ALBERT

Notes: The adsorption of NO on Si(111)7×7 at 90 K and the subsequent reactions induced by thermal heating and photon irradiation have been investigated in detail, using high-resolution electron energy loss spectroscopy and mass spectrometry. It is found that, in addition to molecular and dissociative adsorption of NO, N2O is produced from NO on Si(111)7×7 at 90 K. The product N2O physisorbs on the surface and, at 110 K, partially desorbs and partially dissociates into N2 and O. Molecular adsorption of NO exists in two independent, presumable bridge and atop configurations. There is no observable conversion between the two forms of molecularly adsorbed NO. At 147 K, the bridge NO is thermally activated, which leads to simultaneous NO desorption and, to a much larger extent, N2O synthesis. Dissociation of bridge NO also occurs at about 147 K, at least at low NO exposures. The atop NO is not involved in the thermal reactions at 147 K, but dissociates between 200 and 300 K. Photon irradiation in the UV, visible, and IR induces desorption and dissociation of NO on Si(111)7×7 at 90 K via nonthermal mechanisms. Furthermore, evidence for photosynthesis of N2O on the surface is observed. It is suggested that the N2O synthesis during thermal heating and photon irradiation proceeds via reaction between an adsorbed NO molecule and a hot N atom produced from NO dissociation.

Notes: We apply a recently developed combined molecular dynamics–local Langevin equation method to the simulation of the scattering of Ar by the (100) face of a face-centered cubic solid. The kinetic energies of the Ar are chosen to be low compared to the typical energies used in sputtering. We find that even at low energies, a significant amount of surface damage is inflicted by the Ar, leading to ejection of metal atoms into the gas phase, the formation of dislocations, and the production of isolated atoms trapped on the surface. We study both the probability that such events occur and individual trajectories which display the dynamic processes through which sputtering takes place or defects are created.

Notes: We studied the dimerization equilibrium of 2-methyl-2-nitrosopropane in argon at 25.0 °C from about 500 to 1200 kg m−3. It is clarified that in the simplest solvent, argon, there exists a definite difference in the density dependence of the equilibrium constant between the high and medium density regions. In the high density region, the equilibrium constant shows a sizable increase with the solvent density as in the liquid phase. Near the critical density, however, the equilibrium constant shows little density dependence.

Notes: The influence of periodic orbits on quantum eigenfunctions and the dynamical evolution of wave packets is studied on a model 2D potential. The phase space structure is portrayed by constructing the diagram of characteristics of the most important families of periodic orbits. It is found that the topologies of eigenfunctions can be assigned to certain periodic orbits. Families bifurcating from the principal families, as well as irregular families which are not associated with the principal families, mark the eigenfunctions. The evolution of wave packets and the spectra which are produced from them show that quantum mechanically there is no communication among different resonance regions even at energies where chaos predominates in classical mechanics. This quantum regularity and localization is explained by testing the conjecture that the area of turnstiles (regions in the surfaces of section through which trajectories pass to get in or out the resonance zone) is less than Planck's constant. It is demonstrated that the systematic study of phase space structure through the families of periodic orbits in association with the solution of the time dependent Schrödinger equation for wave packets localized initially on such orbits provide the numerical techniques for studying highly excited species.

Notes: We report the results of a numerical study of how traveling wave branches originate and terminate in a prototypical excitable chemical system, viz. the Kernevez model. These results furnish a qualitative understanding of recent experimental observations on sustained spiral waves in a continuously fed unstirred reactor [W. Y. Tam et al., J. Chem. Phys. 88, 3395 (1988)], in particular of the hysteretic nature of the transition between the wave state and the homogeneous steady state.

Notes: A study of barrier crossing dynamics is presented for a simple two-degree-of-freedom system coupled to a Nosé–Hoover heat bath. The characteristics of the deterministic heat bath are found to give rise to a variety of rate processes. Examination of the microscopic dynamics indicates how bath fluctuations drive the reactive dynamics and gives insight into how different deterministic heat baths can be constructed to model specific bath fluctuation effects.

Notes: When electron correlation effects are small, the set of energy levels available to both the localized and the delocalized individual electrons (the band structure) is the starting place for determining the macroscopic electronic properties of a substance. Calculating the band structure in any disordered medium, however, requires facing the problem that there will always be a distribution of geometries in the material—at least the local parts of which must be accounted for in order to get any reasonable results. In a liquid this requirement means that the liquid structure plays an important role. We show in this paper that the band structure in a liquid is completely and rigorously determined by the equilibrium behavior of an "effective'' liquid with artificial internal degrees of freedom. This mapping implies that standard liquid theory methods (which automatically build in the correct liquid structure) can be used to find the electronic energy levels. As illustration, we use the mean-spherical approximation (MSA) to derive a simple expression for the density of states that is accurate at all but the lowest densities. We further show that this particular MSA theory is identical to an apparently different theory derived recently by Logan and Winn—which makes both theories identical to the so–called EMA theory of Roth. An even more general correspondence exists between our exact formalism and the exact formalism of Logan and Winn, though any given approximation might be more natural in one approach then in another.

Notes: An optical Kerr effect, which is the transient electric birefringence of isotropic molecular fluids when the molecules align themselves with external field, is investigated. Calculations have been carried out on the rotational diffusion of rod-like molecules in the presence of an arbitrary field strength. We will provide an exact recurrence relation with nonlinear effects and this recurrence relation is solved numerically to see how the nonlinearity affects the answer for various field strengths.

Notes: The theory of the correction for the difference in the fluctuations among the different ensembles is applied to derive the time correlation function formula required to compute the distinct diffusion coefficients in the barycentric reference frame by molecular dynamics. In the process it is found that the dynamical variables in the time correlation function of any mass transport coefficient of any reference frame are the same in the canonical as in the molecular dynamics ensemble. The result agrees with the accepted formulas for computing the kinetic part of the mutual diffusion coefficient in a binary mixture of nonelectrolytes and for computing the electrical conductivity in molten salts and electrolyte solutions. The static velocity correlations in the molecular dynamics ensemble are analyzed without recourse to the other ensembles. In sharp contrast with the ensembles which are not constrained with respect to the total momentum, the laboratory-frame velocities of two different particles at the same time are found to be correlated. The result obtained is equivalent to the initial value of the canonical time correlation function of the barycentric distinct diffusion coefficients derived in the preceding paper.

Notes: Starting with the Green–Kubo formulas for the phenomenological coefficients of irreversible thermodynamics, the molecular expression for the distinct diffusion coefficient Dd[R],αβ in any internal reference frame R is derived. It follows that Dd[R],αβ is given by the time integral of a time correlation function (tcf) Λd[R],αβ(t) whose dynamical variables are the velocities of two different particles of species α and β, each relative to a microscopic reference velocity wˆR . These relative dynamic variables have the welcome feature of being orthogonal to the total momentum of the system so that Λd[R],αβ(t) vanishes at long times. Hence the divergence of the barycentric distinct diffusion coefficients, suggested by their usual interpretation in terms of a time correlation function that lacks an explicit dependence on the reference frame, is resolved. The influence of the reference frame on the behavior of the barycentric distinct tcf 's is analyzed. In great measure this influence is exposed by considering initial values of the tcf 's; in turn the initial values are decisive in establishing the sign of the distinct diffusion coefficients in ideal or nearly ideal mixtures. Finally, by applying new linear relations connecting the self and distinct tcf 's, a complete description of some of the barycentric distinct tcf 's may be given in the limiting cases of a pure fluid and a tracer in a host liquid.

Notes: The revised capillarity approximation (RCPA) of McGraw and Reiss [J. Stat. Phys. 20, 385 (1979)] is reexamined using established liquid drop models to evaluate the effects of physical cluster excluded volume on the thermodynamic barrier to nucleation near a critical temperature Tc. A significant heightening of the nucleation barrier is predicted, and the anomalous behavior observed experimentally near Tc is examined within the framework of classical nucleation theory using the RCPA. Definitive methods for experimentally separating this pure nucleation effect from those dependent on slowing down of the post-nucleation completion times for phase separation are described.

Notes: The photoreduction of CH3Cl was used to detect the transmission of electrons through layers of H2O and Xe on Ni(111) under ultrahigh vacuum (UHV) conditions. At a laser wavelength of 248 nm with H2O spacers, the electron intensity exhibited a nearly exponential decay to zero signal with the 1/e point at about 2 monolayers. At 193 nm with H2O spacers, the signal decayed with the 1/e point at about 4 monolayers to a constant value due to the direct photofragmentation of the CH3 Cl possible at this wavelength. At 248 nm with Xe spacers, the signal decayed by a factor of 4 in the first 2 monolayers, after which the signal was nearly constant and nonzero for at least 64 monolayers, indicating a long electron mean free path. CH3Br is shown to behave similarly with H2 O spacers at 248 nm.

Notes: The adiabatic dynamics of an electron in a polar solvent has been studied by means of a molecular dynamics technique for the coupled electron–solvent system. The method hinges on a separation of time scales for the solvent motion and certain classical dynamical degrees of freedom which represent the ground-state wave function of the electron. For the solvated electron these variables are taken to be the positions and widths of Gaussians that comprise a distributed basis set. The method is applied to an electron in liquid ammonia, and the conditions for the validity of the classical separation of time scales are investigated. Time correlation functions describing the electron diffusion coefficient, and the fluctuation and dissipation of electron energy, are compared with analogous quantities for a classical solute (a chloride ion) in the same solvent. The electron diffusion coefficient in ammonia is calculated to be about twice that of a chloride ion. Experimental mobility data on the electron–ammonia system suggest that the real diffusion coefficient is a factor of 3 larger than calculated for the model used here. Possible reasons for this discrepancy are noted.

Notes: The kinetics of aggregation of polystyrene latices occuring in the presence of an electrolyte or poly(4-vinylpyridine) in aqueous media are examined. The aggregate size distribution was recorded at different times by counting the flocs containing a given number of elementary colloids. The moment of order zero ( the total number of clusters per unit volume of the suspension), as well as the second moment of the size distribution are computed. In the presence of an excess of electrolyte or an optimum amount of polymer, a dynamic scaling function of the form cg (t)=t−2 F[g/t] and a narrow time invariant polydispersity is found. Under slow flocculation conditions, a large time dependent polydispersity sets in. Large aggregates then react preferentially with equally large ones. This behavior is discussed in term of a sticking probability , using the formalism of the reaction limited cluster–cluster aggregation.

Notes: The highly correlated CCSDT-2 model of electron correlation is applied to computation of the equilibrium geometry and harmonic vibrational frequencies of O3. Comparison of the present results with those of a recent study [J. F. Stanton, W. N. Lipscomb, D. H. Magers, and R. J. Bartlett, J. Chem. Phys. 90, 1077 (1989)] reveal the importance of certain simultaneous three-particle correlation effects (T3) which enter perturbation theory in fifth and higher orders, particularly for asymmetric structures of ozone and, consequently, the B2 stretching frequency. The CCSDT-2 value for ω3 is 1182 cm−1, in reasonable agreement with the experimental value of 1089 cm−1, and considerably better than values obtained with other methods which make more approximations in treating T3. To treat this difficult problem, it appears to be necessary to include terms involving T3 which are nonlinear in the cluster amplitudes into the CC equations.

Notes: The intensities of vibrational lattice modes of α and β s-triazine and pyrazine are calculated using four models of increasing complexity for the description of the molecular multipole electrostatic moments and polarizabilities. For s-triazine crystals it is found that high-order electrostatic interactions should be taken into account in order to reproduce the experimental infrared absorption intensities. For pyrazine crystal it is found that high-order molecular polarizabilities give a non-negligible contribution, and that it cannot be reparametrized in terms of a molecular dipolar polarizability.

Notes: We employ laser double resonance techniques to measure the rates of gas-phase collisional deactivation of vibrationally excited HF by NO molecules, in order to compare the relaxation efficiency of this free radical species with CO and other closed-shell molecules whose vibrational dynamics are well known. Although the near-resonant energy gaps for HF collisional energy transfer are less favorable for NO than for CO, we find that NO is as much as an order of magnitude more efficient than CO in relaxation HF vibrations. Since the NO and CO collision partners have rather similar dipole moments (0.153 vs 0.112 D), rotational constants (1.17 vs 1.93 cm−1), and molecular weights, the disparity in vibrational relaxation efficiency may come from chemical factors, particularly the open- vs closed-shell electronic character, associated with long-range interactions. Ab initio calculations and natural bond orbital (NBO) analysis of the structure and energetics of NO:HF and CO:HF complexes indicate that the NO monomer is better able to form effective n→σ* donor–acceptor H bonds to HF over a wide range of nonlinear "acceptance angles.'' Compared to CO, NO presents a significantly attractive potential to HF over an appreciably wider range of collision orientations, leading to structures in which HF stretching couples to other internal modes of the transient complex and vibrational excitation is efficiently quenched. Our results strongly suggest the important role that chemical factor can play in the dynamics of fast vibrational relaxation processes.

Notes: Nonrelativistic and relativistic Hartree–Fock (HF) and configuration interaction (CI) calculations have been performed in order to analyze the relativistic and correlation effects in various diatomic gold compounds. It is found that relativistic effects reverse the trend in most molecular properties down the group (11). The consequences for gold chemistry are described. Relativistic bond stabilizations or destabilizations are dependent on the electronegativity of the ligand, showing the largest bond destabilization for AuF (86 kJ/mol at the CI level) and the largest stabilization for AuLi (−174 kJ/mol). Relativistic bond contractions lie between 1.09 (AuH+) and 0.16 A(ring) (AuF). Relativistic effects of various other properties are discussed. A number of as yet unmeasured spectroscopic properties, such as bondlengths (re), dissociation energies (De), force constants (ke), and dipole moments (μe), are predicted.

Notes: Geometry optimizations and band structure calculations are reported on two polymeric vinylene derivatives of pyrene: poly(1,6-pyrenylene vinylene) and poly(2,7-pyrenylene vinylene). The results are indicative of a dramatic sensitivity of the electronic structure on the mode of connection of the two vinylene units to the pyrene rings, in agreement with the chemist's intuition. An analysis of the atomic π orbital contributions to the one-electron wave functions shows that the HOMO–LUMO characteristics of both compounds are markedly different. The 1-6 mode of connection leads to an electronic structure similar to that of polyparaphenylene with a bandgap of 2.9 eV. The 2-7 mode of connection results in a significantly lower bandgap, around 1.7 eV, due to the stabilization of quinoid resonance structures.

Notes: A theoretical study of the relationship between interatomic distances and the spectral positions of valence- and K-shell σ* photoionization resonances is reported for a selected series of molecules. Three-dimensional graphical representations of the occupied and virtual-valence σ-symmetry orbitals of these compounds reveal their striking similarity to the wave functions of a particle in a cylindrical well, substantiating qualitative notions long employed in free-electron molecular orbital (FEMO) approximations. Accordingly, the molecular potential along the symmetry axis in these compounds is modeled after a finite square well, with a depth approximately equal to the energy of the lowest σ-symmetry valence molecular orbital and a width determined from analogies to FEMO theory. Calculated minimal-basis-set molecular-orbital energies for both occupied and virtual states are seen to correlate accurately with the simple square-well energy level formula (π2 /2)(n2/l 2 ) when measured in Hartree atomic units from the bottom of the well. The calculated σ* orbital energies are furthermore in excellent agreement with experimentally and theoretically determined valence-shell photoionization resonance positions, the latter consequently also satisfying the square-well correlation formula. A similar situation obtains for experimentally and theoretically determined K-shell resonance positions, although energy shifts from minimal-basis values are evident in these cases. These circumstances are clarified quantitatively on basis of Feshbach–Fano considerations, in which minimal-basis-set virtual-valence σ* orbitals play the roles of zeroth-order states subject to modification by interactions with nonresonant background continua. Concluding remarks contrast and compare molecular-orbital and square-well approaches to photoionization resonances with those based on multiple-scattering and barrier models. The present results appear to clarify the origins of recently reported empirical correlations of bond lengths with resonance positions, and help to determine their range of applicability.

Notes: The viscosity of a suspension of spherical Brownian particles is determined by Stokesian dynamics as a function of the Péclet number. Several new aspects concerning the theoretical derivation of the direct contribution of the Brownian motion to the bulk stress are given, along with the results obtained from a simulation of a monolayer. The simulations reproduce the experimental behavior generally observed in dense suspensions, and an explanation of this behavior is given by observing the evolution of the different contributions to the viscosity with shear rate. The shear thinning at low Péclet numbers is due to the disappearance of the direct Brownian contribution to the viscosity; the deformation of the equilibrium microstructure is, however, small. By contrast, at very high Péclet numbers the suspension shear thickens due to the formation of large clusters.

Notes: The nature of long-range many-body interactions in metallic fluids is examined with emphasis on their possible role in the unique features of these systems observed near the liquid–vapor critical point. A reexamination of recent theoretical results demonstrating the existence of van der Waals forces between "pseudoatoms'' (ion cores and associated screening electrons) reveals a direct correspondence with dispersion forces in insulating systems. In the limit of high conduction electron number densities ρ, pseudoatoms have an effective frequency-dependent polarizability α(ω)=α(0)ω2p/(ω2p−ω2), where α(0)=z/4πρ, with z the ion valence, and ωp is the classical electron gas plasma frequency (4πρe2/m)1/2. It is the dynamic nature of the interactions (arising from fluctuations of the pseudoatoms) that permits such a long-range interaction to exist. The dimensionless parameter α(0)ρ which in insulating fluids characterizes the relative importance of triplet (Axilrod–Teller) to pair dispersion interactions is thus system independent and significantly larger than in nonmetallic fluids. The nature of this dynamic polarizability is further examined in the context of a transport theory for a classical plasma based on the Boltzmann equation. The statistical mechanics of fluctuating pseudoatoms at finite temperature is studied both for the metallic fluid and in the Wigner crystal. These various approaches suggest that the pseudoatom interaction may be viewed as a potential mediated by the exchange of plasmons, just as conventional van der Waals forces arise from the exchange of virtual excitations of atomic levels. A description of the critical point in terms of pseudoatom interactions appears to explain qualitatively the extreme liquid–vapor asymmetry of the coexistence curves of cesium and rubidium as arising from the magnitude of three-body interactions. Additionally, it suggests that the thermal energy at the critical point scales with the plasmon energy, consistent with experiment.

Notes: A simple matrix method is given that generates the probability distribution for the coordinates of the end of a polymer chain (modeled as a random walk on a regular lattice with a finite range of intrachain correlation). From this information the exact end-to-end distance distribution can be constructed. The method can be applied to specific sequence copolymers and to the distance distribution for arbitrary points within a polymer chain.

Notes: We report the results from a theoretical investigation of the structures and reactivity of various isomers of Ni clusters in the size range from 4 to 13 atoms. The geometries of the clusters were optimized using binding energy values calculated by the corrected effective medium (CEM) theory. Two different potential energy surfaces were used to describe the interaction between D2 and the Ni clusters. The first used the form and parameters that were determined in the study of H2 dissociative chemisorption on Ni surfaces, while both used atomic positions appropriate to the clusters, the second used the same form but determined the parameters by comparison to CEM values of the H/Ni13 interaction. Using these PES, we investigated the dissociation dynamics of D2 on NiN (N=4–13) clusters by classical trajectory techniques. We found that: (1) for clusters of size less than Ni9 , the rate constants varied strongly with cluster size; and (2) for all size clusters, the rate constants were very sensitive to different isomers. This isomeric variation of the rate constant is discussed in terms of various structural features in different isomers, a number of which do not have any analogy in the dissociative chemisorption on low Miller index surfaces.

Notes: We present analytical justification for our previously described exchange pseudopotential. We show how the fermi quantum partition function can be constructed from the Boltzmann (distinguishable particle) wave functions if the states that correspond to like-spin electrons occupying the same quantum state are excluded. A class of weighting functions that satisfy this constraint approximately is discussed. Our previous pseudopotential falls under this class. Essentially, our pseudopotential forces the unwanted states to have high energy and, hence, to make negligible contribution to the partition function. Exchange potentials of the form discussed in this article should be useful for studying systems where the (allowed) correlated Boltzmann wave functions have negligible amplitude for like-spin fermion–fermion distances less than the diameter of the individual particle wave packets. For example, in the case of two spin-up (or spin-down) fermions, if one fermion is located at r, then ||Ψ(r,q)||2 is negligible if q(approximately-equal-to)r. This should be the case for systems where a tight binding model is appropriate or for systems with strong interparticle repulsions.

Notes: The interaction of water on clean and caesium dosed Cu{110} has been investigated as a model for the simulation of electrochemical double layers in ultrahigh vacuum. We present new data from high resolution electron energy loss spectroscopy (HREELS) and electron stimulated desorption ion angular distributions which provide a detailed picture of the growth and structure of water on clean Cu{110}. Water adsorbs molecularly on this surface and forms hydrogen bonded clusters at all coverages. The presence of caesium has a strong influence on the adsorption of water. For aitch-thetaCs=0.06 water is molecularly adsorbed with an increase in the heat of adsorption. The strongly bound water is associated with an increase in the work function and with striking changes in the HREELS spectra: the libration mode at 750 cm−1, which dominates the clean surface water spectrum, is absent in the presence of caesium. These effects are attributed to a reorientation of water in the vicinity of the caesium adatom.

Notes: Microwave and infrared spectra of HF–HCl and HCl–HF have been obtained using a molecular-beam electric-resonance optothermal spectrometer, which operates by quadrupole-field focusing of polar molecules onto a bolometer detector. The HF–HCl microwave measurements extend to Ka=1, the previous Ka=0 results of Janda, Steed, Novick, and Klemperer, allowing the determination of the Ka dependence and asymmetry of the Cl quadrupole coupling constant. For the metastable HCl–HF isomer no previous spectroscopic measurements have been reported. Here, microwave spectra are observed for the Ka=0 and 1 states and interpreted in terms of an L-shaped hydrogen-bonded structure for the complex, with a 3.235 A(ring) center-of-mass separation between the HF and HCl subunits. The DJ distortion constant indicates that the harmonic stretching force constant for HCl–HF is ∼35% larger than that of HF–HCl. Infrared spectra of the Ka=0–0 and 1–0 subbands of the H–F stretching band for HF–HCl and of the Ka=0–0 subband of the H–F stretch for HCl–HF are also reported. The vibrational predissociation linewidths depend on vibration, Ka state, isotopic species, and isomer excited.

Notes: Using a reduced equation of motion for the density matrix which accounts for spontaneous emission and superradiance, we analyze the fluorescence and transient grating (TG) decays from a dilute, optically thin distribution of molecular aggregates. We find that the fluorescence is a limited form of superradiance, where cooperativity is restricted to the number N of two level systems which make up the aggregate. The dependence of the linear aggregate decay rate on the amount of inhomogeneous broadening, which is randomly distributed according to Gaussian line shape of width (1/e)2σ, is calculated. It is shown that σ must be comparable in magnitude to the nearest neighbor dipole–dipole coupling V and not the superradiant decay rate, in order to quench the superradiance.

Notes: We have used ab initio methods to characterize the Ne–HF van der Waals complex. The interaction energy was determined using size consistent, correlated CEPA wave functions expanded in a Gaussian basis chosen to represent both intraatomic effects and the low order multipole moments and polarizabilities of Ne and HF. The calculated well depths are −65 cm−1 for linear Ne–HF and −39 cm−1 for linear Ne–FH, with an intervening saddle point at −27 cm−1. The induction contribution to the energy is significantly greater for Ne–HF than for Ne–FH, but dispersion remains the dominant attraction over the region of interest. Converged variational and close-coupling calculations using the ab initio potential surface reveal three bound levels of the Ne–HF stretch mode, and several metastable levels correlating asymptotically with rotationally excited HF( j=1). Though nearly degenerate, the lifetimes of the two metastable Π (body frame Λ=±1) bending levels differ markedly because of different rotational coupling strengths to the Σ (body frame Λ=0) bending state, which undergoes rapid rotational predissociation. From the calculated line positions, widths, and intensities we have synthesized far infrared and infrared spectra of Ne–HF and Ne–DF.

Notes: The infrared spectra of the ν3, ν4, and ν6 bands of formaldehyde in the region from 890 cm−1 to 1580 cm−1 have been obtained at high resolution using tunable diode laser (TDL) and Fourier transform-infrared (FT-IR) spectroscopy. The transition frequencies have been analyzed using a Hamiltonian including terms through sextic in centrifugal distortion and including five interstate vibration–rotation coupling terms. Excited state pure rotational transitions are also included in the data, and their frequencies are reproduced well. Individual measured line intensities are used to determine dipole derivatives and band strengths using the fully coupled, asymmetric top eigenvectors.

Notes: Visible laser excitation (460–725 nm) of dilute rare gas/I2 (2000:1) matrices resulted in emission from the I2A 3Π(1u) state. Reanalysis of the A→X spectra provided revised molecular constants for matrix isolated I2. A state lifetimes of 70±20, 80±20, and 110±30 μs were observed in Ar, Kr, and Xe hosts, respectively. Excitation spectra for the A state closely followed the I2 continuum absorption spectrum, indicating that transfer from the B 3Π(0+u) and 1Π(1u) states was effective in populating I2(A). At dilution ratios of 600:1 or lower the I 2P1/2–2P3/2 transition was observed in conjunction with the A–X bands. Excitation studies showed that isolated I atoms, trapped during the deposition process, were excited by energy transfer from nearby I@B|2 molecules. A vibronic progression, similar to the A–X bands, but shifted to longer wavelengths, was noted in concentrated Rg/I2(300:1) matrices. This system, which was emitted with a lifetime of about 10 ms, most probably originated from perturbed I2 A' 3Π(2u). Intermolecular energy transfer was observed in matrices that contained I2 codeposited with O2. Electronic excitation of I2 resulted in a long-lived emission from O2 a 1Δg. Matrices containing high concentrations of iodine also exhibited O2(a)→I(2P1/2) transfer.

Notes: Polarized resonance Raman measurements of highly oriented films of cis polyacetylene have been carried out. The intensities of the three Raman fundamentals in the four different scattering configurations have been measured and corrected for the strong anisotropy of the optical properties. The experimental data are indicative of an almost perfect alignment of the cis polyenic chains along the stretching direction. Preliminary measurements of the depolarization ratios carried out on isotropic samples in the resonance region appear to be inconsistent with the view that the visible band is dominated by a single transition polarized along the chain axis.

Notes: Laser-induced fluorescence excitation spectra have been recorded for many vibronic transitions of the A(2Π)–X(2Π) and B(2Σ+)–X(2Π) systems of NCS under supersonic free jet expansion conditions. New assignments have been made for many of these bands and several of the assignments from previous work have been revised. Vibronic energies have been determined for levels involving excitation of all three vibrational modes in the ground electronic state and both excited electronic states. A detailed rotational and vibrational analysis has been carried out for levels involving excitations of the two stretching modes and the many rotational and potential function constants have been determined for the A and X states. The variation of the spin–orbit constant with vibrational level has also been investigated. Results of this paper provide the ground work for a detailed analysis of the Renner–Teller effect in the bending vibrational mode of this radical for both the X2(Π) and A(2Π) states to appear in a forthcoming paper. They also provide a spectroscopic database for future dynamical and kinetics studies of processes involving this radical.

Notes: A general phase modulation algorithm has been developed for the stored waveform inverse Fourier transform (SWIFT) excitation method used in Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR or FTMS). The algorithm, based on the time shifting theorem and the uncertainty principle, shows that the quadratic phase modulation is the theoretically optimal method for square magnitude spectral profiles. For more complicated magnitude spectral profiles, the corresponding phase functions can be generated through the algorithm by using a nonlinear grid on the frequency domain. The degree of dynamic range reduction can be estimated from a simple equation.

Notes: An accurate and computationally tractable theoretical procedure for the calculation of the nonresonant, electronic components of the third-order molecular polarizabilities, γ(0;0,0,0), γ(−3ω;ω,ω,ω), and γ(−2ω;ω,ω,0), can be constructed. This procedure partitions γ into a σ-electron component (γσ) and a π-electron component (γπ). The γσ term is evaluated using the bond-additivity approximation; the γπ term is calculated using the semiempirical INDO all-valence-electron molecular orbital method combined with full single- and double-excitation configuration interaction (SDCI) of singlet π-electron configurations, and Orr and Ward's sum-over-states expression for γ. The INDO-SDCI method is also used to calculate one- and two-photon spectroscopic properties of the 1ππ* states salient to γπ for the molecules of interest. It is shown that single-excitation CI alone is not sufficient for the calculation of γπ for linear polyenes and benzene. Calculations of the effect of chain length and conformation on the values of γ for ethylene, cis and trans linear polyenes, and benzene indicate that γ is strongly influenced by conjugation chain length. A simple relationship can be established between the calculated value of γπ(0;0,0,0) for the trans linear polyenes investigated and that for ethylene, the molecule with the solitary π-electron C–C bond: γπ(0;0,0,0)(approximately-equal-to)γπ(0;0,0,0)ethylene NC–C3, where NC–C=1,3,5,7, and NC–C is the total number of C=C and C–C bonds in the given polyene, i.e., the length of the π-bonding network. As γ increases with chain length, so does the ratio γπ/γ. Virtual electronic transitions involving excited π-electron states with extensive charge separation and double excited configurational character are important contributors to γπ for the linear polyenes and benzene. An approximation of γπ(0;0,0,0) for the linear polyenes can be written in terms of the linear π-electron polarizabilities for the ground state and 1 1Bu π-electron excited state. Although thisapproximation is strictly applicable to the centrosymmetric linear polyenes, it does suggest a very interesting criterion for the selection of organic molecules with large third-order polarizabilities. Namely, the change in polarizability between the ground state and a strongly one-photon absorbing excited state is an important factor to consider when selecting candidate molecules.

Notes: The photoion–photoelectron coincidence (PIPECO) spectra for (N2)+2 in the wavelength range 650–866 A(ring) have been measured at different nozzle stagnation pressures. The formation of stable (N2)+2 from fragmentation of excited (N2)+n cluster ions initially produced by photoionization of (N2)n, n≥3, is efficient. For nozzle expansion conditions which minimize the production of (N2)n, n≥3, the intensities for the N+2(A˜,B˜)⋅N2 PIPECO bands are found to be negligibly small compared to that of the N+2(X˜)⋅N2 PIPECO band, indicating that the electronically excited N+2(A˜,B˜)⋅N2 dimer ions are dissociative in temporal ranges 〈42 μs. Assuming that the radiative lifetimes for N+2(A˜,B˜) and N+2(A˜,B˜)⋅N2 are identical, we estimate that the dissociative lifetimes for N+2(A˜)⋅N2 and N+2(B˜)⋅N2 are (approximately-less-than)10 μs and (approximately-less-than)60 ns, respectively. The ionization energy for (N2)2 is determined to be 14.50±0.08 eV (855±5 A(ring)), suggesting that N+2(X˜)⋅N2 is bound by 1.09±0.08 eV. The PIPECO data for (N2)+2 presented here and those for (CO)+2 reported previously support the perturbed monomer ion model for the photoionization of a van der Waals cluster. Namely, the formation of N+2⋅(N2)n−1 by photoionization of (N2)n, n≥2, can be viewed as a photoionization process of N2 perturbed by the presence of other N2 molecules in the clusters. We suggest that the rapid dissociation of electronically and vibrationally excited dimer ions is a general mechanism for the suppression of autoionization features in the photoionization efficiency spectrum for an ionized van der Waals dimer.

Notes: Energy transfer of vibrationally highly excited CF3I molecules (E(approximate)18 000 cm−1) in collisions with argon, propane, and octane was studied using hot UV absorption spectroscopy of CF3I. the preparation of the excited CF3I was achieved by IR multiphoton absorption which, under the conditions applied, produces a narrow initial energy distribution of CF3I near to the dissociation energy. The average energies 〈ΔE〉 transferred per collision were found to be proportional to E for the bath gases propane and octane; they showed a stronger increase with energy at low excitation energies in the bath gas argon. The energy dependence of 〈ΔE〉, therefore, is not only governed by the properties of the excited molecule (e.g., its density of states) but also by features of the collision partner (e.g., the magnitude of 〈ΔE〉). At energies near to the dissociation energy for all colliders, 〈ΔE〉 was found to approach values similar to those obtained from single UV photon excitation experiments with highly excited triatomic and large polyatomic molecules.

Notes: A classical trajectory study of energy flow between two metal-molecule bonds separated by a heavy metal atom in a collisionally perturbed complex O 3/4 C⋅⋅⋅M⋅⋅⋅C 3/4 O is presented. The left M⋅⋅⋅CO bond is excited by collision and subsequent flow of excitation is followed by classical mechanics. In the heavy metal system, the behavior of energy flow blockage by the metal atom prevails at certain total energies of the complex. The blockage causes a long time delay for energy flow between two M⋅⋅⋅CO bonds and bond dissociation. Such energy flow blockage is found to be absent in the light metal system, where the amount of energy transfer to the complex is much larger. The energy flow patterns associated with the blockage are discussed in detail. Specific systems considered are the collisional excitation of OC⋅⋅⋅Pt⋅⋅⋅CO and OC⋅⋅⋅Ni⋅⋅⋅CO by Ar.

Notes: The flowing afterglow/Langmuir probe (FALP) technique has been extended to enable the neutral products of electron–ion dissociative recombination in thermalized afterglows to be identified by spectroscopic methods. Absolute number densities of H atoms in the afterglow have been determined using vacuum ultraviolet (VUV) absorption at the Lα wavelength. By exploiting the reaction H+NO2→OH+NO, all of the H atoms can be incorporated into OH molecules and thus observation of the intensity of laser induced fluorescence (LIF) If, obtained by exciting the (1,0) band of OH(A 2Σ←X 2Π), allows a calibration to be obtained of If against the known number density of OH X 2Π(ν‘=0) in the afterglow. Following this procedure, a recombining HCO+2 /electron afterglow was probed for production of ground state OH X 2Π(ν‘=0) using LIF and it was established that OH(ν‘=0) resulted from 17% of the recombining ground state HCO+2 ions. It was also established that a further 17% of the recombinations resulted in OH(ν‘〉0), i.e., that, in total, (34±6)% of the HCO+2 ions recombine to produce OH X 2Π radicals, either directly or via the electronically excited A 2Σ state. Details of the calibration procedure for H and OH number densities, of the ion chemistry involved in the production of the HCO+2 afterglow plasmas and of the checks carried out to establish that the fluorescence observed was from OH produced in the recombination reaction are presented. During these experiments, the rate coefficient at 300 K for the H+NO2 reaction was determined to be 1.3×10−10 cm3 s−1 from observations of the H-atom decay as a function of NO2 number density in the afterglow (in good agreement with previous determinations). Also the rate coefficient for the quenching reaction of OH(ν‘〉0) with NO to produce OH(ν‘=0) was determined to be 6×10−11 cm3 s−1.

Notes: A numerical study of dichotomous noise-induced transitions is presented for an example of a two-dimensional excitable system exhibiting bistability between a limit cycle and a fixed point: the simple Oregonator model. The decay of the average population number in the fixed-point region is investigated for various noise correlation times and for two different initial system preparations. The mean first passage time taken to leave the fixed-point region is determined and is compared with analytical results obtained from a simple stochastic model.

Notes: The high-resolution infrared spectrum of the K=1–0 subband of the H–F stretching vibrational band of the hydrogen-bonded HF–DF complex has been recorded using a molecular-beam electric resonance optothermal color-center-laser spectrometer. The spectrum exhibits minor perturbations and vibrational predissociation linewidths of 23±2 MHz full width at half-maximum for comparison to the 11±1 MHz widths found for the corresponding mode of the homonuclear HF–HF dimer.

Notes: In this work we study the direct, incoherent energy transfer from excited donors to acceptors, which are attached to chainlike polymers. We consider both multipolar and exchange-type interactions between the donors and the acceptors. The polymers are modeled through random walks (Gaussian or self-avoiding, depending on the solvent). With the use of the end-to-end distribution function of the walks, we calculate the time dependence of the ensemble averaged decay of the excitation of the donor. For multipolar and exchange-type interactions, we find Kohlrausch–Williams–Watts (KWW) stretched exponential and exponential–logarithmic decay patterns, respectively. The decay forms reflect the quality (good or poor) of the solvent directly and they may be used to test transitions in the polymer conformation.

Notes: The steady-state catalytic oxidation process of carbon monoxide on platinum metal surfaces is studied using two irreversible kinetic computer simulation models: (a) An extended version of the model introduced by Ziff, Gulari, and Barshad (ZGB) with the effects of CO desorption and diffusion as well as finite reaction probability taken into account. The different physical processes, diffusion and desorption are studied independently and their effect on the equilibrium window, i.e., the regime where steady CO2 formation occurs is determined. (b) An interaction model where adatom–adatom nearest-neighbor (nn) interactions are taken explicitly into account through Boltzmann terms J1, J2, and J3 which are the energies of the CO–CO, O–O, and CO–O interactions, respectively. The phase diagrams in the temperature–CO-partial pressure (T,pCO−) plane are determined for different values of the nn interactions. The behavior of the system is dependent on the sign of J1(=J2 in the simulations) as well as the sign of the difference J1−J3. There is thus a clear analogy with a two-component equilibrium lattice gas with nn interactions.

Notes: A kinematic study of the velocities and angular distribution of I− ions detected following Rydberg electron transfer in K(nd)/CF3I collisions is reported. The data show that post-attachment interactions between the product ions are important at intermediate n (9(approximately-less-than)n(approximately-less-than)20) even though the I− ions are formed with appreciable kinetic energies.

Notes: From the far infrared spectrum of 2-methylpropanal (isobutyraldehyde), (CH3)2CHCHO, in the gaseous state, the asymmetric torsion for the gauche conformer was observed as a series of Q branches at 75.0, 80.9, 85.8, and 90.3 cm−1 with similar transitions observed for (CD3)2CDCHO. Also from an investigation of the Raman spectrum of the gas a transition at ∼110 cm−1 has been tentatively assigned as the "double jump'' of the asymmetric torsion for the trans conformer. The microwave spectrum of the excited vibrational states for the light molecule has been investigated in the region from 12.5 to 40 GHz and, from relative intensity measurements, the frequencies of the asymmetric torsions were determined to be 72±7 and 50±20 cm−1 for the gauche and trans rotamers, respectively. Additionally, "tunneling'' splitting of 2.0±0.4 and 60.0±0.3 MHz has been observed for the second and third excited states, respectively, of the asymmetric torsion of the gauche conformer. From these data the asymmetric potential function has been calculated and the following potential constants have been evaluated: V1=−131±35, V2=−225±21, V3=555±10, V4=−70±5, V5=95±5, and V6=−140±10 cm−1 where the 1←0 transition of the gauche has been assigned at 75.0 cm−1. This potential function leads to an enthalpy difference of 250±66 cm−1 (715±189 cal/mol) which is in good agreement with the value of 248±35 cm−1 determined from variable temperature studies of the Raman spectrum of the gas. Analysis of the transitions observed between 250 and 150 cm−1 in the far infrared spectrum of the gas indicated that the two methyl rotors are coupled for the gauche conformer and the barriers to internal rotation of the methyl tops have been calculated to be 987 cm−1 (2.82 kcal/mol) and 1132 cm−1 (3.24 kcal/mol).

Notes: Cyanoacetylene underwent polymerization reaction in a solid phase at pressures above 1.5 GPa. The Raman study of the reaction product showed that the polymer had a conjugated linear backbone with CN pendant groups. The Raman spectra for this substituted polyacetylene demonstrated a resonance behavior similar to that reported for trans-polyacetylene. The optical gap associated with the π–π* transition in the conjugated system was smaller than that of trans-polyacetylene, probably due to the resonance interactions between the CN triple bonds and the conjugated double bonds.

Notes: The state-to-state integral cross sections for the inelastic scattering of CH(X 2Π) with He were measured in a newly constructed crossed molecular beam machine. Use of laser-induced fluorescence in an unconventional flux mode of detection provided single fine-structure state specific detection of the products. Two types of measurements were performed to further our understanding of the collision dynamics of open shell systems: (1) the product state distribution at a fixed and well-defined collision energy and (2) the dependence on collision energy of product state-resolved cross sections. A qualitative understanding of the collision dynamics can be obtained by properly factoring out features dependent on the fine-structure states, i.e., effects involving individual Λ-doublet states and features dependent on the rotational level alone, i.e., effects remaining after summing over all four fine-structure states associated with a given rotational quantum number. As for the fine-structure effects, a preferential population of product Λ-doublet states with reflection symmetry Π(A‘) was observed. The physical origin of this observed electronic orbital alignment can be attributed to a quantum interference phenomenon, as detailed in the accompanying paper. At the rotational level, the dominance of rotational rainbow scattering is unambiguously identified from both the existence of dynamical thresholds and a strong correlation between rotational level distributions at fixed translational energy and level specific excitation functions. These effects combined with other experimental observations lead us to visualize the CH+He scattering dynamics in a novel fashion. The collision can be regarded as a series of approximately independent sequential events each mediated by different regions of the interaction potential during the course of the whole encounter.

Notes: We have studied the excitation of N2(B 3Πg, v=1–12) in the interaction between N2(A 3Σ+u) and N2(X 1Σ+g, v≥5). The N2(A) and N2(B) are observed spectroscopically between 220 to 400 nm and 560 to 900 nm, respectively, while the N2(X,v) number densities are determined by metastable-helium Penning ionization. The experiments are performed in a discharge-flow reactor with separate discharge sources of N2(A), N2(X,v) and He*(23S). The excitation rate coefficient is (3±1.5)×10−11 cm3 molecule−1 s−1. Observations of N2(A) decay indicate that the N2(A) is removed by N2(X,v) with an apparent rate coefficient of about 3.5×10−12 cm3 molecule−1 s−1. The discrepancy between the excitation and removal rate coefficients probably results from N2(A) regeneration via cascade from the excited N2(B). The appearance of vibrationally excited N2(A) when N2(X,v) is added to a flow of N2(A, v=0) demonstrates this regeneration process. The reaction appears to be a transfer of electronic energy from the N2(A) to the N2(X,v) rather than an excitation of the N2(A) to N2(B) resulting from the input of energy from the N2(X,v).

Notes: A recently proposed procedure for computing the S matrix for collinear molecular collisions is extended from the single-mode case to the multimode case. In this procedure the semiclassical Hamiltonian describing the molecular collision is integrated in a finite-dimensional, faithful nonunitary representation of the dynamical group. When the integration is completed, the resulting group operation is mapped into the infinite-dimensional unitary representation to describe the S matrix which acts on the molecular Hilbert space and describes collisional excitation of the internal vibrational modes. This procedure is implemented to study the two-mode collision process N2+O2 and the two three-mode collision processes N2+CO2 and NO+CO2. The results compare favorably with other treatments of these collinear collision processes.

Notes: The shapes of the real-phonon and pseudophonon sideband holes (PSBH) which occur in hole-burned spectra in amorphous solids are both wavelength and burn time dependent. The theory previously proposed to simulate hole shapes in the short burn time limit is extended to examine the PSBH shapes at arbitrary times and at different burn wavelengths. The simulated spectra are compared with experimental data for tetraphenyl porphin in polystyrene. It is also shown how the simulated spectra may be used to deconvolve the antihole spectrum from the hole spectrum.

Notes: The gas phase infrared spectra of the hydrated hydronium cluster ions H3O+⋅(H2O)n(n=1, 2, 3) have been observed from 3550 to 3800 cm−1. The new spectroscopic method developed for this study is a two color laser scheme consisting of a tunable cw infrared laser with 0.5 cm−1 resolution used to excite the O–H stretching vibrations and a cw CO2 laser that dissociates the vibrationally excited cluster ion through a multiphoton process. The apparatus is a tandem mass spectrometer with a radio frequency ion trap that utilizes the following scheme: the cluster ion to be studied is first mass selected; spectroscopic interrogation then occurs in the radio frequency ion trap; finally, a fragment ion is selected and detected using ion counting techniques. The vibrational spectra obtained in this manner are compared with that taken previously using a weakly bound H2 "messenger.'' A spectrum of H7 O+3 taken using a neon messenger is also presented. Ab initio structure and frequency predictions by Remington and Schaefer are compared with the experimental results.

Notes: The charge-transfer mechanism in 1,4 cis-polyisoprene is studied using solid state 13C NMR spectroscopy as a function of doping level of iodine. The CP/MAS NMR data show that the concentration of double bonds decreases as a consequence of doping. The overall results are consistent with previous observations and indicate that cation radicals –C+||–C⋅||– are formed upon doping and charge transfer. The double bonds are reformed as the doped samples are compensated with ammonia indicating reversibility of the doping. The measurements also indicate that the chain mobility is restricted in the doped regions compared to the undoped regions.

Notes: Using multipulse techniques, principal values of the proton nuclear magnetic resonance (NMR) chemical shift tensors were measured at room temperature for the monoclinic form of oxalic acid, potassium hydrogen oxalate, and potassium hydroxide. The isotropic and perpendicular shifts of the first two compounds, which are of medium hydrogen bond strength, were found to fit the linear correlation with O⋅⋅⋅O distance, RO⋅O, established by Rohlfing, Allen, and Ditchfield for medium and strongly O–H⋅⋅⋅O hydrogen bonded solids. Although KOH is very weakly hydrogen bonded, the shifts were also found to conform to the correlation, at least as well as does the other data, thus extending this to weakly hydrogen bonded solids. An empirical correlation of the isotropic and perpendicular shifts with exp(−RO⋅O/ρ), where RO⋅O is in A(ring) and ρ is 0.94 A(ring), is given here which has better agreement with the data and has an interpretation in terms of a simple ionic model of an O–H⋅⋅⋅O bond.

Notes: The application of the algorithm of Zeiger and McEwen to the analysis of noisy photon correlation data is investigated. For the particular case where the data are sampled at equidistant time intervals a complete solution is given allowing reliable reconstruction of the spectrum of exponential decay rates without any a priori knowledge. A particular attractive feature of the method is that the singular value analysis of the Hankel matrix of autocorrelation functions offers a practical criterion for the decomposition of the data into a signal and a noise part. Some tests of the method are illustrated with experiments on monodisperse latices, gold sols, and binary mixtures of monodisperse latices. In the latter case comparable and even better results are obtained in significantly shorter computing time when compared to an analysis with Contin and the maximum entropy method. Since the present method does not require any a priori parameter setting, it is also complementary to these methods.

Notes: The dissociation reaction of HgI2 is examined experimentally using femtosecond transition-state spectroscopy (FTS). The reaction involves symmetric and antisymmetric coordinates and the transition-state is well-defined: IHgI*→[IHgI]

Notes: Zero-point energy excitation has a profound effect on the relaxation of benzene CH and CD overtone states. Only adding a fraction of the zero-point energy for each normal mode in the initial conditions results in smaller overtone relaxation rates. If no zero-point energy is added to C6H6, the n=3 and 5 CH overtones do not relax within 1 ps. Adding zero-point energy to different types of normal modes has nonequivalent effects on overtone relaxation. Zero-point excitation of modes with HCC bend character is particularly effective in enhancing relaxation of the overtones.

Notes: The photoionization of O2 near its ionization limit has been studied with coherent vacuum ultraviolet radiation produced by third harmonic generation in free jet expansions of the rare gases. High resolution (∼2 cm−1) photoionization spectra were obtained in the ionization threshold region from 103 to 98 nm which includes three vibrational levels of the H 3Πu (3sσ) Rydberg state. The H, v=0 photoionization spectrum was assigned by simulating the H 3Πu←X 3∑+g Rydberg excitation, yielding spectroscopic constants as well as the overall autoionization lifetime. The v=1 and 2 levels have distinctly different rotational band contours which reflect perturbations with bound and dissociative levels of nearby "dark'' states. The photoionization dynamics were probed further through measurements of photoelectron angular distributions for the v+=0 and 1 vibrational levels of O+2. In addition to strong variations in the asymmetry parameter (β) across the H state autoionization resonances, spectrally narrow variations in β were found in the surrounding continuum. These latter results suggest the presence of weak resonance features imbedded in the background continuum which nonetheless strongly influence the photoelectron ejection dynamics.

Notes: The rate of intramolecular electronic excitation energy transfer in 9,9' bifluorene has been investigated in a variety of solvents, using both time-correlated single photon counting and femtosecond fluorescence upconversion technique. The kinetics of energy transfer were determined in both cases by time dependent fluorescence anisotropy measurements. The energy transfer dynamics between fluorene moieties has been found to occur on a time scale of approximately 600 fs in different solvents and has been correlated with the T2 value calculated from the absorption linewidth and the β value obtained from jet measurements. The dihedral angle between the fluorene moieties was also calculated from the anisotropy measurement and compared with the values obtained from a solution phase NMR determination.

Notes: Three-dimensional potential energy and electric dipole moment functions for the electronic ground state of H2O+ have been calculated from highly correlated multiconfiguration reference configuration interaction (MRCI) electronic wave functions. The analytic representations of these functions have been used in vibrational and perturbational calculations of the rovibrational absorption spectrum of H2O+. The quartic force fields in normal coordinates have been employed in the evaluation of the equilibrium spectroscopic constants in H2O+, D2O+, and HDO+ by perturbation theory. The equilibrium structure, vibrational band origins, centrifugal distortion constants and rotational energy levels agree very well with the available experimental data. Absolute vibrational band intensities have been calculated from the dipole moment functions and are compared with theoretical integrated band intensities. The radiative lifetimes of excited vibrational states exhibit mode specific variations. The rotationally resolved room temperature absorption spectra have been evaluated ab initio for the pure rotational and the ν2, 2ν2, ν1, ν3, and 3ν2 transitions. The rovibrational electric dipole transition matrix elements and absolute line intensities are given for the most intense transitions. These data take full account of anharmonicity effects and vibration–rotation coupling.

Notes: A simple model is proposed for correcting problems with zero point energy in classical trajectory simulations of dynamical processes in polyatomic molecules. The "problems'' referred to are that classical mechanics allows the vibrational energy in a mode to decrease below its quantum zero point value, and since the total energy is conserved classically this can allow too much energy to pool in other modes. The proposed model introduces hard sphere-like terms in action–angle variables that prevent the vibrational energy in any mode from falling below its zero point value. The algorithm which results is quite simple in terms of the cartesian normal modes of the system: if the energy in a mode k, say, decreases below its zero point value at time t, then at this time the momentum Pk for that mode has its sign changed, and the trajectory continues. This is essentially a time reversal for mode k (only!), and it conserves the total energy of the system. One can think of the model as supplying impulsive "quantum kicks'' to a mode whose energy attempts to fall below its zero point value, a kind of "Planck demon'' analogous to a Brownian-like random force. The model is illustrated by application to a model of CH overtone relaxation.

Notes: The photoion–photoelectron coincidence spectra for C2H+ and C2H+2 have been measured in the wavelength range of 645–765 A(ring). The C2H+2(A˜ 2Ag,B˜ 2∑+u) ions prepared with internal energies above 17.39 eV are found to dissociate completely into C2H++H in the temporal range 〈12 μs. An upper bound of 17.33±0.05 eV is determined for the appearance energy of the process C2H2+hν→C2H++H+e− at 0 K.

Notes: Vibrationally mediated photodissociation, in which one photon prepares a highly vibrationally excited molecule by vibrational overtone excitation and a second photon dissociates the vibrationally excited molecule, is a means of studying the spectroscopy and photodissociation dynamics of highly vibrationally excited states. Applying this dissociation scheme to nitric acid (HONO2) excited in the region of the third overtone of the O–H stretching vibration (4νOH) and detecting the OH fragment by laser induced fluorescence determines the energy partitioning and identifies the influence of vibrational excitation prior to dissociation. Vibrationally mediated photodissociation using 755 and 355 nm photons deposits more energy in relative translation than the isoenergetic single photon dissociation with 241 nm light. The former process also produces three times more vibrationally excited OH fragments, and both processes form electronically excited NO2, which receives over three-quarters of the available energy. In these experiments, vibrational overtone excitation enhances the cross section for the electronic transition by about three orders of magnitude. The observed differences are consistent with the motion of the vibrationally excited molecule on the ground electronic state surface strongly influencing the dissociation dynamics by allowing access to different electronic states in the photolysis step.

Notes: A procedure which offers computational advantages over the customary formalism is presented for calculating the shifts of the energy levels of discrete states embedded in a continuum. Numerical procedures are described and applied to the vibrational level shifts of the B 3Σ−u state of 16O2 interacting with the 5Πu continuum.

Notes: The mechanism of the radiative species Si*(1P0) formation from SiH4 was investigated with ab initio molecular orbital–configuration interaction calculations on two reaction routes: (a) the simplest one-step reaction mechanism SiH4→Si+2H2 and (b) the second simplest two-step reaction mechanism SiH4→SiH2+H2→Si+2H2. The conclusions obtained are as follows: (i) the one-step mechanism (a) operates effectively in the generation of the radiative species Si*(1P0), (ii) no effective path was found for the formation of Si*(1P0) in the two-step mechanism, (iii) the radiative species Si*(1P0) originates from the fifth lowest excited singlet state 11T1 of SiH4 at the Td structure in the one-step mechanism (a), whereas the nonradiative species Si(1D) is produced spontaneously from the fourth lowest excited singlet state 11E in both reaction mechanisms of (a) and (b), (iv) the nonradiative species SiH2(11B1) is spontaneously generated from the lowest excited singlet state 11T2 of SiH4, (v) the electronic state of the radiative Si*(1P0) has Rydberg character and the emission of Si*(1P0) at 288.2 nm is approximately expressed by 3p4s(1P0)→3p2(1D) where 4s is a Rydberg atomic orbital (AO). It was also concluded that the contributions of 3d AOs are important for making a description of the quantitative energy separations among the lower lying singlet states, 1D, 1S, and 1P0 of Si atoms.

Notes: Inelastic scattering of electronically excited Na atoms by ground state O2 molecules was studied theoretically using a multiple-curve-crossing model. The movement of the collisional system within the potential grid describing the Na–O2 and Na+– O−2 pairs was visualized for two initial electronic states of Na (5S and 4D) at collision energies ranging between 0.2 and 1.0 eV. Visualization of the collision path was used to better understand prominent features of the redistribution of energy by collision within the system. Knowledge of the collision path for inelastic scattering also helped in discussing a competing reactive channel.

Notes: In this paper we present two new formulations for the time-independent quantum mechanical calculation of photodissociation amplitudes. The first is based on a variational L2 amplitude density approach, and the second is based on a new scattered wave variational principle, both approaches having been developed previously for the treatment of general reactive scattering problems. It is shown that, apart from an inhomogeneity term, the algebraic equations which must be solved are identical in form to those already successfully treated in recent three-dimensional, converged quantum reactive scattering studies. The new variational principles should provide a practical method for carrying out converged, three-dimensional quantal calculations for photodissociation processes in which any number of fragmentation pathways are possible.

Notes: Electron mobilities μ were measured in dense gaseous krypton as a function of density normalized electric field E/n at 3.8≤n/1026 molecule m−3≤40 and 152≤T/K≤250. At each density a constant value of the mobility μ0 is attained at low E/n. At fields higher than a threshold (E/n)th μ first increased, passed through a maximum, and then decreased. In the saturated vapor nμ0 decreased with increasing (n,T) while at constant n, nμ0 increased with T. The density dependence of nμ0 is compared to the dielectric screening model of Baird [Phys. Rev. A 32, 1235 (1985)].

Notes: Mutual diffusion in poly(ethylene oxide)/polymethyl methacrylate blends is investigated by using the photon correlation spectroscopic technique. The mutual diffusion coefficient and the static structural factor are measured as a function of concentration for various temperatures. The Onsager coefficient is obtained by simultaneous measurements of the diffusion coefficient and static structural factor. The result is employed to test theoretical predictions on the concentration dependence of the mutual diffusion coefficient of the polymer blend. The experiment results are found not in agreement with theoretical predictions.

Notes: Thermal broadening of spectral holes burnt into excitonic states of long chain molecular aggregates of pseudoisocyanine iodide is investigated over a temperature range from 350 mK to 80 K. The results differ very much from the usual behavior of small molecules in glasses. We found that extended states are almost completely decoupled from the amorphous host lattice and spectral diffusion effects play a minor role. The homogeneous linewidth is independent of temperature below 10 K. Above, thermal broadening occurs in two steps: there is a weak onset around 15 K and a strong onset around 65 K. The data can be fitted by a superposition of two exponentials.

Notes: Recently we have shown that the lowest triplet state (T0) of pyridine, incorporated in a single crystal of benzene, may be studied by electron spin-echo (ESE) spectroscopy. From the nitrogen hyperfine structure in the ESE detected electron paramagnetic resonance (EPR) spectra, we were able to conclude that pyridine, a planar molecule in the ground state, becomes nonplanar upon excitation into T0. Here we report the results of a detailed investigation of this distortion and of the electronic nature of the lowest triplet state of pyridine-d5. We have performed electron spin–echo envelope modulation (ESEEM) spectroscopy. From the modulation spectra, the electron-nuclear double resonance (ENDOR) frequencies corresponding to the various deuterium nuclei are obtained. Analysis of the dependence of these frequencies on the orientation of the magnetic field with respect to the triplet fine-structure axes system allows for a determination of the deuterium hyperfine and quadrupole tensors. From these tensors and the known nitrogen hyperfine tensor, the structure and spin-density distribution of pyridine in its lowest triplet state are deduced. Pyridine adopts upon excitation into T0 a boatlike structure, in which the nitrogen atom is tilted by about 40° with respect to the plane through the ortho- and meta-carbon atoms and the para-carbon/para-deuterium fragment by about 10° with respect to this plane. Thereby the hybridization of the ortho-carbon atomic orbitals strongly deviates from that for aromatic hydrocarbons and becomes almost sp3; the hybridization of the atomic orbitals on the other carbon atoms changes much less, while the nitrogen atomic orbitals remain sp2 hybridized. Approximately half of the spin density is found to be localized on the nitrogen atom with a remarkable distribution over the atomic orbitals: the π orbital carries 40%, the n orbital 10%. The rest of the spin density is distributed over the para- (30%) and ortho-carbon atoms (10% each). The lowest triplet state of pyridine is neither an nπ* state nor a ππ* state, but a state of mixed character as the result of a strong vibronic coupling between the 3B1 (nπ*) and 3A1 (ππ*) states.

Notes: A method is presented for calculating energy levels of atom–rigid-diatom systems for various values of the total angular momentum (J) of the complex. The technique is based upon the collocation method for the vibrational motions of the system and the Galerkin approach for the total rotation. Unlike the Rayleigh–Ritz variational principle, the method does not require the evaluation of integrals over the Hamiltonian and so is very simple to implement. An important feature of the method is that the wave function is obtained in an analytic form and so it is a simple matter to calculate many quantities of spectroscopic interest such as rotational constants and spectral intensities. It is also shown that contracted basis sets can be used in conjunction with the collocation method to enhance the efficiency of the calculation. The method is demonstrated by calculating rovibrational levels of the van der Waals complex Ar–HCl for J up to 10.

Notes: Phase transformation kinetics that occurs by a nucleation and growth process is investigated. A simple discrete space and time model is used for the dynamics and analytical results are obtained for the volume fraction of the material transformed for both finite systems and a special example of an inhomogeneously nucleated system. The theory is developed for two cases, initial nucleation, and continuous nucleation. The results are compared with simulations of the model.

Notes: We investigate the one-body structure of fluids confined within model pores when the confinement is sufficient to force the fluid into a quasi lower-dimensional regime. The concept of the quasi-two-dimensional regime of atomic fluids confined to symmetric planar slit geometry is introduced and statistical mechanical arguments are shown to predict a universal form for the density profile; or to be more precise, to imply that the many-body contribution to the one-body structure is a particular universal function of position. Computer simulation data of Lennard-Jones fluid confined to planar slit pores are presented, which clearly illustrate the statistical mechanical prediction. Generalizations of this phenomenon to include fluid mixtures, molecular fluids and nonplanar geometries are discussed.

Notes: Measurements of the interfacial tension γ between immiscible polymers of low molecular weight have not resolved the question of how γ depends on the degree of polymerization N. In this paper the issue is addressed theoretically by solving equations derived earlier, but previously solved only in the infinite molecular weight limit. The basic origin of molecular weight dependence is entropic adsorption of chain ends to the interfacial region. It is found that the interfacial tension and concentration profile across the interface approach the infinite degree of polymerization limit like N−1. No sign is found of the N−2/3 approach reported to fit the data, although not unambiguously. If correct, N−2/3 may be a law applicable over a limited range, and not asymptotically.

Notes: A family of mechanistic models are proposed to describe cooperative molecular displacements in glass-forming liquids. These models assume that within each cooperative domain motion occurs by sequences of discrete, localized events, but that each of these events involves synchronous displacement of a smaller cluster of neighboring components. The size and properties of this cluster, described as a "primitive subunit,'' are assumed to reflect intrinsic details of local structure, not sensitive to external degrees of freedom. However, the length of the sequence of events by which each subunit moves is assumed to be a statistical function of some communal degree of freedom such as the free volume. Two examples are explored for the case of a polymer melt in which idealized conformational rearrangements are constrained by steric interferences. The distribution of lengths for cooperative sequences of events required to remove these interferences is derived as a function of a parameter β related to free volume. It is shown that the mean length of such sequences diverge to infinity for some nonzero (critical) value of β, and that the divergence obeys a scaling law. The divergence is a time-invariant feature of the model, similar in physical significance to the equilibrium phase transition which has been proposed as the underlying basis for the glass transition. However, in the present models it follows from mechanistic constraints, independent of any thermodynamic considerations.

Notes: The spectral densities of motion J1(ω) and J2(2ω) were measured at 15.3 MHz in the nematic phase of p-methoxy-d3-benzylidene-d1-p-n-butyl-d9-aniline for the methine and butyl chain deuterons. The present study concentrates on measurements near the clearing temperature as well as extends measurements to temperatures lower than those of an earlier report. We propose to interpret these spectral density parameters using a new model of molecular reorientation suggested by Vold and Vold. The reorientation motion of the molecules in a Maier–Saupe potential is governed by three rotational diffusion rate constants Dα, Dβ, and Dγ. The subscripts in the rate constants are the Euler angles ΩL,M which are involved in time-dependent transformation between the molecule fixed frame and the laboratory (director) frame. Dα and Dβ are the principal components of a rotational diffusion tensor in the director frame. The motion about the molecular ZM axis (γ motion) is assumed to be independent of the α,β motions. The internal chain dynamics is treated using the superimposed-rotations model in which rotation about each C–C bond can either be free or jump among t (trans) and g± (gauche) sites. The diffusion constants Dα, Dβ, and Dγ are found to show the Arrhenius temperature dependence. Similar behaviors are observed in the derived rotational diffusion constants for internal rotation of the C–C bonds in the butyl chain of MBBA.

Notes: We describe the apparatus necessary for the application of a previously reported classical path method to three-dimensional reactive collisions. The case of D+H2 is used as an example.

Notes: When energized sufficiently either vibrationally or electronically, ROH (where R is methyl or ethyl) can dissociate to form H atoms and RO radicals. We have determined the translational energy release (〈ETr 〉=0.82Eavl ) and angular distribution (β=−0.60±0.03) from the laser induced fluorescence spectra of H atoms produced in the 193 nm photodissociation of CD3OH. We have also determined that the quantum yield for producing H from CD3OH is 0.86±0.10. In contrast, the reaction of O(1D)+CH4 which produces vibrationally excited CH3OH, has a quantum yield for producing H atoms of roughly 0.25 with only 22% of the available energy released as translation. We conclude that although the total available energy is the same in both cases, the dissociation of photoexcited methanol is prompt whereas the dissociation of chemically activated methanol shows some degree of internal vibrational equilibration.

Notes: The reaction between ground state atomic oxygen and ethylene was studied under single collision conditions using the crossed molecular beam method. At an average collision energy of about 6 kcal/mol, the two major primary reaction channels are (a) the formation of CH3 and CHO and (b) the formation of C2H3O and H. Product angular distributions and time-of-flight spectra were measured and the translational energy release was determined for each channel. The observed results and calculated potential energy surfaces suggest that after the addition of O(3P) to ethylene forming a triplet diradical, channel (a) occurs by way of intersystem crossing to the singlet state, 1,2-H migration and subsequent C–C bond rupture, whereas channel (b) proceeds mostly through the direct dissociation of the intermediate triplet diradical, except for a small contribution from H atom elimination of the singlet acetaldehyde intermediate.

Notes: A mechanism for the enhanced splitting detected in the millimeter-wave rotational spectra of the first excited S–S stretching state of HSSH (disulfane) has been studied. The mechanism, which involves a potential coupling between the first excited S–S stretching state and excited torsional states, has been investigated in part by the use of ab initio theory. Based on an ab initio potential surface, coupling matrix elements have been calculated, and the amount of splitting has then been estimated by second-order perturbation theory. The result, while not in quantitative agreement with the measured splitting, lends plausibility to the assumed mechanism.

Notes: The X˜ 1A1, a˜ 3B2, and b˜ 3A2 states of vinylidene are observed in the ultraviolet (351.1–364.0 nm) photoelectron spectra of X˜ 2B2 H2CC−, X˜ 2B2 D2CC−, and X˜ 2A' HDCC−. The X˜ 1A1 state exhibits vibrational structure well above the barrier for isomerization to acetylene. A strict lower bound to the lifetime of the singlet state against rearrangement is τ〉0.027 ps, with an estimate of τ≈0.04–0.2 ps based on a simulation of the line shapes including rotational broadening. A vibrational analysis of the singlet and lower triplet state bands provides vibrational frequencies and estimates of the changes of molecular geometries between the anion and the neutral species. A qualitative potential energy surface for the CH2 rock mode, which closely corresponds to the reaction coordinate for isomerization, is extracted from the experimental data. The adiabatic electron affinity is EA(X˜ 1A1 H2CC)=0.490±0.006 eV and the triplet term energies are T0(a˜ 3B2 H2CC)=2.065±0.006 eV and T0(b˜ 3A2 H2CC)=2.754±0.020 eV. Experimental values for the bond dissociation energy of vinyl radical, D0(H2CC–H)=80.0±5.0 kcal/mol, and the acetylene–vinylidene isomerization energy, ΔHI=46.4±5.5 kcal/mol, are derived. Combining the latter value with the upper limit of Field and co-workers, ΔHI≤44.1–44.7 kcal/mol, yields ΔHI≈41–45 kcal/mol.

Notes: A vibrational analysis for the excitation and fluorescence spectra of the S1–S0 transition in jet-cooled indole and N-deuterated indole is presented. This analysis yields the frequencies for eight low-lying S1 vibrational modes of both isotopic species. Single vibronic level fluorescence spectra also enable us to refine many of the S0 vibrational frequencies. Quantum interference effects in the energy-resolved fluorescence decays for one vibrational level of S1 indole are observed as a result of picosecond laser excitation. Through analysis of the fluorescence spectra and the modulated fluorescence decays, the coupling responsible for the quantum beats was found to result from a zero-order in-plane mode coupled to a level which is described by several zero-order out-of-plane modes. The effect of the deuterium substitution on the quantum beats is also investigated.

Notes: A picosecond time-domain light scattering technique is used to study the viscoelastic KNO3–Ca(NO3)2 60:40 glass-forming liquid mixture. By using scattering angles between 1.92° and 85.67°, acoustic frequencies from 50 MHz to 4 GHz are sampled. Together with existing ultrasonic and Brillouin scattering data, a temperature-dependent distribution of relaxation times is found to be well fit with a Cole–Cole distribution whose width changes from several decades in the glassy state to nearly single relaxation time in the high-temperature liquid state. The characteristic relaxation time is found to obey the Vogel–Tammann–Fulcher law with T0=338 K.

Notes: We present a simple model of coupled locally anharmonic oscillators which can be used to describe the optical nonlinearities in conjugated organic monomeric, oligoneric, and polymeric structures. We show that the method can very readily be used to explain the dependence of the band gap, the polarizability α, and the second hyperpolarizability γ on the number of repeat units of conjugated chain compounds by adjusting two parameters: the local anharmonicity term and the oscillator coupling constant. To illustrate the usefulness of this model, we have calculated the dependence of the band gap, the polarizability α, and the second hyperpolarizability γ, as the function of the number of repeat units for the oligomers of thiophene and benzene. The results predicted by the coupled anharmonic oscillator model are in good agreement with those of the experimental studies on thiophene and benzene oligomers recently reported by our group. In addition, the predicted power dependences of orientationally averaged 〈α〉 and 〈γ〉 on the number of repeat units are compared with those predicted by a free electron model, PPP method, sum-over-states method, and ab initio calculations.

Notes: Time-resolved photodissociation of styrene ions at 308 nm was observed in the ion cyclotron resonance ion trap. The rate of m/z 78 product ion formation was shown to be a sensitive function of the internal energy of the dissociating parent ions. The dissociation rates, ranging up to 5×105 s−1, were near the upper limit of the technique, and an improved signal equation was derived for quantitative interpretation of the data. By observing the m/z 78 photoappearance rate as a function of pressure and delay time between electron impact ionization and the laser pulse, the relaxation of initially excited styrene ions was characterized, giving a collisional relaxation rate constant of 4.5×10−10 cc molecule−1 s−1, and an IR radiative relaxation rate constant of 0.65 s−1. At 11.5 eV electron impact ionizing energy the parent ions were found to be formed with an average initial excess internal energy of 0.45 eV.

Notes: The excitation of H2S at 66 000 cm−1 by two-photon absorption is shown to produce significant quantities of H2 molecules in very high vibrational, but low rotational levels. The vibrationally excited H2 has been identified by a 2+1+1 resonance enhanced multiphoton ionization (REMPI) process in whch the E˜,F˜ state is excited as an intermediate state. The major contribution of this double well state comes from the outer F˜ well because of the large H2 bond distance in the nascent H2(v) photoproduct. Excitation of the E˜,F˜ state by a 291 nm photon dissociates the molecule to the H+H*(n=3) products, and the excited atom is then ionized by a fourth photon. The analysis of the photoelectron spectra (PES) demonstrates that the intermediate H2S* molecule preferentially dissociates to vibrationally excited H2 molecules rather than ionizing by the absorption of an additional photon. It appears likely that the photodissociation of H2S is a practical method for producing highly vibrationally excited, but rotationally cool, H2 molecules for further dynamical studies.

Notes: Quasiclassical trajectory calculations have been performed for motion on an ab initio potential energy surface to investigate energy transfer from N2(A 3∑+u, v=1, 3 and 6) to H2 and D2. Because of the unusual features of the surface, both vibrational relaxation and electronic quenching of N2(A) are observed, the latter process resulting in dissociation of the hydrogen molecule. It is deduced that coupling of the vibrational motions of the N2 and H2 molecules initiates the energy transfer process. The results are compared with experimental information on the quenching of N2(A) by σ-bonded molecules.

Notes: Bimolecular rate constants for the thermal chemical reactions of muonium (Mu) with the halogen gases—Mu+X2→MuX+X—are reported over the temperature ranges from 500 down to 100, 160, and 200 K for X2=F2,Cl2, and Br2, respectively. The Arrhenius plots for both the chlorine and fluorine reactions show positive activation energies Ea over the whole temperature ranges studied, but which decrease to near zero at low temperature, indicative of the dominant role played by quantum tunneling of the ultralight muonium atom. In the case of Mu+F2, the bimolecular rate constant k(T) is essentially independent of temperature below 150 K, likely the first observation of Wigner threshold tunneling in gas phase (H atom) kinetics. A similar trend is seen in the Mu+Cl2 reaction. The Br2 data exhibit an apparent negative activation energy [Ea=(−0.095±0.020) kcal mol−1], constant over the temperature range of ∼200–400 K, but which decreases at higher temperatures, indicative of a highly attractive potential energy surface. This result is consistent with the energy dependence in the reactive cross section found some years ago in the atomic beam data of Hepburn et al. [J. Chem. Phys. 69, 4311 (1978)]. In comparing the present Mu data with the corresponding H atom kinetic data, it is found that Mu invariably reacts considerably faster than H at all temperatures, but particularly so at low temperatures in the cases of F2 and Cl2. The current transition state calculations of Steckler, Garrett, and Truhlar [Hyperfine Interact. 32, 779 (986)] for Mu+X2 account reasonably well for the rate constants for F2 and Cl2 near room temperature, but their calculated value for Mu+Br2 is much too high. Moreover, these calculations seemingly fail to account for the trend in the Mu+F2 and Mu+Cl2 data toward pronounced quantum tunneling at low temperatures. It is noted that the Mu kinetics provide a crucial test of the accuracy of transition state treatments of tunneling on these early barrier HX2 potential energy surfaces.

Notes: The time evolution of spins-1/2 subject to radiation damping, which is commonly encountered for solvent peaks at high field, is examined in detail. The well-known analytic results for rectangular pulses on undamped spin-1/2 systems are extended to the radiation damped case, and reveal surprisingly complex dynamics. Explanations in terms of Bloch vectors are also presented, and composite pulse sequences which would also be insensitive to radiation damping are proposed. In addition, gradient optimization programs were developed to find shaped π and π/2 pulses insensitive to radiation damping. The optimized pulses compensate for radiation damping effects even when the characteristic damping time is shorter than the pulse length.

Notes: We present a theoretical description of the aggregation equilibrium of nonswollen and water-swollen micelles in oil (or vice versa). While we specialize on spherical, noninteracting aggregates, we treat the competition between phase separation, dissolution in monomers and formation of micelles with variable extent of swelling. The present model is based on an interfacial free energy of the surfactant monolayer which includes stretching and bending contributions. This free energy allows one to desribe both, the macroscopic interface in a two-phase system and the internal interface surrounding the aggregates on equal footing. The extent of swelling of the micelles at the CMC is determined by the bending energy of the saturated surfactant monolayer. Increasing the splay modulus or decreasing the spontaneous curvature favors swollen micelles (microemulsion droplets) and ultralow interfacial tensions. We present explicit results for the size distributions of the aggregates and interfacial tensions. Considering shape fluctuations of the aggregates we conclude that spherical micelles may be stabilized by small interfacial tension alone, while stability of strongly swollen microemulsion droplets requires a finite splay modulus of the monolayer.

Notes: Clusters of argon, krypton, and xenon are grown in a free jet and ionized by electron impact. The size of these clusters, (Rg)+n, extends up to n(approximately-equal-to)1000. Individual cluster sizes are mass resolved up to n(approximately-equal-to)570 in the case of Ar+n. The well known, but puzzling differences in the size distributions of Kr and Xe clusters disappear beyond n(approximately-equal-to)130, while those between Ar and Xe disappear beyond n(approximately-equal-to)220. The most pronounced "magic numbers'' in the distributions of large cluster ions occur at n=147 (148 for Ar), 309, and 561, in striking agreement with the number of atoms required to build icosahedral clusters with 3, 4, and 5 complete coordination shells, respectively. Closure of the 6th icosahedral coordination shell is indicated by another strong intensity drop at n(approximately-equal-to)923 in the unresolved part of the spectra. Several additional intensity extrema are observed between major shell closures. A simple structural model, assuming an icosahedral core decorated by the additional atoms, accounts for these anomalies reasonably well up to n=561.