nandi, and Muenter and the perpendicular band to a C2h complex by Bryant, Eggers, and Watts. The parallel band exhibits three Ka=0 and three asymmetry-doubled Ka=1 series. The transitions show a clear intensity alternation with Kc with two of the Ka=0 series missing every other line. In addition, the perpendicular band has the same ground-state combination differences as the parallel band. To explain these apparent anomalies in the spectrum, we invoke a model consisting of a T-shaped complex with interconversion tunneling between four isoenergetic hydrogen-bonded minima. In this picture, the parallel and perpendicular bands arise from excitation of the acetylene units parallel and perpendicular to the hydrogen bond. The observation of rotation–inversion transitions in the microwave spectrum, in addition to the pure rotation transitions of Prichard, Nandi, and Muenter, verifies the model. The measured microwave splittings yield a tunneling frequency of 2.2 GHz which is consistent with a ∼33 cm−1 barrier separating the four minima.

Notes: We report the spectroscopic characterization of the X(10+) and A(30+) states of CdNe, CdAr, CdKr, and CdXe. The van der Waals molecules were created in a free jet supersonic expansion and studied by low and high resolution laser-induced fluorescence. CdAr was also studied by dispersed fluorescence. A method of analyzing rotationally structured vibrational bands of overlapping isotopic spectral contributions is discussed. Spectroscopic parameters are obtained from computer simulations of CdNe and CdAr spectra and from analysis of vibrational isotope splittings for CdKr and CdXe. CdNe: r‘e (X state)=4.26±0.05 A(ring), re(A state) =3.62±0.05 A(ring), D'e(A state) =77 cm−1; CdAr: r‘e(X) =4.33±0.04 A(ring), r'e(A) =3.45±0.03 A(ring), De(A) =325 cm−1; CdKr: D'e(A) =513 cm−1 and CdXe: De(A) =1086 cm−1.

Notes: The resonance enhanced multiphoton ionization spectra from 400 to 500 nm of the GeF and GeCl radicals are reported. In GeF bands involving transitions to the A 2Σ+, C' 2Π, D' 2Π, E 2Σ+, and G 2Δ states were observed. The bands involving transitions to the valence A 2Σ+ state were accounted for by 1+2 and 1+1+1 resonance enhanced multiphoton ionization (REMPI) mechanisms. Bands associated with the other Rydberg states were generated by a 2+1 REMPI mechanism. The spectrum of the GeCl radical arose from 2+1 REMPI excitation through the C' 2Πr (5pπ) and C 2Πr (4pπ) Rydberg states. The C' 2Πr state spin–orbit coupling constant was determined by direct measurement of C' 2Π1/2←←X 2Π1/2 transitions and "one photon forbidden'' C' 2Π3/2←←X 2Π1/2 transitions. A least-squares fit determined the spectroscopic constants of the C' 2Πr 5pπ Rydberg state of GeCl: Te=42 190±5 cm−1, Ae=81.0±3.5 cm−1, ωe=509.2±5.3 cm−1, and ωexe=3.3±1.1 cm−1.

Notes: Experimental results obtained for the dissociation reaction ICN*→[I⋅⋅⋅CN]

Notes: The microwave spectrum of the Ar–formamide van der Waals complex has been obtained using a pulsed-nozzle Fourier-transform microwave spectrometer. The rotational constants of the complex are: A=10 725.7524(48) MHz, B=1771.0738(22) MHz, and C=1548.9974 (16) MHz. The complex is shown to be nonplanar with an inertial defect of −6.21 u A(ring)2. The Ar atom is located at 3.62 A(ring) from the center of mass of the formamide unit at Ar–O, Ar–N, and Ar–C distances of 3.55, 3.79, and 3.93 A(ring), respectively. The shortest Ar–H distance is 3.25 A(ring) which is similar to that observed for Ar–vinyl cyanide (3.21 A(ring)). Stark effect and hyperfine analyses yield the following values for the electric dipole moment components and 14N quadrupole coupling constants for the complex: μa=0.922(1) D, μb=3.407(5) D, χaa=−1.164(7) MHz, χbb=1.906(5) MHz, and χcc=−0.742(6) MHz.

Notes: A model for outer-sphere bimolecular electron transfer reactions is presented which is based on a rate expression correct in the adiabatic, nonadiabatic, and intermediate dynamical regimes for an overdamped solvent coordinate. A complete discussion of the time-scale separations needed to define a bimolecular rate for these electron transfer reactions is given. Bimolecular reaction rates are predicted to display distinct solvent and electronic coupling dependencies, defined by a few experimentally variable parameters, which we now identify. Also, a significant intermediate dynamical regime is predicted to exist in which the rate is sensitive to both electronic coupling and nuclear relaxation. Evidence that both extreme dynamical regimes have been accessed experimentally is recalled, and strategies are presented for fully mapping out the dependence of rate on the electronic coupling and relaxation times.

Notes: A relatively comprehensive theoretical description of one-color and two-color photoionization of molecules by strong laser fields is developed. The molecular system is modeled by a number of discrete electronic configurations and a number of electronic ionization continua in a diabatic representation, allowing for intramolecular coupling of the discrete states and the continua. The vibrational degrees of freedom are included in compact operator notation without invoking the Born–Oppenheimer approximation. The relevant radiative dipole couplings are treated nonperturbatively on an equal footing with the intramolecular couplings. The important Coulomb threshold effects (accumulating Rydberg series at the electronic–vibrational ionization thresholds) are included via a threshold expansion of the complex level-shift operators representing intramolecular and radiative couplings. The weak-field, long-time ionization rate (golden-rule formula) is rederived from the general theory both in the one-color and the two-color case. In the two-color case, strong-field effects caused by either one of the two lasers are briefly discussed. We derive a simplified multichannel-quantum-defect model from the general two-color formalism which nicely reproduces a number of characteristic features of vibronic autoionization recently observed in several polyatomic molecules.

Notes: The endoergic substitution reactions Br+R–Cl→Cl+R–Br (R=o-, m-, and p-CH3C6H4, C6H5, C6F5; ΔH°(approximately-equal-to)15 kcal/mol) have been studied using the crossed molecular beams method in the collision energy (Ec ) range 20–35 kcal/mol. The CH3C6H4Br and C6F5Br products were found to be mostly forward scattered with respect to the incident Br beam indicating that the lifetimes of the Br–R–Cl collision complexes are short compared to their rotational periods. The product translational energy distributions and excitation functions for these reactions are well reproduced by statistical calculations that assume that only a few vibrational modes in the collision complexes participate in intramolecular energy redistribution prior to Cl elimination. Ring substituents are found to affect both the extent of energy redistribution in the complexes and the probability of Br addition. For example, no substitution product was observed with m-CH3C6H4Cl or C6H5Cl. The relative magnitudes of the cross sections are explained in terms of possible features of the potential energy surfaces along their reaction coordinates.

Notes: The various absolute partial cross sections for producing the excited neutral fragments, OH(A 2Σ+), H(n=2–4), and H2(a 3Σ+g), from photoexcitation of H2O in the 15–20.5 eV region are presented. It is found that a large fraction of the neutral products which amounts to 0.35 in the 15.7–17.7 eV region can be attributed to the undetected neutral products of (1) H(n=1)+OH(X 2Π), (2) H2(X 1Σ+g)+O(1D), and (3) H2(X 1Σ+g)+O(1S). All these products will possess high kinetic energy. The present results clearly show strong competitions among processes leading to photoionization and various photodissociation processes.

Notes: We present a modification of the time-independent adiabatic switching algorithm which is applicable to nonintegrable Hamilton systems. This algorithm transforms the Hamiltonian into a generalization of the Birkoff–Gustavson normal form. We discuss in detail the difficulties which occur when singular motions are encountered and explore the dynamical nature of the adiabatically switched tori using traditional methods of classical mechanics. A quantization scheme based upon this algorithm is proposed and numerical results, i.e., the semiclassical energies, are reported for a simple model of water.

Notes: Numerical methods are described for the calculation of time correlation functions and the stress relaxation moduli for the model developed in the preceding paper [J. Chem. Phys. 89, 3892 (1988)]. Illustrative results are presented for the storage and loss modulii, the shear compliance, and the translational diffusion constant, with parameters appropriate to polybutadiene.

Notes: The shear viscosity, pressure, and configurational (SPC) internal energy of liquid water at four densities along the 303.15 K isotherm are calculated using nonequilibrium molecular dynamics. The interaction between the water molecules is modeled by the SPC pair potential of Berendsen and co-workers. The simulation results are compared with experimental data at the same temperatures and densities. The shear viscosity is found to be approximately 50% lower than the experimental results but has a similar slope as a function of density. The results for the full SPC potential are compared to those of a model fluid whose pair potential is obtained by setting the electrostatic part of the SPC interaction to zero. The results for the shear viscosity are remarkably similar to those of SPC water, suggesting that the electrostatic interactions make a very small contribution to the viscosity for this fluid.

Notes: The motion of electrons in nonpolar dense media is studied in the framework of gas kinetic transport theory. The observed external electric field dependence of electron mobility is explained by means of the solution of Boltzmann-type transport equation. Numerical calculations of the mobility, mean energy, and energy distribution function of electrons in liquid Ar and in liquid CH4 are presented and a good agreement with experimental data is obtained. Also the cases of gaseous Ar are studied.

Notes: In this paper we further explore the applicability of a vibrational T2 process based on the extended-exponential modulation model [Rothschild, Perrot, and Guillaume, J. Chem. Phys. 87, 7293 (1987)] to Raman correlation data of concentrated aqueous solutions of LiSCN and KSCN [Kato¯, Mol. Phys. 48, 1119 (1983); Kato¯ and Takenaka, Mol. Phys. 46, 257 (1982)]. In general, the values of dispersion parameter α in the modulation function exp[−(t/τ)α], obtained from the fit of the theory to the isotropic correlation data of the CN oscillator, predict the prevalence of interrelated, collective dynamic processes in the medium that are the cause of the instantaneous oscillator transition frequency shifts (motional narrowing). In particular we predict, from the observed concentration dependence of α, strong short-time (fraction to several ps) cation–water–anion interactions that, in the more concentrated LiSCN–H2O systems at 303 K, are above a site percolation threshold with a value of α∼0.3 (close to that found in glasses). The expectation value of t, 〈t〉=τΓ(1+1/α), becomes critical near a concentration of 5 mol/l and shows a pronounced Vogel–Fulcher-type temperature dependence (T0=250 K) in the 10 mol/l LiSCN–H2O system over a range 0.45–76 ps. However, since α approaches its limiting value=1 at the highest temperature reported (353 K), the large-cluster cation–water–anion distributions in LiSCN–H2O must be rather tenuous. In contrast, the characteristics of α and of 〈t〉 for the KSCN–H2O systems agree with the relatively weak cation–water forces; the (inverse) concentration dependence of α is linear, its temperature dependence is flat, and the Vogel–Fulcher-type temperature behavior of 〈t〉 for the 10 mol/l solution stretches merely from 0.7 to 1.4 ps.

Notes: We present Monte Carlo simulation results for an isolated polyelectrolyte chain where the charged groups interact via a screened Coulomb interaction of Debye–Hückel form. The chain consists of hard spheres on a backbone with fixed bond angles and either free rotation or a torsional potential that produces hindered rotation. Parameters have been chosen to roughly mimic polyacrylic acid (PAA). We give results for the mean square end-to-end distance 〈r2〉, mean square radius of gyration 〈s2〉, and finite-chain persistence length for chains of up to 320 beads and salt concentrations ranging from 0.1 to 0.0001 mol dm−3. Both 〈r2〉 and 〈s2〉 show power law behavior for chains larger than 80 beads at 0.1 mol dm−3 —the scaling exponents are intermediate between rigid rod and self-avoiding walk values. No scaling behavior can be observed at lower concentrations because of the limitations on the chain lengths that can be simulated. At 0.1 and possibly at 0.01 mol dm−3 the infinite-chain persistence length can be obtained by extrapolation of the distribution of angles of the chain bonds relative to the central bond. These values are in rough agreement with the worm-like chain calculations of Skolnick, Fixman, and Odijk. Our calculations would be expected to give larger values for the persistence length than theories which include ionic degrees of freedom, such as the Poisson–Boltzmann calculations of Le Bret and Fixman. At 0.1 mol dm−3, however, their results are larger than ours, a result we find puzzling.

Notes: The photodissociation of jet-cooled NOCl following excitation in the lowest, structured absorption band is reported. We find that NO(2∏3/2) is produced preferentially, and that the rotational distributions depend strongly on the number of bending quanta in the upper electronic state. In addition, the NO vibrational excitation depends on parent NO stretch excitation. NO(v‘=0) is predominantly produced when v'1 =0 is excited, while NO(v‘=1) is mainly formed following excitation of v1 =1.

Notes: Expressions are derived for the vibrational second hyperpolarizability γv of SF6, in the static limit and for the nonlinear-optical processes of the dc Kerr effect, electric-field-induced second harmonic generation (ESHG), and third harmonic generation (THG). These expressions give γv in terms of vibrational transition dipole, polarizability, and first hyperpolarizability matrix elements (μ,α,β). The matrix elements needed in the numerical evaluation of γv are obtained from the literature and from the Raman scattering measurements reported here. In the static limit it is found that γv(approximate)20 γe, where γe is the electronic contribution to γ. At optical frequencies γv is still comparable to γe in magnitude, and it shows significant frequency dependence.

Notes: The scanning method is a computer simulation technique for polymer chains, which is especially suitable to handle chains with finite interactions and chains that are subject to various geometrical constraints. A chain is constructed step by step with the help of transition probabilities, obtained by scanning the possible continuations of the chain in future steps (called future chains). We discuss in detail the efficiency of the method and for that we study certain autocorrelation functions for three lattice models: self-avoiding walks (SAWs) on a square lattice, a random walk model for polymer adsorption and trails with attractive interactions. We demonstrate that for SAWs the scanning method is significantly more efficient than the related method of Rosenbluth and Rosenbluth. We also develop and test a new procedure in which the transition probabilities are obtained, not by exact enumeration of all the future chains, but from a relatively small sample of future chains, generated by another scanning procedure. This "double scanning'' process is expected to be useful for complex macromolecules such as polypeptides.

Notes: The surface structure of single crystal metals in reflected in the angular distributions of neutral atoms desorbed during keV Ar ion bombardment. Results are presented here of a study of desorption from a stepped surface, Rh{331}. Recently, classical dynamics simulations of the desorption process were shown to give excellent agreement with ejection data from Rh{111}, and the same model potential has been used to simulate desorption from Rh{331}. The agreement with experiment remains excellent despite the different coordination of the surface atoms on Rh{111} and Rh{331}. The azimuthal ejection distributions are strongly affected by the steps.

Notes: A significant increase in the total integral cross section in (Na)2−Ne collisions with the vibrational level v'' of the sodium molecule in the electronic ground state, at a collision energy of the order of 1 mev is reported. (AIP)

Notes: The observation of an anomaly in the optical luminescence excitation spectrum of oxygen in the region of the oxygen K edge is reported. Dispersed luminescence spectra were obtained for x-ray excitation at the pi and sigma resonances, at the anomaly, and in the continuum. These spectra indicate enhanced production of O2+2 ions at both the sigma resonance and at the anomaly. The anomaly thus is attributed to a shake-up or shake-off state associated with an antibonding sigma molecular orbital of oxygen. This work also demonstrates that optical luminesence spectra provide state-specific information about the products of core hole excitation and relaxation.

Notes: We have measured the near ultraviolet absorption spectrum of 2,3-dideuterobutadiene to provide a complete set of experimental Bu←X vibrational intervals and bandwidths for all symmetrically deuterated butadienes. These vibrational intervals and bandwidth ratios are compared with the ground state vibrational frequencies and frequency ratios of the molecules. The prominent vibrational frequency interval observed in transitions to the Bu state of butadiene is demonstrated to arise predominantly from a kinetic coupling of the C=C stretching and CH wagging vibrations. The experimental bandwidth ratios are shown to correlate with single quanta of the ground state au CH2 twist frequency interval ratios. From the latter, a plausable decay path for the Bu excited state of butadiene is deduced. The implications of these conclusions on prior and present attempts to determine the butadiene Bu equilibrium geometry and to understand polyene spectroscopy, photochemistry, and photophysics are discussed.

Notes: Fluorescence spectra from the Rydberg A 2Σ+ (v=0) level of NO trapped in Ne, Ar, Kr, and Xe matrices have been obtained for two different sites. The main site is characterized by broad A(0,v‘) bands (FWHM(approximately-equal-to)80 meV), blue gas-to-matrix shifts of 70 to 300 meV, and by absorption–emission Stokes shifts of 300 meV in Xe to 800 meV in Ne matrices. The line shapes and Stokes shifts are treated within a configuration coordinate model by a moment analysis. A linear and quadratic coupling is invoked with relaxation energies ranging for the excited state from 160 meV in Xe to 540 meV in Ne and for the ground state from 150 meV in Xe to 280 meV in Ne and an increase of the cage radius of 3% in Xe to 15% in Ne. The "red'' site fluorescence shows emission bands with matrix shifts of −140 meV in Xe to 80 meV in Ne and absorption–emission Stokes shifts of 210 meV in Xe to 830 meV in Ne. Red site "hot'' A (v=1) fluorescence is also observed. Red sites are attributed to local disorder around the molecule in Ar, Kr, and Xe matrices and to hexagonal close packed (hcp) pockets within the face centered cubic (fcc) lattice in the case of Ne matrices.