Volume 119, Issue 5, 01 August 2003
 ARTICLES

 Theoretical Methods and Algorithms

Massive thermostatting in isothermal density functional molecular dynamics simulations
View Description Hide DescriptionThis paper demonstrates the excellent temperature control and rapid equipartioning of the kinetic energy of the massive generalized Gaussian moment thermostat (MGGMT, one thermostat is coupled to each degree of freedom) in isothermal density functionalmolecular dynamics (MD) simulations on the Born–Oppenheimer potential energy surface. The MGGMT is implemented in the approach and, as far as we know, it is the first time in literature that the MGGMT is combined with density functional methods. The performance of the MGGMT approach is illustrated with MD simulations of the iron porphyrin–imidazole–carbon monoxide [FeP(Im)(CO)] complex and compared with constant energy MD simulations on the same system. Both MD approaches lead to similar average structures of the FeP(Im)(CO) complex. The examination of the frequency distribution functions reveals that the structural dynamics are not seriously affected by the dynamics of the parameters introduced by the MGGMT. The equipartitioning rates in the MGGMT simulations are significantly faster than in the constant energy simulation. We recommend the MGGMT approach as an very efficient equilibration technique in MD simulations and it emerges as a useful technique for, e.g., simulated annealing and nonequilibrium MD simulations.

Quantumclassical approximation beyond Redfield theory
View Description Hide DescriptionA quantumclassical approximation, capable of describing the evolution of open quantum systems well beyond the applicability limits of Redfield theory is suggested. The theory is based on the short lifetime of the quantum correlations between the quantum and the classical subsystem, caused by energy dispersion (per degree of freedom) of the canonical bath. The resulting quantumclassical approximation has the form of two auxiliary differential equations and fully accounts both for the arbitrary long memory of the heat reservoir and detailed balance. These equations allow direct solution in the time domain without constructing/diagonalizing Liouville space operators, and, in combination with molecular dynamics techniques to simulate bath dynamics, may be applied to quantum subsystems with a fairly large number of levels. A simple example of a twolevel system, coupled to a single correlation time canonical bath, was considered to demonstrate different regimes of approaching the canonical equilibrium state.

Molecular dynamics approach to vibrational energy relaxation: Quantumclassical versus purely classical nonequilibrium simulations
View Description Hide DescriptionWe present an efficient method for the direct solution in the time domain of the equations of a novel recently proposed nonMarkovian quantumclassical approximation, valid well beyond the applicability limits of both Redfield theory and Fermi’s Golden Rule formula. The method is based on an ab initiomolecular dynamics description of the classical bath and is suitable for applications to systems with a fairly large number of quantum levels. A simple model of the breathing sphere in a LennardJones fluid was used to compare the results of the quantumclassical and purely classical treatments of vibrational energy relaxation.
 Gas Phase Dynamics and Structure: Spectroscopy, Molecular Interactions, Scattering, and Photochemistry

Thermal electron emission from the hot electronic subsystem of vibrationally cold
View Description Hide DescriptionWe compare theoretical results on statistical electron emission from electronically hot but vibrationally cold with recent experimental results involving excitation with ultrashort laser pulses. Both photoelectron spectra and ion yields of as well as fragment ions are compared with the predictions of the statistical electron emission model. Quantitative agreement is obtained between the calculations and the experimentally measuredphotoelectron spectra, total ion yields and singlydoubly charged branching ratio. The electronphonon coupling time is fitted to a few hundred femtoseconds, consistent with independent measurements. The data allow a determination of the thermal properties of the electron system, which are consistent with the theoretical input. The data also allow a fit of the averaged photon absorption cross section.

Conformations of indan and 2indanol: A combined study by UV laser spectroscopy and quantum chemistry calculation
View Description Hide DescriptionThree conformational isomers of 2indanol are identified by use of resonance enhanced twophotonionization (R2PI) and single vibronic level dispersed fluorescence spectroscopy in a supersonic jet expansion. By combining the experimental results with the predictions of the ab initio quantum chemistry calculations at the level of theory, the major species is identified as a conformational isomer in which the hydroxyl hydrogen is involved in an intramolecular hydrogen bonding with the πelectrons of the aromatic ring. The theoretical estimate of the hydrogen bond energy is ∼6.5 kJ/mol. A comparative investigation with indan reveals that this weak hydrogen bonding in the former significantly affects the puckering potential of the fivemember side ring. The dispersed fluorescence data indicate for a much higher ringpuckering barrier in the ground state than what has been suggested recently by measuring rotational spectra of the unsubstituted indan.

The interaction of oxygen with small gold clusters
View Description Hide DescriptionPresented in this work are the results of a quantum chemical study of oxygen adsorption on small and clusters. Density functional theory(DFT), second order perturbation theory (MP2), and singles and doubles coupled clustertheory with perturbative triples [CCSD(T)] methods have been used to determine the geometry and the binding energy of oxygen to The multireference character of the wave functions has been studied using the complete active space selfconsistent field method. There is considerable disagreement between the oxygen binding energies provided by CCSD(T) calculations and those obtained with DFT. The disagreement is often qualitative, with DFT predicting strong bonds where CCSD(T) predicts no bonds or structures that are bonded but have energies that exceed those of the separated components. The CCSD(T) results are consistent with experimental measurements, while DFT calculations show, at best, a qualitative agreement. Finally, the lack of a regular pattern in the size and the sign of the errors [as compared to CCSD(T)] is a disappointing feature of the DFT results for the present system: it is not possible to give a simple rule for correcting the DFT predictions (e.g., a useful rule would be that DFT predicts stronger binding of by about 0.3 eV). It is likely that the errors in DFT appear not because of gold, but because oxygen binding to a metal cluster is a particularly difficult problem.

Crossedbeam scattering of The integral cross sections
View Description Hide DescriptionThe title reaction was investigated in a crossedbeam experiment. A resonanceenhanced multiphon ionization technique was used to interrogate the internalstate distributions of the product at three different collision energies. Only the (umbrella) mode excitation was observed. Its distribution changes from a monotonically declined distribution at low energy to a slightly inverted one at higher collision energy. Although the rotational excitations of were small, a strong preference for was found, indicative of the dominance of the tumbling rotation motion of the product. The vibrationresolved excitation functions were also measured for A reaction barrier of 0.5 kcal/mol was deduced.

Ab initio derived analytical fits of the two lowest triplet potential energy surfaces and theoretical rate constants for the system
View Description Hide DescriptionThis work presents two new analytical fits of the ground potential energy surface (PES) and the first excited PES involved in the title reaction, considering the Nabstraction (1) and the Oabstraction (2) reaction channels, and the reverse reaction (−1). The PESs are derived from ab initio electronic structure calculations by means of secondorder perturbation theory on a complete activespace selfconsistentfield wave function (CASPT2 method). Stationary points and extensive grids of ab initio points (about 5600 points for the PES and 4900 points for the PES) were fitted along with some diatomic spectroscopic data to better account for the experimental exoergicity. Thermal rate constants were calculated (200–5000 K) for all mentioned reaction processes by means of the variational transitionstate theory with the inclusion of a semiclassical tunneling correction. Excellent agreement with the experimental data was observed for reaction (1) and its reverse, within all the temperature range, substantially improving the results derived from previous analytical PESs. The contribution of the PES to the reaction rate constant was small even at high temperatures (e.g., only 10.8% at 2500 K). Moreover, the main contribution to reaction rate constant was due to the PES, differing from what happens for reaction (1). The Oabstraction reaction channel accounts for a 3.0% of the total reaction at 5000 K, consistent with the very limited experimental information available.

Laserinduced fluorescence, electronic absorption, infrared and Raman spectra, and ab initio calculations of 1,2dihydronaphthalene: Investigation of the outofplane ring modes for the ground and excited states
View Description Hide DescriptionThe laserinduced fluorescencespectra and dispersed fluorescence spectra of jetcooled 1,2dihydronaphthalene have been analyzed to investigate the ring inversion process in both the and excited states. Ultraviolet absorption, infrared, and Raman spectra were also recorded to complement the analyses.Ab initio calculations predict the inversion process to involve four outofplane ring motions, and linear combinations of these were made to model the inversion process. The data show the barrier to inversion in the ground state to be (the triplezeta ab initio value is The experimental data indicate that the barrier increases substantially in the excited state, for which the calculated barrier is with a basis set.

Instantaneous nonvertical electronic transitions with shaped femtosecond laser pulses: Is it possible?
View Description Hide DescriptionIn molecular electronic transitions, a vertical transition can be induced by an ultrashort laser pulse. That is, a replica of the initial nuclear state—times the transitiondipole moment of the electronic transition—can be created instantaneously (on the time scale of nuclear motion) in the excited electronic state. Now, applying pulse shaping via the modulation of the phases of each spectral component of an ultrashort pulse, it is tempting to ask whether it is also possible to induce instantaneous nonverticaltransitions to bound electronic states, provided that the phases of each spectral component of the pulse are set to appropriate values at the discrete frequencies corresponding to the energy levels of the potential. We analyze the problem in the weakfield limit, and show that such a phase requirement cannot be encoded into an ultrashort pulse. This result is equivalent to the statement that it is not possible to move matter faster than the time associated with the natural (fieldfree) dynamics of the system.

Formation of ozone: Metastable states and anomalous isotope effect
View Description Hide DescriptionA clear explanation for an anomalous isotope effect in ozone formation is given in terms of the energy transfer mechanism, where the metastable states of ozone are formed first, and then stabilized by collisions with other atoms. Unusual nonstatistical properties of metastable states spectra discovered earlier [J. Chem. Phys. 118, 6298 (2003)] are incorporated into the kinetics model, where different metastable states are treated as different species, and the stabilization step is treated approximately. The population of the ozone metastable states builds up and decays through three possible channels. When different isotopes of oxygen are involved the three channels become open at different energies because of the differences in the quantum zeropointenergies (ΔZPE) of the different molecules. The spectrum of metastable states is anomalously dense below the ΔZPE threshold and these states are accessible only from the lower entrance channel. Also, these lowlying metastable states are stabilized very efficiently (by collisions with third body) because they are energetically close to the bound states. Such processes significantly enhance the formation rates of ozone isotopologues through the lower channels over the formation rates through the upper channels. Numerical results obtained for give isotope effects in the right direction and of the right order of magnitude. Consideration of should improve the comparison with experiment.

First determination of the stretching frequencies by jet cooled intracavity laser absorption spectroscopy around
View Description Hide DescriptionUsing intracavity laser absorption spectroscopy combined with a supersonic slit jet, we have observed and analyzed seven vibronic cold bands of between 10 800 and The vibronic energies, relative intensities, and rotational constants, and the spinrotation constants have been determined. The rotational constants play a crucial role in the vibronic assignments. The seven observed states are vibronically mixed. However, three of them have a dominant electronic character while the four others have a dominant character. The vibrational assignments and energies of the three levels are: (1, 0, 0) at (0, 2, 0) at and (0, 0, 2) at The frequencies of the symmetric stretch, and bending, modes derived from the (1, 0, 0) and (0, 2, 0) levels are in agreement with ab initio calculations. In contrast, the frequency of the antisymmetric stretch, derived from the observed (0, 0, 2) level, agrees only with the ab initio value obtained by Kaldor while other ab initio values ranging from 390 to were reported. The four levels with a dominant character have been tentatively vibrationally assigned. The present experimental data allows for a first experimental determination of the two stretch vibrational frequencies which are necessary to model the vibronic interactions, i.e., the conical intersection between and potential energy surfaces.

A realistic multisheeted potential energy surface for from the double manybody expansion method and a novel multiple energyswitching scheme
View Description Hide DescriptionWe report a new multisheeted double manybody expansion potential energy surface that reproduces most known topological features of the title system. Near spectroscopic accuracy is conveyed to the groundstate sheet of in symmetry in the vicinity of the minimum by merging it with a spectroscopically determined Taylorseriesexpansiontype form via a novel multiple energyswitching scheme. A high energy ridge for insertion of into has also been imposed to mimic the result of accurate ab initio complete active space selfconsistent field and secondorder perturbation theory on CAS wave function calculations carried out for such geometries. This ridge decreases for geometries yielding a minimum barrier height for the reaction of 0.273 eV at a bent N–O–O structure defined by and Both the location and height of this barrier are in good agreement with existing ab initio calculations and the recommended values. Another salient feature is a shallow minimum on the potential energy surface that is separated from the absolute minimum by a conical intersection. Such a feature is accurately predicted by the newly reported ab initio calculations and well mimicked by the global double manybody expansion/energyswitching potential energy surface here reported. This is therefore commended both for spectroscopic and reactive dynamics studies on the title system. A final comment in relation to the conical intersection and the energyswitching scheme goes to the expected accuracy of current approaches for spectroscopically determined effective singlevalued forms.

Fluorescence spectra of Σ bands: Experiment and theory
View Description Hide DescriptionWe report new calculations and measurements of fluorescenceproperties of the title system, which is important in astrochemical processes. Dispersed fluorescence spectra show extensive vibrational progressions that depend on the initial state. Observed and calculated lifetimes are in good accord, save for (bent molecule notation), and calculated lifetimes are longer than the ones. The calculated laserinduced fluorescencespectrum is compared with experimental absorption data and with previous calculations, finding that the present treatment underestimates the intensity of the (0,4,0)Σ band.

Examination of the correlation energy and second virial coefficients from accurate ab initio calculations of raregas dimers
View Description Hide DescriptionCalculations of raregas dimers (He–He, Ne–Ne, Ar–Ar, He–Ne, He–Ar, and Ne–Ar) at the coupledcluster single double (triple) level of theory with large basis sets including bond functions and counterpoise corrections are reported over a wide range of 100 internuclear separations. These results are compared to experimental curves obtained from fitting to rovibrational spectra, and to second virial coefficients and Boyle temperatures. Accurate analytic potentials are developed for the total interactionenergy, Hartree–Fock (exchange) energy, and correlation(dispersion)energy; the transferability of the latter is demonstrated to very high accuracy even in the region of considerable wave function overlap. These calculations represent an important set of benchmarks that can be used to develop improved empirical molecular mechanical force fields and new quantum models.

Perturbational ab initio calculations of relativistic contributions to nuclear magnetic resonance shielding tensors
View Description Hide DescriptionWe present perturbational ab initio calculations of the leadingorder oneelectron relativistic contributions to the nuclear magnetic resonance shielding tensor based on the Pauli Hamiltonian. The scalar relativistic and spin–orbit interaction effects, including both relativistic corrections to the wave function (“passive” relativistic effects) and relativistic magnetic perturbation operators (“active” effects), are considered for (X=O, S, Se, Te, Po), HX (X=F, Cl, Br, I, At), and noble gas (Ne, Ar, Kr, Xe, Rn) systems. The perturbational corrections are calculated using linear and quadratic response theory applied to nonrelativistic reference states. We use the uncorrelated selfconsistent field as well as correlated, multiconfigurational complete active space selfconsistent field models. Results for the and heavyatom shielding constants and anisotropies are compared with Dirac–Hartree–Fock and quasirelativistic data.
 Condensed Phase Dynamics, Structure, and Thermodynamics: Spectroscopy, Reactions, and Relaxation

Integral equation theories for orientionally ordered fluids
View Description Hide DescriptionIntegral equationtheories have been quite successful at providing structural information for isotropic fluids. In particular, the expansion method originally proposed by Blum and Torruella [J. Chem. Phys. 56, 303 (1972)] has been pivotal in making these theories feasible for molecular fluids. Recently, integral equationtheories have also been applied to nematic phases for simple onecomponent fluids. In this article, we derive the formalism appropriate for fully, and partially, orientationally anisotropic fluids. Appropriate expansions of the pair and singlet functions are employed to obtain tractable expressions for the three coupled equations which must be solved. The numerical solution of these equations is discussed and simplifications resulting from symmetry are considered in detail. Numerical results are presented for fluids of dipolar ellipsoidal molecules in an external field.

Condensedphase effects on absolute cross sections for dissociative electron attachment to CFCs and HCFCs adsorbed on Kr
View Description Hide DescriptionWe present measurements of absolute dissociative electron attachment (DEA) cross sections to and adsorbed on the surface of Kr as a function of electron energy (0–10 eV). The DEA cross sections are measured to be at ∼0 eV, at 0.65 eV, and at 0.89 eV for and respectively. This cross section is similar to the gasphase value for while for the latter two molecules, it is orders of magnitude higher than the gaseous values. These results can be explained by considering the changes in the survival probability of the anion resonance and in the electron capture probability due to the decrease of the nuclear wave function overlap in the Franck–Condon region.

Selectively observing the amplitude modulation under magic angle sample spinning
View Description Hide DescriptionMagic angle sample spinning (MASS) averages the inhomogeneous magnetic field due to the susceptibility contrast in porous media by modulating the local magnetic field(frequency modulation). Molecular diffusion introduces a homogeneous broadening, which modulates the amplitude of the signal (amplitude modulation). The depth of amplitude modulation is determined by the interplay of molecular diffusion and MASS averaging and contains rich information on local magnetic fields, through which the spatial structure of the sample may be obtained. In this paper, we present two methods to quantify the amplitude modulation: a phase suppressed method and a more conventional twodimensional (2D)exchange method. The phase suppressed method directly observes the amplitude modulation in the time domain. The approximate equations are derived to extract the physical information from the time domain data. The conventional 2Dexchange spectrum contains both the frequency modulation and the amplitude modulation terms. The amplitude modulation introduces cross peaks in the 2D spectrum. The integration of the 2D spectrum along lines parallel to the main diagonal line will give a onedimensional spectrum, that is the Fourier transform of the amplitude modulation.

Electronic transport in disordered alkanes: From fluid methane to amorphous polyethylene
View Description Hide DescriptionWe use a fast Fourier transform block Lanczos diagonalization algorithm to study the electronic states of excess electrons in fluid alkanes (methane, ethane, and propane) and in a molecular model of amorphous polyethylene (PE) relevant to studies of space charge in insulating polymers. We obtain a new pseudopotential for electron–PE interactions by fitting to the electronic properties of fluid alkanes and use this to obtain new results for electron transport in amorphous PE. From our simulations, while the electronic states in fluid methane are extended throughout the whole sample, in amorphous PE there is a transition between localized and delocalized states slightly above the vacuum level (∼+0.06 eV). The localized states in our amorphous PE model extend to −0.33 eV below this level. Using the Kubo–Greenwood equation we compute the zerofield electron mobility in pure amorphous PE to be Our results highlight the importance of electron transport through extended states in amorphous regions to an understanding of electron transport in PE.