Search

The reproducing kernel Hilbert space (RKHS) method has been previously shown to be accurate and efficient in the construction of potential energy surfaces (PES) by interpolating fully gridded high level ab initio data. This paper extends the RKHS method to handle partially filled data calculated over arbitrarily shaped regions, while keeping nearly intact its accuracy and efficiency. The extension permits points or regions to be added to or removed from the grid as needed before doing expensive ab initio calculations, thus enabling the construction of RKHS PESs from the data distributions that are most likely to occur in practice. The utility of the new technique is demonstrated using data from the lowest global RKHS PES for the reaction O([sup 1]D)+H[sub 2], showing that ignoring the irrelevant regions of the PES does not adversely impact the accuracy of the surfaces if the relevant region is adequately sampled. © 2001 American Institute of Physics. [ABSTRACT FROM AUTHOR]

The adiabatic global potential energy surface of the H[sub 3][sup +] system for the lowest triplet excited state of A[sup ′] symmetry was computed for an extensive grid of conformations around the minimum region at full configuration interaction ab initio level, using a much more extended basis set than in a previous paper from the same authors. An accurate global fit (rms error lower than 27 cm[sup -1] for energies lower than dissociation into separated atoms and lower than 5 cm[sup -1] for energies lower than the dissociation channel) to these ab initio points and also to part of the previous calculated points (for a total of 7689 energies in the data set) of the lowest triplet excited state of A[sup ′] symmetry is obtained using a diatomics-in-molecules approach corrected by one symmetrized three-body term with a total of 109 linear parameters and 1 nonlinear parameter. This produces an accurate global potential which represents all aspects of the bound triplet excited state of H[sub 3][sup +] including the minima and dissociation limits, satisfying the correct symmetry properties of the system. The vibrational eigenstates have been calculated using hyperspherical coordinates with symmetry adapted basis functions with the proper regular behavior at the Eckart singularities. The accuracy of the vibrational levels thus obtained is expected to be better than 2 cm[sup -1] with respect to unknown experimental values. Due to the presence of three equivalent minima at collinear geometries (D[sub ∞h]) the lower vibrational levels are close to triple degenerate. Since the interconversion barrier between the three minima is about 2640 cm[sup -1], these states split for the upper excited vibrational levels. Such splitting can provide a key feature to identifying the unassigned transitions amongst the many H[sub 3][sup +] lines that have been observed in hydrogen plasmas. © 2001 American Institute of Physics. [ABSTRACT FROM AUTHOR]

In this paper, the weakest bound electron potential model (WBEPM) theory is employed to study the transition probability of many-valence-electron systems Cu I, Ag I, and Au I. According to this theory, a coupled equation is used to obtain the parameters Z[sup *], n[sup *], and l[sup *], which are needed in the calculation of transition probability. In the calculation, spectral fine structure is taken into consideration, and node numbers of silver and gold are adjusted. Compared with the accepted values, the results reported here are fairly good, and agreement is within the uncertainty of the accepted values. In our discussion, we give the explanation of node number adjustment from the view of WBEPM theory. © 2000 American Institute of Physics. [ABSTRACT FROM AUTHOR]

Due to the behavior of the pseudo-orbitals near nuclei, basis set selection in shape-consistent effective potential calculations can be problematic in studies using correlation techniques based on virtual orbital expansions (configuration interaction, etc.). In this paper we quantify the magnitude of the problem and give simple rules for the selection of correlating functions in a manner that parallels the use of simple diffuse primitives in all-electron work. Our calculations show the problem to be substantial for s and p basis sets, and to involve both one- and two-center effects, but is negligible for higher l. We also show that the problems can be avoided by the inclusion of one additional high exponent s primitive in the basis set and by augmenting p primitives with a higher exponent function so as to give zero first derivatives at the nucleus. By this recipe, computed s and p shell correlation energies in single and double promotion configuration interaction calculations can be increased by 3%-7%, as compared to basis sets using simple diffuse primitives for correlation. © 2000 American Institute of Physics. [ABSTRACT FROM AUTHOR]

In this paper, we describe implementation of the semiclassical Liouville method for simulating molecular dynamics on coupled electronic surfaces in the electronic adiabatic representation. We cast the formalism in terms of semiclassical motion on Born-Oppenheimer potential energy surfaces with nonadiabatic coupling arising from the coordinate dependence of the adiabatic electronic eigenstates. Using perturbation theory and asymptotic evaluation of the resulting time integrals, we derive an expression for the probability of transition between adiabatic states which agrees with the result given previously by Miller and George [W. H. Miller and T. F. George, J. Chem. Phys. 56, 5637 (1972)]. We also demonstrate numerically the equivalence of semiclassical trajectory-based calculations in the adiabatic and diabatic representations by performing molecular dynamics simulations on a model two-state system and comparing with exact quantum mechanical results. Excellent agreement between the exact and semiclassical treatments is obtained in both representations. © 2000 American Institute of Physics. [ABSTRACT FROM AUTHOR]

The potential-energy surfaces of the 1 [sup 1]A[sub g], 2 [sup 1]A[sub g], and 1 [sup 1]B[sub u] states of trans-1,3,5-hexatriene (THT) are explored in the vicinity of the ground state equilibrium structure. The S[sub 0] geometry optimization and force field calculation have been carried out with the restricted Hartree-Fock plus Mo\ller-Plesset second-order perturbation theory method. Vibronic coupling constants for the normal coordinates of a[sub g] and b[sub u] symmetry were computed with the complete-active-space self-consistent-field (CASSCF) and single state multiconfigurational second-order perturbation theory (CASPT2) electronic structure models. The CASSCF/CASPT2 method unequivocally places the vertical excitation energy of the dark 2 [sup 1]A[sub g] "phantom state" below the 1 [sup 1]B[sub u] level and predicts an energy difference of ca. 0.5 eV. The results are consistent with time-resolved photoionization yield and photoelectron spectroscopy experiments that indicate the existence of a low lying S[sub 1]-S[sub 2] conical intersection which induces rapid 1 [sup 1]B[sub u]→2 [sup 1]A[sub g] internal conversion on a time scale of 40 fs to 50 fs [Cyr and Hayden, J. Chem. Phys. 104, 771 (1996)]. Based on the vibronic coupling constants five totally symmetric vibrations with high Franck-Condon and/or tuning activity have been identified. The S[sub 1] and S[sub 2] states interact primarily via the two b[sub u] normal modes ν[sub 24] and ν[sub 26]. Other a[sub g] and b[sub u] normal vibrations do not appear to couple significantly to the lowest lying π→π[sup *] transition. The modeling of the ultrafast relaxation processes following optical excitation of the 1 [sup 1]B[sub u] state of THT and the calculation of absorption and resonance Raman spectra are discussed in the following paper. © 2000 American Institute of Physics. [ABSTRACT FROM AUTHOR]

It is shown that the quadratic steepest descent procedure developed in three previous papers goes smoothly and naturally into a limiting form, presenting no computational problems, when the Hessian becomes singular. This is in contrast to quasi-Newton algorithms which, in the case of singular or near-singular Hessians, yield unrealistic search directions and are kept from breaking down only by introducing arbitrary additional assumptions. The reason is that quasi-Newton algorithms give inadequate weight to the first order variation of the energy, i.e., its gradient. [ABSTRACT FROM AUTHOR]

Reports that in the article 'Theoretical characterization of the potential energy surface for H+O[sub2]...HO[sub2]...OH+O. IIII. Computed points to define a global potential energy surface,' from the Journal of Chemical Physics volume 94, p. 7068, the CASSCF/CCI calculations described in the paper were incorrectly identified as [4s3p2d1f/3s*2p] instead of [4s3p2d1f/3s*2p1d].