Volume 68, Issue 4, 15 February 1978
Index of content:

Computer simulation of the nucleation and thermodynamics of microclusters
View Description Hide DescriptionGas liquid equilibrium in finite systems is studied by computer simulation using molecular dynamics and Monte Carlo techniques. In overexpanded liquids, cavitation is observed. At some point in the expansion the system undergoes a transition from a liquid with cavities to a droplet in equilibrium with vapor. This transition is observed in the p V diagram of the system. A method is devised for counting the number of atoms in a cluster and thereby determining the clusterdistribution function. The interface of a microcluster is compared to that of a planar sheet. The two are found to be very similar in density profile. In the course of this study homogeneous nucleation in a supersaturated gas is observed for the first time in a molecular dynamics study. Simple theories of nucleation in a finite system are considered. The free energy of formation of a droplet is found to have a maximum—a barrier to nucleation—and a minimum—stable equilibrium between a droplet and liquid. When gas imperfection is included, the barrier increases, and the stable cluster is destabilized.

Determination of the ^{14}N quadrupole tensor of phenazine in its lowest excited triplet state by ODNQR
View Description Hide DescriptionThe first complete determination of the ^{14}N nuclear quadrupoletensor in the photoexcited triplet state of an organic molecule was made feasible by an alternative to ODENDOR. Instead of using a simultaneous desaturation method (cw‐ODENDOR), a procedure involving sequential population transfer between specific electron–nuclear spin levels was adopted. This new method is demonstrated for phenazine‐d _{8} in a Shpolskii matrix and yields ‖e ^{2} q Q/h‖= (4.36±0.02) MHz; η=+0.21±0.03.

Collective vibrational modes of adsorbed CO
View Description Hide DescriptionThe interpretation, in terms of dipole interactions, of the dependence on coverage of the stretching frequency of CO adsorbed of metal sufaces is re‐examined. We take into account substrate image effects and dielectric screening of the adsorbed layer. We also provide a theory of the statistical fluctuations in random, partially filled layers. The dipole interactions provide a frequency shift of about 10 cm^{−1}, which is only about one third of that usually observed.

Ionization of xenon atoms in selected high Rydberg states by collision with CH_{3}I, C_{7}F_{14}, C_{6}F_{6}, and CH_{3}Br
View Description Hide DescriptionAbsolute rate constants for the ionization of highly excited xenon atoms in collisions with CH_{3}I, C_{7}F_{14}, C_{6}F_{6}, and CH_{3}Br have been measured and the major charged reaction products have been identified for all but CH_{3}Br. Theoreticalmodels view such collisions as being dominated by the interaction between the target molecules and the ’’essentially free’’ Rydbergelectrons. It is thus predicted that the negative ions produced during collisional ionization are identical to those resulting from free electron attachment, and that the rate constants for collisional ionization are the same as those for free electron attachment. The present data are reviewed in light of these predictions.

Electron mobilities in gaseous, critical, and liquid xenon: Density, electric field, and temperature effects: Quasilocalization
View Description Hide DescriptionThe mobility μ of electrons in xenon varies with the density n : μαn ^{ x }. For thermal electrons in the gas, x=−1.0 when n<4×10^{20} molecule/cm^{3}, independent of temperature. This corresponds to pressures ≲20 atm at 300 K and is characteristic of simple scattering by a polarization field. At higher gas densities x becomes more negative, then passes through a minimum, and μ has a positive temperature coefficient at constant n. At these densities thermal electrons form quasilocalized states in, or suffer enhanced scattering by, microscopic regions of relatively high density. The localization effect maximizes in the critical fluid, at n=5×10^{21} molecule/cm^{3} and T=290 K, and appears to be negligible again at n?6.8×10^{21}. In the liquid phase x?+9 up to n=1.2×10^{22} molecule/cm^{3}, then changes to x?−8 at higher densities. The positive value of x is due to the formation of a conduction band, and the negative value to scattering by the repulsive cores of the closely spaced atoms. In the gas and the low density liquid (n<1.2×10^{22}), μ at low electric field strengths E is independent of E; at higher E, μ increases with E then passes through a maximum and decreases. The maximum in μ reflects the Ramsauer–Townsend minimum in the scattering cross section σ_{ε} of the atoms as a function of electron energy ε. Quasilocalization is inhibited by the application of moderate electric fields, and this enhances the observed dependence of μ on E at 4×10^{20}<n<7×10^{21}. The electron drift velocity v _{ d } tends to become independent of E at high E in the gas, liquid, and solid phases. The ’’plateau’’ drift velocity was 1.0 km/s at n?2×10^{21}, 1.2 at 5×10^{21}, 3.0 in the normal liquid at 1.4×10^{22}, and 5.5 km/s in the solid at 1.6×10^{22} molecule/cm^{3} (Miller, Howe, and Spear). The velocity plateau in all phases is attributed to the steep rise in σ_{ε} on the high energy side of the Ramsauer–Townsend minimum. The maxima in μ vs E curves for heavy ions are discussed.

The relative permittivity of supercooled water
View Description Hide DescriptionApproximate relative permittivity data for water down to −35°C have been derived from measurements of the Maxwell–Wagner interfacial polarization which occurs at the surface of emulsified water droplets. Although the emulsifying agent strongly influences this polarization in a complex way, a novel method of data analysis enables the permittivity and conductivity to be obtained in a simple and direct manner. The derived data are in essential agreement with literature data near 0°C. To within an estimated experimental uncertainty of ±2% the present data conform to a critical exponent expression with the same critical temperature (228 K) found for many other properties of supercooled water, though the divergence is very weak.

A theoretical study of phase transitions in bidimensional monolayers
View Description Hide DescriptionWe propose a theoretical study of phase transitions in long hydrocarbon chain monolayers, lying on aqueous substrate. We use a model taken from that applied by Di Marzio to liquid crystals: each molecule can occupy several aligned sites on the holder (1 to 5 for us). We also take into account the attractive interactions between the hydrocarbon chains in the framework of a mean field approximation: it gives a first order phase transition in a diagram showing surface pressure variations in terms of the molecular density, noted (π, ρ) diagram. We also introduce an order–disorder transition and we show that it is either of the first order, or of the second order, according to the physical conditions. The (π, ρ) isotherms are in good qualitative agreement with experimental results.

Thermodynamic properties of a quasi‐ionic alloy from electromotive force measurements: The Li–Pb system
View Description Hide DescriptionElectromotive force measurements of the activity of lithium in lithium–lead liquid alloys are reported in the temperature range T=770–932 K. The partial molar excess Gibbs free energies are negative. Variations with composition exhibit an inflection point around an atomic fraction of lead equal to 0.2 which corresponds to the ratio of the valences of Li and Pb. The inflection point, as well as structural data, may be interpreted by assuming an at least partly ionic character of the alloy. A set of partial structure factors defined by Faber and Ziman are calculated as a function of the composition at the long‐wavelength limit. Their variations support the hypothesis of a local short‐range order around the above‐mentioned composition (0.2). Finally, a modeling approach based on an electron transfer from Li to Pb is used to explain some of the characteristics of this class of alloys.

Stochastic theory of phase transition in multiple steady state chemical system: Nucleation
View Description Hide DescriptionWe study on a simple reaction–diffusion system the role of stochastic fluctuation in first order type transition between two stable steady states. We focus especially on the large fluctuations in a small spatial region that give rise to nucleation. The theory of nucleation developed is based on a birth‐and‐death type master equation. The result is different from the Langevin approach, and the physical basis of the difference is discussed.

Measurement of the rate coefficients for the bimolecular and termolecular ion–molecule reactions of He_{2} ^{+} with selected atomic and molecular species
View Description Hide DescriptionThis work reports the measurement of bimolecular and termolecular charge transferreactions of He_{2} ^{+} into nonassociative product channels. In this study ion destruction frequencies have been experimentally determined from the selectively excited fluorescence of N_{2} ^{+} in high pressure afterglows of mixed gases excited by intense electron beam discharges. Data have been obtained as functions of helium pressure over the range from 400 to 1500 torr and as functions of the partial pressure of reactant from 50 to 200 mtorr. From these data pressure‐dependent rate coefficients have been extracted and subsequently resolved into contributions from bimolecular and termolecular components for reactions of He_{2} ^{+} with Kr, H_{2}, O_{2}, NO, HBr, HCl, H_{2}O, N_{2}O, NO_{2}, C_{3}H_{8}, NH_{3}, and CCl_{2}F_{2}, selected because of their widely varying values of polarizability and dipole moment. The bimolecular components have been found to agree with the NOAA flowing afterglow results, where available, and to compare favorably with theoretical values obtained from Langevin, ADO, and locked dipole approximations. The sensitivity of the method has been sufficient to detect termolecular components as small as 2×10^{−30} cm^{6} sec^{−1} and values were found to range widely from 9×10^{−30} cm^{6} sec^{−1} for H_{2} to 140×10^{−30} cm^{6} sec^{−1} for HBr. A termolecular analog to the Langevin limit has been constructed which explains these values in terms of the rates at which third‐body encounters change glancing collisions into inwardly spiraling trajectories. The data reported are in general agreement with this model and indicate that the reaction probabilities are near unity in both bimolecular and termolecular channels. The size of these termolecular rates suggests the general importance of three‐body ion–molecule reactions in higher pressure plasmas such as those found in e‐beam lasers.

Bounds for the long‐range behavior of electronic wavefunctions
View Description Hide DescriptionUpper bounds to expectation values of many‐electron operators for atomic wavefunctions are given. These estimates, which incorporate the ionization potentials of the atom, lead to detailed information about the asymptotic behavior of subcontinuum wavefunctions if more than one electron–nucleus distance tends to infinity. They compare favorably with results given in the literature.

Lattice parameters and thermal expansion of solid CD_{4}
View Description Hide DescriptionPrecise measurements of the lattice parameters of solid CD_{4} at saturated vapor pressure are reported for temperatures from 4.4 to 60 K. The x‐ray results show two first order phase transitions near 26.9 and 22.0 K. For the two cubic phases, the lattice parametera has been measured to an accuracy of 15 ppm, while for the low temperature tetragonal phase, the parameters c and a have been obtained to roughly 150 ppm. The thermal expansion for all three phases has been calculated from the data. Several interesting properties of the phase transitions are described.

Electron affinity of the HC_{2} radical
View Description Hide DescriptionLarge scale a b i n i t i o self‐consistent field and configuration–interaction calculations were carried out to evaluate the electron affinity of the HC_{2} radical. The adiabatic electron affinity is calculated to be 2.14, the vertical 2.09, and the vertical detachment energy 2.17 eV. Corrections for the zero‐point vibrational energy amount to less than 0.01 eV. From these values, the heat of formation of HC_{2}, and the C–H bonddissociation energy of C_{2}H_{2} were estimated to be 121±3 and 119±3 kcal mole^{−1}, respectively.

Further empirical testing of the suitability of a nonrandom integration method for classical trajectory calculations
View Description Hide DescriptionA nonrandom method of approximating multidimensional integrals is compared to the traditional Monte Carlo method in the determination of average energy transfer values from classical trajectories. Two atom–diatomic molecule collision systems are studied. Estimates of error show that, for a given number of trajectories, the nonrandom method tends to be more accurate than the Monte Carlo method.

Pseudospectral dipole oscillator strength distributions and some related two body interaction coefficients for H, He, Li, N, O, H_{2}, N_{2}, O_{2}, NO, N_{2}O, H_{2}O, NH_{3}, and CH_{4}
View Description Hide DescriptionPseudospectral dipole oscillator strength distributions (DOSDs) are presented for ground state Li, N, O, H_{2}, N_{2}, O_{2}, NO, N_{2}O, H_{2}O, NH_{3}, and CH_{4}. The pseudo DOSDs are constructed by requiring the pseudodipole (excitation energy–oscillator strength) pairs to reproduce known values of the dipole oscillator strength sums S (k) evaluated from original DOSDs which are either completely continuous or contain a substantial continuum contribution. The pseudo DOSDs, which are a discrete representation of the original DOSDs, can be used to evaluate a variety of molecular properties in almost a trivial way, without sacrificing accuracy, by using discrete pseudospectral representations of the properties. The pseudostate DOSDs, together with a b i n i t i o pseudostates for the ground state H and He atoms, are used to evaluate orientation averaged dipole–dipole dispersion energy constants, C_{6}(a,b), and the corresponding relativistic coefficients, W _{4}(a,b), for all pairs of atoms and molecules taken form H, He, Li, N, O, H_{2}, N_{2}, O_{2}, NO, N_{2}O, H_{2}O, NH_{3}, and CH_{4}. The results are used to assess the accuracy of semiempirical formulae for C _{6}(a,a) and W _{4}(a,a) and of combination rules for C _{6}(a,b) and W _{4}(a,b). The method of using pseudo DOSDs for the evaluation of molecular properties, relative to using the original DOSDs, becomes more important as the number of integrals over DOSDs in the definition of a property increases.

Spin–spin dipolar and exchange interactions in crystalline bisgalvinoxyl biradical
View Description Hide DescriptionA static magnetic susceptibility study has been carried out on a stable crystalline bisgalvinoxyl biradical in the temperature range 55–330 °K. The study has established that a negative intramolecular spin–spin exchange interaction (J) exists between two unpaired electrons in the biradical, giving rise to a ground singlet state and a thermally accessible triplet state. The singlet–triplet separation (2‖J‖/k) derived from the observed temperature dependence of the magnetic susceptibility is 490±10 °K (341±7 cm^{−1}). The existence of the triplet state of the bisgalvinoxyl has been further confirmed by the observation of the ESR zero‐field splitting of its powder sample, by reducing the triplet entities at the low temperatures. The ESR spectrum at 77 °K can be clearly analyzed as the characteristic spectrum of a nonaxially symmetrical triplet, with the zero‐field splitting parameters ‖D‖=76.0±0.3G, ‖E‖=14.5±0.3G, and g _{ x x }, g _{ y y }, and g _{ z z }, equal to 2.0042±0.0002, 2.0059±0.0002, and 2.0032±0.0002, respectively. When the temperature is increased, the zero‐field components are broadened, then collapsed into a single, sharper line, because of temperature‐dependent intermolecular exchange interactions. The D‐ and E‐parameters were calculated for the assumed molecular structures, using McLachlan’s spin densities of galvinoxyl, which is a monoradical half of the bisgalvinoxyl.

The molecular Zeeman effect in vinyl formate using a conventional Stark modulated spectrometer and a Fourier transform spectrometer
View Description Hide DescriptionThe rotational Zeeman effect in vinyl formate was observed. The rotational molecular g values are g _{ a a }=−0.1759±0.0010, g _{ b b }=−0.0342±0.0006, and g _{ c c }=−0.0059±0.0006. The magnetic susceptibilityanisotropies in units of 10^{−6} erg G^{−2} mole^{−1} are 2χ_{ a a }−χ_{ b b }−χ_{ c c }=9.5±1.3 and 2χ_{ b b }−χ_{ a a }−χ_{ c c }=20.8±1.1. The molecular quadrupole moments in units of 10^{−26} esu cm^{2} are Q _{ a a }=−0.1±1.4, Q _{ b b }=+1.2±1.4, and Q _{ c c }=−1.1±2.2. The relative performance of Stark modulated and Fourier transform spectrometers is examined.

Crossed beam studies of O^{−}+D_{2}→OD^{−}+D
View Description Hide DescriptionUsing the crossed‐beam machine EVA we have measured the product angular and energy distributions of the reaction O^{−}+D_{2}→OD^{−}+D in the relative energy range of 1.2–4.7 eV (5.7–23.1 eV LAB). Below 2.5 eV the product distribution is centered about the center of mass, indicating a long‐lived complex. Above 2.5 eV the distribution slowly moves forward. Most of the available energy goes into internal energy of the products.

The potential energy curves of the X ^{1}Σ^{+} and A ^{1}Σ^{+} states of CsH
View Description Hide DescriptionPotential energy curves for the X ^{1}Σ^{+} and A ^{1}Σ^{+} states of CsH are constructed using Rydberg–Klein–Rees (RKR) methods modified (a) to include the effect of nonzero rotational quantum number and (b) to simultaneously use data from more than one isotopic variant, respectively. In each case these potentials are considered to be the most complete and reliable available. Intensity calculations are also presented based on Franck–Condon factors obtained for these new potentials; comparison with experiment suggests a strong increase of A–X transition moment with increasing internuclear distance, as previously indicated for analogous transitions in LiH and NaH.

Dynamics of vibrational motion of chemisorbed atoms or molecules; the infrared line shape
View Description Hide DescriptionWe present a theory which relates the line shape of infrared absorbtion by chemisorbed atoms or diatomics to the dynamics of lattice vibration. We start from the relationship between the line shape and the thermodynamicGreen’s function and use Dyson’s equation to compute the latter with the aid of many‐body perturbation theory. The derivation includes local modes and surface states as well as anharmonic interactions. Illustrative calculations show that a variety of line shapes are possible and we identify their dynamical origin.