Library

Notes: In order to make an exact estimation of discretization error for finite/infinite element methods, it is necessary to develop analytical solutions for some transient mass transport problems in infinite media. These transient mass transport problems may be viewed as the benchmark problems for the discretization error estimation of a new numerical method so that they generally have the following characteristics: (1) their initial and boundary conditions can be exactly modelled by the finite/infinite element method; (2) their solutions can be rigorously expressed in a closed form. In this paper, several of the aforementioned problems have been constructed and solved mathematically for transient mass transport problems in both 1D and 2D infinite media.

Notes: In this communication, the application of the transmission line modelling (TLM) technique to the solution of the differential Riccati equation is described. A comparison is made between the TLM, fourth-order Runge-Kutta and the first-order Gear methods, for the case where one is applying the Kalman filter to the estimation of a voltage in a passive analogue circuit. In the particular example studied, the state equation is second-order. It is found that, when the system is underdamped, the fourth-order Runge-Kutta method has the best performance, followed by the TLM method. When the system is overdamped and stiff, the TLM method yields results that are closest to the analytical solution. Finally, a discussion is presented of the effects of errors in the TLM solution of the Riccati equation on the accuracy of the solution to the continuous Kalman filter estimation equation.

Notes: The paper describes the influence of contact area on the service life of dynamically loaded gears with a crack in a tooth root. Numerical results obtained with the finite element method using the program package BERSAFE are compared with the experimental results obtained on the universal testing machine INSTRON 1255. Numerical and experimental results are in good agreement. The analysis shows that the contact area directly influences the speed and direction of the crack propagation and the service life of the gear with a crack in a tooth root.

Notes: Mindlin-type plate and shell finite elements are the most widely used ones in the analysis of laminated composite structures. In the commercially available finite element programs, the transverse shear correction factors for these elements are usually the same as for structures made of isotropic materials. In the paper the shear correction factors are determined using energy principles for general laminates in cylindrical bending. Results for various materials and laminates are presented.

Notes: A B-spline finite element method is used to solve the regularized long wave equation numerically. This approach involves a Galerkin method with quadratic B-spline finite elements so that there is continuity of the dependent variable and its first derivative throughout the solution range. Time integration of the resulting system of ordinary differential equations is effected using a Crank-Nicolson approximation. Standard problems are used to validate the algorithm, which is then used to model the smooth development of an undular bore.

Notes: In the paper, a simple modification of the internal dissipation term in coupled thermoplastic finite element equations is presented. A modified form of the heat capacity and the thermomechanical coupling matrices are derived. These modifications are based on a decomposition of the internal dissipation term into two parts, which depend on the total strain rate and the rate of temperature change, respectively.

Notes: We examine a method to impose boundary conditions on arbitrary boundaries, introduced to make domains of infinite extent finite for the purpose of numerical calculations, when a finite element discretization based on linear, bilinear or trilinear elements is used, in one, two or three dimensions, respectively. In particular, we look at the so-called ‘free’ boundary condition, which consists in retaining the boundary integrals generated by the weighted-residuals formulation along the open boundaries and adding them to the stiffness matrix. We show that this procedure is exactly equivalent to imposing on the boundary nodes a Sommerfeld radiation condition in one dimension, and a slightly modified form of the Sommerfeld boundary condition in two and three dimensions. We also show that the procedure is not applicable to the purely elliptic case.

Notes: Recent advances in planar mesh generation of arbitrary domains incorporate methods for placing points as well as connecting them into a triangulation. Postprocessing techniques such as Laplacian smoothing and mesh relaxation enhance the shape of triangles in these meshes but do not address the construction of meshes near boundaries or provide a criterion to determine the number of interior points of the initial triangulation. The paper addresses these issues by investigating the construction of degree-6 triangulations, the primary goal of the mesh relaxation method.

Notes: The paper presents a distributed computing approach for concurrent finite element analysis of truss structures. The client-server model of distributed computation is applied to the plane frame finite element algorithm and implemented on SUN workstations. The remote procedure call combined with the ability to create threads of computation allows one to efficiently implement distributed and concurrent programs. The present approach is demonstrated through a typical truss structure to illustrate the details of the implementation. This approach can be used for very large finite element systems.

Notes: This work is devoted to the simulation by finite elements of nearly incompressible inviscid flows in real 3D geometries, by means of an Euler code based on the SUPG (streamline upwind Petrov-Galerkin) method, explicit forward Euler pseudo-temporal time integration and periodic and absorbing boundary conditions, among other features. The main goal is the application to flow around turbomachinery, with special emphasis on the performance analysis of a given machine, that involves several numerical computations at different operation points. Finally, these results are summarized in the form of characteristic curves.

Notes: A posteriori error estimation for typical Euler-Bernoulli beam bending problems is developed using a recovery operator based on a smoothing technique that is applied in the context of one-dimensional fourth-order problems. The development exploits a discrete superconvergence property of the ‘bending moments’ of the trial space interpolant of a polynomial solution of sufficiently low degree at certain Gauss points, and the superconvegence property of the energy norm difference of the trial space interpolant (of the weak solution) and the finite element approximation. An asymptotically exact error estimator is developed.

Notes: We discuss a two-step superposition method for calculating linear elastic stress intensity factors. The procedure requires the solution to the full cracked problem and the solution to a problem on the same mesh assuming the singularity due to a crack tip in an infinite region. We show that this is equivalent to the well known subtraction of singularity method if the two solutions are characterized by crack tip stress. The advantages of our procedure are that no modifications need to be made to a standard computer program and that once one singular solution is available on a given cracked mesh, solutions with different boundary conditions on the same mesh may be obtained in one step without including any singular crack effects. The mesh required to represent the singular crack tip field may also be studied independently of the complete problem. The additional computational cost of a two-step procedure is minimal since the solution matrix from step one may be reused with a new right-hand side. Numerical experiments using the boundary element method demonstrate the high accuracy and simplicity of the superposition approach.

Notes: The effect of upwinding on iterative performance of convection-diffusion problems is investigated. An analysis of the iterative method considered here leads to a criterion for selecting the optimal upwinding parameter to improve iterative performance for a class of two-dimensional convection-diffusion problems. Supporting numerical experiments are presented.

Notes: A numerical method is presented for the analysis of dynamic interaction between rigid wheel and flexible beam. In contrast to traditional approaches, the wheel nominal motion as well as the contact force are considered here as unknowns. The correct contact force between the wheel and the beam is computed by iteratively reducing the constraint error towards zero, and a simple time integration is employed for the solution of equations of motion. Convergence of the iterative scheme is analysed, and numerical simulations are conducted to demonstrate the accuracy of the solution.

Notes: The paper investigates the convergence of the h-, p- and hp-versions of a variational symmetric boundary element method (BEM) in plane elasticity by numerical experiments. The study discusses mixed boundary value problems on polygonal domains, i.e. problems for which the exact solution is analytic except in a finite number of points. The convergence of the error in energy norm is displayed for all versions of the BEM. All results are also compared with those obtained by the corresponding finite element methods. The theoretically predicted asymptotic convergence rates for all versions of the method can be observed in the numerical experiments. A comparison of computer times is given as well.

Notes: The integral transform method is employed in the analytical solution of the non-classical eigenvalue problem that appears in connection with the analysis of forced convection in ducts, including the effect of fluid heat diffusion in the axial direction. The related eigenfunctions are expanded in terms of eigenfunctions from a simpler auxiliary eigenvalue problem, and the original eigenvalues are determined from the associated matrix eigensystem analysis. Convergence rates of the proposed solutions are illustrated and reference results established for different values of the governing parameters.

Notes: Because of the earth's curvature, all existing channels have a curvature. E. Zeidler gave an existence and uniqueness proof for permanent heavy and capillar-heavy waves in such circular shaped channels. Based on this proof and on a general computational method for constructing the solutions numerically, the case of permanent heavy waves in circular-shaped channels is considered in this work.

Notes: In this paper, I discuss the convergence and divergence rates for a class of generalized matrix multisplitting relaxation methods in a detailed manner. In particular, when the coefficient matrix is an L-matrix, I obtain several sufficient and necessary conditions ensuring the convergence of these methods.

Notes: An independent refinement and integration procedure is developed to couple together independently modelled (global and local) regions in a single analysis. The finite element models can have different levels of refinement and the nodes along the interface between them need not coincide with one another. A spline interpolation function that satisfies the linear isotropic plate-bending differential equation is used to relate the local model interface nodal displacements to the global model interface displacements. The proposed independent refinement and integration procedure is validated by applying it to problems involving in-plane and out-of-plane deformations.

Notes: Both the logarithmic and derivative kernel integrations for potential problems, solved with quadratic isoparametric boundary elements, contain quartic functions of the integration parameter. It is shown that such functions can be written as the product of two quadratic functions with real coefficients. The derivative kernel integration is then represented as the sum of two integrals, which can be evaluated analytically. The logarithmic kernel integration can be similarly split but needs a Taylor series expansion of the Jacobian of the integration to enable analytical integration. The use of this Taylor series expansion means that the method presented is limited to problems involving weakly curved elements.

Notes: In this paper I re-examine the mechanics of the residual bending flexibility correction and show that it is an extravariational trick.

Notes: The determination of the dynamic response of plate and shell structures is carried out employing a discrete representation of the continuum, associated with an explicit integration scheme in the time domain. Comparisons with experimental results certify the accuracy and reliability of the proposed method.

Notes: Constraints arise naturally in the context of rigid inclusions, incompressibility, inextensible fibres and deformed finite elements. Typical methods for handling constraints in the governing matrix equation include Lagrange multipliers, penalty weights and elimination of variables. Each procedure has a particular undesirable feature. Proposed here is an approach in which each constraint is handled directly and sequentially through a modification of the rows and columns of the governing matrix and force vector. Positive definiteness, symmetry and the dimensions of the matrix remain unchanged. Elementary examples involving both the static and dynamic response of a bar are given to illustrate the procedure.

Notes: The focus of the present work is directed towards the construction of a family of special finite elements for the numerical solution of the Reissner-Mindlin plate problem. The model describes the deformation of a plate when transverse shear deformation is taken into account. The model is widely used for thin to moderately thick plates. Despite its simple formulation, the numerical approximation is not straightforward. When standard low-order finite elements are used for the approximation the solution degenerates very rapidly for small thickness (locking phenomenon). To overcome such behaviour, non-standard formulations of the problem are usually combined with low-order finite elements to weaken or possibly eliminate the locking of the numerical solution.In recent years high-order finite elements have been introduced and successfully applied in several fields. Previously we have constructed a family of hierarchic high-order finite elements to solve the Reissner-Mindlin problem in its plain formulation. The locking was strongly reduced but was still active for very thin plates. Meanwhile some mixed-interpolated finite elements have been developed and shown to be locking free. In this paper we combine the two approaches, namely, the hierarchic high-order elements and the mixed-interpolated elements. The result is a family of special finite elements that exhibits both properties of convergence and robustness.

Notes: The paper reveals the conditions under which the approximations of the normal vector and global coordinates of interior points of two segments on a smooth contour can fail using the C1-continuous Overhauser elements. A new continuous modelling of the normal vector is developed using the standard Lagrange elements. The derivation is consistent with the assumptions employed in the boundary element method formulations of potential problems. Test numerical examples are considered in order to analyse the accuracy of the approximations based on the use of standard C0-continuous elements, Overhauser C1-continuous elements and the proposed ‘smooth’ C0 elements.

Notes: The paper describes a new numerical algorithm for pure advection at large Courant numbers. The DISCUS scheme uses an accurate and proven control volume algorithm (QUICKEST) within a Lagrangian treatment of advection. By being in sympathy with the information flow dictated by characteristic lines, the scheme performs exceedingly well at Courant numbers greater than one. In contrast to existing Lagrangian schemes which use the characteristics to determine the grid points from which concentrations are interpolated, in DISCUS the characteristics are used to determine the appropriate location for the control volume. The scheme is of comparable accuracy to Eulerian schemes at Courant numbers less than one, but it increasingly outperforms them as the Courant number increases. Indeed, the results reported illustrate the futility of using an Eulerian scheme at high Courant numbers.

Notes: The use of effective core potentials in the calculation of the geometrical parameters of the ferrocene molecule and its heavier analogs is reported. It is shown that a critical factor in these calculations is the efective core-core (ECC) potential and, in the absence of ECCs for first-row atoms that are involved in short bonds, calculations of the geometrical parameters are not reliable. Good agreement with experimental geometries may be obtained by using the Los Alamos ECPs for atoms of the second and higher rows of the periodic table at the MP2 level. DFT calculations have been performed and found to give numerical results comparable to MP2 in the same basis. © 1995 John Wiley & Sons, Inc.

Notes: In this article, we report an efficient computational procedure for electron scattering matrix elements in the previously developed cubic-grid Gaussian basis sets. The Green function matrix elements derived for the cubic-grid basis set are simpler and easier to calculate than are those available in the literature for conventional Gaussian basis sets. Special features of the cubic-grid basis sets may also be exploited for a very efficient computation of Coulomb and exchange integrals. Inelastic scattering amplitudes for vibrational excitations may be efficiently calcualted in the harmonic approximation by numerical differention of the T-matrix elements. © 1995 John Wiley & Sons, Inc.

Notes: Previously developed cubic-grid basis sets of various sizes were used for the calculation of cross sections for elastic and inelastic electron scattering by the He and Ne atoms and the H2O molecule by the T-matrix expansion method. The aim was to test the invariance of calculated cross sections with respect to the translation of the target molecule and to examine the effect of basis-set size on the results. We also present a simple procedure for accounting for long-range interactions from the part of space that lies outside the volume that contains the cubic-grid basis set. © 1995 John Wiley & Sons, Inc.

Notes: Atomic charge and momentum densities of 91 atoms (He—U) are classified in terms of their L1, L2, and entropic measures of distance from the densities of the preceding atoms. The relationship between these distances and the first ionization energies is also considered. © 1995 John Wiley & Sons, Inc.

Notes: The subject of this article is the self-consistent-field (SCF) treatment of low-lying molecular vibrations in molecules subject to solvent effects and light atom migration. The analyses use a Cartesian Gaussian basis and Gaussian functional expansions of potential energy operators. The objective of the work was to establish approximate and practical methods of analysis of vibrational degrees of freedom in molecules that build on and compare well with the highly accurate treatments of vibrations in small molecular systems of the past decade. An application to a system in which hydrogen bonding contributes the major anharmonic effect illustrates the method. © 1995 John Wiley & Sons, Inc.

Notes: This article presents methods for computing matrix elements with Cartesian Gaussian wave functions of potential energy operators that depend on functions of the form (r-r0)n exp[-a(r - r0)] as well as matrix elements of the class of polynomial many-body potentials developed by Murrell et al. The matrix elements arise in the analyses of anharmonic vibrations in molecules. © 1995 John Wiley & Sons, Inc.

Notes: The extension of the PPP Hamiltonian for alternant cyclic polyenes to noninteger values of the pseudomomentum by imposing modified boundary conditions is discussed in detail. It is shown that a computer program for periodic boundary conditions can be easily adapted to the new boundary conditions. Full CI computations are carried out for some low-lying states of the PPP model of altemant cyclic polyenes (CH)N, (N even) at half-filling. The energy values obtained by using periodic (Bloch) and antiperiodic (Möbius) orbitals are used to perform energy extrapolations for N → ∞. © 1995 John Wiley & Sons, Inc.

Notes: General formalism for evaluation of multiparticle integrals involving J̌2 and J̌z operators over explicitly correlated Cartesian Gaussian functions is presented. The integrals are expressed in terms of the general overlap integrals. An explicitly correlated Cartesian Gaussian function is a product of spherical orbital Gaussian functions, powers of the Cartesian coordinates of the particle, and exponential Gaussian factors, which depend on interparticular distances. This development is relevant to both adiabatic and nonadiabatic calculations of energy and properties of multiparticle systems. © 1995 John Wiley & Sons, Inc.

Notes: The molecular g-tensors for the molecules NO2, CO2-, C3H5, H2CO+, and NF3+ are calculated at the unrestricted Hartree-Fock (UHF) level using the Rayleigh-Schrödinger perturbation approach. All one-electron terms have been evaluated, including the relativistic mass correction, the one-electron spin-Zeeman gauge correction, and a second-order term involving the spin-orbit coupling and the orbital-Zeeman interaction. The relative importance of the first-order terms is found to be quite substantial. For first-row molecules, their total contribution is often of the same order of magnitude as of the second-order contribution, in some instances even larger. A study on the basis-set dependence of these g-tensors has been conducted. At this level, the most important basis-set criterion is shown to be a satisfactory description of the valence regions. Finally, the gauge dependence of these g-tensor calculations has been systematically studied. This dependence seems to be substantially reduced for basis sets which include polarization functions. © 1995 John Wiley & Sons, Inc.

Notes: A scaling procedure is used for the force constants generated by the SINDO1 method in internal coordinates to achieve better agreement with experimental frequencies of molecules. The procedure is subsequently used to calculate frequency shifts for adsorbed molecules. The results for CO2 and CO adsorption at NaCl cluster surfaces are in good agreement with experimental data. © 1995 John Wiley & Sons, Inc.

Notes: Ab initio self-consistent-field calculations are reported for electronic states of beryllium clusters comprised of 93, 105, 111, and 123 atoms. The respective clusters correspond to coordination shells 12-15 of a central Be atom with internuclear separations derived from the lattice constants of the bulk metal. Ab initio effective core potentials have been employed to replace the 1 s electrons, thereby reducing the complexity of the calculations. In addition, use of the full D3h point group symmetry of the clusters results in a substantial reduction of the numbers of two-electron integrals that must be computed and processed. Binding energies, orbital energies, electric field gradient, nuclear-electrostatic potential, diamagnetic shielding constant, second moments, and Mulliken populations are calculated for selected electronic states. Calculated binding energies when compared among the different clusters as well as to smaller and larger fragments from earlier studies provide evidence for the onset of convergence to the Hartree-Fock limit of the bulk. Lowest-state ionization potentials are consistently above and agree to within 14% of the experimental workfunction. The net charge on the central beryllium atom decreases toward zero. The variability of observed bulklike behavior for the different properties indicates that the transition between cluster and bulklike behavior is not sharp and depends on the quantity of interest. © 1995 John Wiley & Sons, Inc.

Notes: The discrete variable representation method (DVR) is applied to the calculation of the J = 0 vibrational energy levels of the ground electronic state of nitrogen dioxide, a molecule which shows a large amplitude bending vibration. The Hamiltonian is expressed in Johnson hyperspherical coordinates and developed on a DVR basis set for each coordinate. A successive diagonalization-truncation method is applied which gives accurate values for the energy levels up to ≃ 7000 cm-1. © 1995 John Wiley & Sons, Inc.

Notes: A procedure for finding transition states ( TS) that does not require the evaluation of the second derivatives during the search is proposed. The procedure is based on connecting a series of points representing products Pi and reactants Ri and taking conservative steps along the difference vector from Pi toward Ri and from Ri toward Pi until the two points coalesce. Although the points Po and Ro represent the product and reactant specifically, other Pi and Ri are determined by minimization in hyperplanes perpendicular to Pi -1 and Ri-1. We apply this technique to six well-known potential functions and compare these results with those obtained from other well-known procedures. © 1995 John Wiley & Sons, Inc.

Notes: Potential energy curves for the ground and low-lying excited states of the AH2+ (A = Mg - Ar) dications have been calculated using high-level ab initio methods with large atomic orbital basis sets. Quasi-bound potential energy curves with local minima and deprotonation barriers have been found for most of the dications studied. The energies, tunneling lifetimes, and widths of the quasi-bound states have been calculated by numerical solution of the radial Schrodinger equation using the Numeov method. All these dications except ArH2+ have low-lying states which support quasi-bound vibrational states. The ArH2+ dication has a 2∏i potential energy curve with a minimum so shallow that it does not support any quasi-bound vibrational states. Results of our calculations are compared with previous ab initio calculations and available experimental data. © 1995 John Wiley & Sons, Inc.

Notes: A bond-energy formula deduced by means of the Hellmann-Feynman theorem is applied to selected simple hydrocarbons. The required potentials at the nuclei are calculated with the help of large basis-set expansions including polarization functions. The carbon-carbon bond energy of ethane is evaluated at ∼ 70 kcal mol-1. The CC bond energies of ethane, ethylene, acetylene, benzene, and cyclopropane are approximately in a ratio of 1: 2.0: 3.0: 1.65 1.0. Limitations and possible improvements in future applications of this energy formula are discussed. © 1995 John Wiley & Sons, Inc.

Notes: Ab initio molecular orbital (MO) calculations with the 3-21G and 6-31G basis sets are carried out on a series of complexes of NH3 with Li+, C≡N-, LiCN, and its isomer LiNC. The BSSE-corrected interaction energies, geometrical parameters, internal force constants, and harmonic vibrational frequencies are evaluated for 15 species. Complexes with trifurcated (C3v) structures are calculated to be saddle points on the potential energy surfaces and have one imaginary frequency each. Calculated energies, geometrical parameters, internal force constants, and harmonic vibrational frequencies of the various species considered are discussed in terms of the nature of association of LiCN with ammonia. The vibrational frequencies of the relevant complexed species are compared with the experimental frequencies reported earlier for solutions of lithium cyanide in liquid ammonia. © 1995 John Wiley & Sons, Inc.

Notes: When two electronic potentials present an avoided crossing, the adiabatic approximation breaks down in the energy region near to the crossing. In particular, the correspondence between exact energy levels of the two-state system and the adiabatic levels of the lower and upper adiabatic potentials becomes ambiguous. This implies that the term “nonadiabatic effect,” used for the difference between exact and adiabatic energy eigenvalues, loses its meaning in the crossing regime unless an unambiguous way of assigning an adiabatic to an exact level is defined. This is important in order to investigate where nonadiabatic schemes, such as the generator coordinate approximation, fit in between the adiabatic approximation and quasi-exact approaches. © 1995 John Wiley & Sons, Inc.

Notes: The application of the Σ-separation method to the calculation of multicenter two-electron molecular integrals with Slater-type basis functions is reported. The approach is based on the approximation of a scalar component of the two-center atomic density by a two-center expansion over Slater-type functions. A least-squares fit was used to determine the coefficients of the expansion. The angular multipliers of the atomic density were treated exactly. It is shown that this approach can serve as a sufficiently accurate and fast algorithm for the calculation of multicenter two-electron molecular integrals with Slater-type basis functions. © 1995 John Wiley & Sons, Inc.

Notes: The intersystem crossing (ISC) between the lowest triplet and singlet states occurring in the reaction of atomic oxygen with ethylene was studied. The importance of spin-orbit coupling (SOC) in oxirane biradicals (ĊR′R″ - CRR* - Ȯ) is stressed through calculations where the spin-orbit matrix elements over the full Breit-Pauli SOC operator has been obtained in the singlet-triplet crossing region. The calculations are performed with a multiconfigurational linear response approach, in which the spin-orbit couplings are obtained from triplet response functions using differently correlated singlet-reference-state wave functions. Computational results confirm earlier semiempirical predictions of the spin-orbit coupling as an important mechanism behind the ring opening of oxiranes and addition of oxygen O(3P) atoms to alkenes. © 1995 John Wiley & Sons, Inc.

Notes: We consider a crystal as partitioned into a localized molecular cluster (containing a defect or not) and an embedding region. Within the Hartree-Fock formalism, an expression is derived for an effective potential due to the embedding region of crystal. This potential is part of the cluster Fock operator and requires input from a perfect crystal calculation. Special features of the derivative are rigorous inclusion of cluster-embedding overlap and orthogonality between single-electron states of the embedding region and the function-space manifold of the cluster; physically correct normalization of the Fock eigenstates; and a nontrivial total-energy algorithm. Computational requirements are qualitatively compared with those for an isolated cluster. The method allows for intracluster (and intraembedding) correlation and can be adapted straightforwardly to local density functional approaches. Fundamental aspects of the embedding problem are addressed in a general formulation that is, nevertheless, oriented toward explicit calculations. © 1995 John Wiley & Sons, Inc.

Notes: An efficient procedure for third-order electron propagator calculations of ionization energies and electron affinities is reported. Diagonal self-energy expressions that are suitable for large molecules are empolyed. The outer-valence Green's function method also is implemented. An integral transformation program for direct and semidirect algorithms is modified to store only nonzero integrals according to Abelian point group symmetry. Contributions to self-energy matrix elements that depend on electron repulsion integrals with four virtual orbital indices are computed in a direct way. Intermediate batches of integrals are created by sort procedures while avoiding storage of transformed integrals in the main memory. This method permits calculation of electron binding energies for C72- with a 231 atomic orbital basis and for Zn(C5H5)2 with a 220 atomic orbital basis on an IBM RISC/6000 Model 550. During these calculations, the CPU is engaged approximately 90% of the time. © 1995 John Wiley & Sons, Inc.

Notes: The intermolecular potentials for the X 2σ and A2Π states of Li… Ar were studied by a variety of multiconfiguration, single-configuration, and perturbation methods (CASPT2). The A 2Π excited state was calculated to have a well depth of 811 cm-1 at an internuclear separation of 2.59 Å, in excellent agreement with the 810 cm-1 derived from experimental data. A smaller well of 77 cm-1 was found for the X 2σ ground state at an intermolecular separation of 4.8 Å. These results are in better agreement with experimental results than were the previously reported pseudopotential calculations. The comparison of CI calculation with the CAPST2 results shows that the latter is able to give good results for interacting metal-rare gas systems. © 1995 John Wiley & Sons, Inc.

Notes: This article is a brief review of the work formalism of electronic structure, its recent developments, and the results of its application to spherically symmetric and nonspherical density atoms. The formalism, which is founded in Schrödinger theory, is derived by physical arguments based on Coulomb's law. The fundamental quantity in the formalism is the pair-correlation density that constitutes the nonlocal quantum-mechanical source charge distribution giving rise to both a local potential representing electron correlations as well as the electron interaction energy. The potential is the work done to move an electron in the force field of the pair-correlation density and the energy the interaction energy between the electronic and pair-correlation densities. (For systems for which the curl of the force field may not vanish, the potential is obtained from the irrotational component of the field, the solenoidal component being neglected). The differential equation governing the system is a Sturm-Liouville equation, and as such, the exact wave function can in principle be obtained as an infinite linear combination of Slater determinants of the self-consistently determined spin-orbitals of the occupied and virtual states. The correctness of the interpretation for the local potential representing electron interaction is evidenced as follows: In the Pauli-correlated and central field approximations, ground-state energies of atoms (2 He -86Rn) lie within 50 ppm of those of Hartree-Fock theory, differing by less than 10 ppm for atoms with Z 〉 35. The densities thus generated clearly exhibit atomic shell structure and also satisfy the Kato-Steiner electron-nucleus cusp condition to 2 ppm. Another attribute of the formalism is that the asymptotic structure of the potential (when both Pauli and Coulomb correlations are considered) is that of the Pauli-correlated approximation. This is rigorously the case as shown for the He atom for which the potential vanishes in the classically forbidden region, the potential there being the exchange potential. As such, it is meaningful to compare the highest occupied eigenvalue of the differential equation in the Pauli-correlated approximation to experiment. A comparison for atoms and atomic ions of this eigenvalue to experimental ionization potentials and electron affinities show them to be consistently superior to the corresponding eigenvalue of Hartree-Fock theory. Transition energies determined from eigenvalue differences are also superior to those obtained from total energy calculations via Hartree-Fock theory when compared to experiment. Further, by considering the carbon atom in one of its degenerate ground states for which the curl of the field due to the Fermi hole does not vanish, it is shown that the solenoidal component of the field is negligible and two orders of magnitude smaller than is the irrotational component. Thus, the approximation of obtaining a path-independent potential for nonspherical density systems from the irrotational component of the field is accurate. Finally, Coulomb correlation effects can be incorporated within the work formalism in practice via the configuration interaction approximation. The self-consistent orbitals thus obtained explicitly incorporate the effects of both Pauli and Coulomb correlations in their structure because the source charge from which they are generated is a pair-correlation density. Furthermore, these orbitals possess the correct asymptotic structure since they are also generated by a potential that is local. The work formalism also provides a physical interpretation for the local potential representing electron correlations of Kohn-Sham density functional theory. Further, the exchange potential of the work formalism satisfies analytically two requisite conditions of the Kohn-Sham theory exchange potential. These are the scaling requirement and the sum rule relating the exchange energy to its functional derivative. The work formalism also leads to a deeper understanding of electron correlations in various approximations within Kohn-Sham theory. For example, it can be rigorously shown that the pair-correlation density in the local density approximation contains a term proportional to the gradient of the density. Thus, in contrast to the Kohn-Sham theory interpretation that electron correlations in this approximation are those of the uniform electron gas assumed valid locally, we learn that the nonuniformity of the electronic density is, in fact, explicitly accounted for by the approximation. This then explains the accuracy of the approximation. © 1995 John Wiley & Sons, Inc.

Notes: The aziridine analogues of the epoxide metabolites of carcinogenic polycyclic aromatic hydrocarbons have greater mutagenic potency than the epoxides. Like their well-studied analogues, the aziridines undergo a pH-dependent decomposition that leads to a reactive carbocation intermediate. In aqueous solution the nucleophile with which the carbocation generally reacts is water. The kinetics of this pH-dependent reaction have been experimentally characterized by others. Although the effect of DNA on this reaction has not been studied, we hypothesize that, like their epoxide analogues, the aziridine derivatives of polycyclic aromatic hydrocarbons undergo a DNA-catalyzed reaction leading through a carbocation to either a DNA-adduct or a hydrolysis product. Using Poisson-Boltzmann calculations in conjunction with Metropolis Monte Carlo simulations and energy-minimized conformations, we predict the DNA-dependence of the acidcatalyzed decomposition of the K-region aziridine, phenanthrene[9,10]imine. © 1995 John Wiley & Sons, Inc.

Notes: A QSAR analysis of a series of Valproic Acid (vpa) derivatives is given, which shows the importance of hydrophobic and electronic effects as determinants of the anticonvulsivant activity. The statistical analysis allows one to infer that the electron acceptor capability of the carboxylic carbon atom may guide electrostatic interactions of the molecules with the receptor site, in those cases where the lipophilic requirements are satisfied. Both the anticonvulsivant activity and the calculated lipophilic parameters (log P values) are taken from the literature, whereas the electronic descriptors result from Intermediate Neglect of Differential Overlap calculations at the Configuration Interaction level, (INDO/S-CI parametrization), for the most stable conformers of each derivative. The protein environment is modeled as a dielectric continuum in a Self-Consistent Reaction Field approach. The conformational analysis is based on AMI calculations. © 1995 John Wiley & Sons, Inc.

Notes: A new form of second-order multireference perturbation theory coupled with finite-field perturbation theory is applied to evaluate some one-electron molecular properties. Several possible definitions of the zeroth-order Hamiltonian are considered and results tested against bench-mark full CI calculations. © 1995 John Wiley & Sons, Inc.

Notes: The correlation present in the nondegenerate ground state of an interacting Fermi system is discussed in terms of reduced density matrices and their cumulant expansion. By generalizing a result obtained for the interacting uniform electron gas (correlation induced exchange-hole narrowing), possible measures of the correlation strength in terms of natural occupation numbers (the eigenvalues of the true one-particle density matrix) are introduced. These quantities-the v-order nonidempotency and the information entropy of the natural occupation numbers-result from the correlated many-body wave function and characterize the ground-state correlation in addition to the usual correlation energy. The uniform electron gas serves as a first illustrative example. © 1995 John Wiley & Sons, Inc.

Notes: In the present work, we reexamined the gradient expansion of the exchange energy of an electron gas with a slowly varying charge density. We stay within the exchange-only approximation of Sharp, Horton, Talman, and Shadwick but go to second order in the deviation from the homogeneous limit. The coefficient of the lowest-order gradient correction is obtained analytically both for a bare and a screened Coulomb interaction - the former yielding the value previously obtained by Kleinman numerically and by Engel and Vosko analytically. A screened Coulomb interaction gives Sham's coefficient in the limit of infinite screening length. The cause of the difference between the coefficients of Kleinman and Sham is clearly exhibited. The coefficients of the two next highest-order gradient corrections, one of which originates in second-order response theory, is shown to diverge as the screening length becomes large. The bare Coulomb interaction gives finite coefficients of which the one originating from linear response is obtained analytically and differs from the presumably correct result obtained by Engel and Vosko. This discrepancy demonstrates the extreme sensitivity of the analytical expressions to different regularization procedures. We suggest that coefficients should rather be chosen according to the performance of the resulting gradient approximations in weakly perturbed electron gases. © 1995 John Wiley & Sons, Inc.

Notes: The focal point of the present work is the single-particle kinetic energy density tensor in D dimensions. This quantity enters both differential and various integral forms of the virial theorems, which are again set up in D dimensions. Major new results lie in (i) demonstrating that, by one-dimensional quadrature, it is possible to construct the Pauli potential directly from the kinetic energy tensor, without the need for functional differentiation and (ii) generating the gradient expansion for the kinetic energy tensor, in D dimensions. © 1995 John Wiley & Sons, Inc.

Notes: The geometries of the two most important tautomeric forms of adenine and the corresponding methyladenines are investigated by means of quantum chemical calculations at different level of sophistication, ranging from semiempirical methods to correlation corrected ab initio methods at second-order Møller-Plesset perturbation theory level (MP2). The relative stability of the N(7)H ↔ N(9)H tautomeric forms of adenine are investigated with highly correlation corrected methods, MP3 and MP4. The relative stability is also corrected for solvent interactions and compared with experimental information. N(9)H-adenine is predicted to be the most stable tautomer in both vacuum and in solution. The relative stability is predicted to be between 24.5 and 35.0 kJ/mol in vacuum depending on computation method. In water solution N(7)H-adenine is stabilized more by the solvent, and the corresponding relative energies were found to be between 4.9 and 10.2 kJ/mol. We also found that correlation effects are essential to describe the ground state geometry with a high accuracy. The geometries predicted by semiempirical methods and ab initio calculations without correlation correction show large deviations in some parts of the molecule compared to the MP2 results as well as compared with experimental geometries. © 1995 John Wiley & Sons, Inc.

Notes: Ab Initio crystal orbital calculations on three-dimensional crystals/crystallohydrates of a number of diprotonated mononucleotides have been performed using the CRYSTAL92 routine package. The present results help to gain a deeper insight into the physical mechanisms of nucleic acid semiconductivity, as well as into the essence of intermolecular interactions and solvent effects in solid-state samples of nucleic acids. © 1995 John Wiley & Sons, Inc.

Notes: Experimentally measured rates for the oxidation of p-substituted benzyl amines by bovine monoamine oxidase type B (MAO-B) derived from the literature were examined with respect to the effects of molecular (semiempirically (AM1) derived) electronic, steric, and lipophilicity parameters. These properties included vertical and adiabatic ionization potential, LUMO energy, the LUMO-HOMO difference, molecular hardness, absolute electronegativity, calculated log P values, molecular volume, surface area, and ovality. Substrate oxidation rates (log kcat/Km) were found to correlate with molecular ovality and vertical ionization potential while the rate of enzymatic (flavin) reduction associated with substrate oxidation (log kred) was described by a two-parameter model containing an ovality and an absolute electronegativity term. These results are consistent with an initial one-electron substrate oxidation mechanism. In previous work, use of classical Hansch analysis suggested that electronic terms were not important in the enzymatic reactions. This discrepancy may be related to nontransferability inherent in fragment approaches which assume that the substituent of interest behaves similarly in all molecular scaffolds. Analysis of substrate binding (log Kd) to the enzyme was described by a two-parameter model containing a calculated log P term as well as LUMO energy. The significant correlation found with LUMO energy is consistent with studies suggesting that this property is important for drug-receptor interactions. © 1995 John Wiley & Sons, Inc.

Notes: The integrated molecular transform (FTm) is a unitary structure index that has been successfully used for the correlation of 2- and 3-dimensional structure representations with their physicochemical and pharmacological properties. In the present instance the reported pKa values in a series of 30 compounds consisting of five subseries were correlated with their FTm indices. The omission of four outliers gave a moderate correlation across the entire series; within the individual subseries the correlations were considerably improved even with inclusion of the outliers. In general, the logarithmic transform of the pKa did not improve the correlations. This method gives a relatively simple means of estimating pKa in several structure classes. © 1995 John Wiley & Sons, Inc.

Notes: We investigated Raman and infrared spectra of the Watson-Crick type of the guanine cytosine base pair and of the individual guanine and cytosine nucleic acid bases by ab initio Hartree-Fock theory using the 6-31G* basis set. IR and Raman intensities and Raman depolarization ratios were predicted using the double-harmonic approximation. The effects of a polar solvent were modeled by the self-consistent reaction field (SCRF) approximation. Variations in geometries, harmonic force constants, and vibrational spectra of the studied nucleobases due to the specific hydrogen-bonding interactions are discussed. © 1995 John Wiley & Sons, Inc.

Notes: We present here a perturbative analysis of the coupled cluster response method for molecular static properties with Euler and extended coupled cluster functionals under cubic truncation. Comparative analysis is meant to cater to both pedagogical and practical interests. Comprehensive tables for energy-derivative expressions and equations at the stationary point are presented. © 1995 John Wiley & Sons, Inc.

Notes: The ground-state potential energy curve of the F2 molecule as well as spectroscopic constants were calculated by means of the second-order quasi-degenerate many-body perturbation theory within a full (eight) valence orbital space using a DZP basis set. The problem encountered with a large number of valence electrons is avoided by a proper redefinition of the Fermi vacuum. A comparison with other related multireference techniques is also provided. © 1995 John Wiley & Sons, Inc.

Notes: Correlation holes of electrons with the same (Fermi hole) and different (Coulomb hole) spins in the ground (X1Σ+), first (A1Σ+) and second (B1II) excited states of LiH were constructed from full configuration interaction (CI) wave functions. It was found that the shapes of both the Fermi and Coulomb holes in these states are dependent on the location of the reference electron. When the reference electron is chosen to be close to the Li nucleus, the Fermi correlation results in a large negative hole for all three states. However, the A1Σ+ excited state is further characterized by displaying a second hole around the H nucleus, and in the B1II state, the hole is elongated along the molecular axis. Coulomb correlation shows up strongly in the A1Σ+ state and, in addition, there is clearly correlation of electrons at the two nuclei. These features of the correlation holes were compared with those from a two-Slater-determinant model wave function. The Hartree, Fermi, and Coulomb screening potentials in these states were also studied in the light of possible modeling of the correlation functionals for the excited states. © 1995 John Wiley & Sons, Inc.

Notes: The link between a uniform scaling of the electron density and a scaling of the electron-electron interaction is reviewed. The effective potential along the coupling constant path which connects a noninteracting and a fully interacting system with the same electron density is considered. The effective potential for an arbitrary coupling constant is here expressed in terms of the exchange-correlation potential at a coupling constant of unity. The effective potential is then investigated for ionization processes. Use of the fact that the ionization energy is determined by the exponential decay of the electron density allows us to derive new formulas for the ionization energy. Based on the Taylor expansion of the effective potential along the coupling constant path, a density functional perturbation theory is introduced which leads to a formally exact Kohn-Sham KS formalism. To first order, this formalism gives identities for the exchange potential in terms of KS orbitals and orbital eigenvalues. Moreover, higher-order terms give identities for the correlation potential as well as for the exchange potential. These identities are pointwise as well as integrated. Hence, various new requirements for the exchange and correlation functionals are derived. New insight into the optimized effective potential method is gained by discussing it in the light of the results obtained here. © 1995 John Wiley & Sons, Inc.

Notes: Available approximations to Exc[{nσ}], the exchange-correlation energy functional of spin-density-functional theory, do not include self-interaction corrections (SIC). This leads to Kohn-Sham (KS) potentials, Vxcσ, that fail to satisfy some important analytic properties known to be exhibited by the exact potential. To resolve these difficulties, we consider a KS theory for orbital-dependent exchange-correlation energy functionals that explicitly includes SIC. Recent work by Krieger, Li, and lafrate (KLI), which considers the analytic properties of the spin-polarized optimized effective potentials (OEP), Vxcσ0, i.e., the KS potentials corresponding to Exc = Exc[{φiσ,}], is reviewed as well as the properties of VxcσKLI, an easily calculated approximation to the exact result which, unlike Vxcσ0, can be employed for systems of arbitrary symmetry. In addition, we compare the results of the exact and approximate OEP calculation of the properties of the ground state of atoms and singly charged negative ions in the exchange-only case in which Ex = ExHF [{φiσ}] where HF = Hartree-Fock. We conclude that VxσKLI maintains most of the important analytic properties of Vxσ0, and provides an excellent numerical approximation to the exact result. We also give detailed consideration to the calculation of the ionization potential, I, and the electron affinity, A, in the exchange-only approximation for atoms with Z ≤ 20. We find that the KLI results for both I and A are always within 0.1 milli-au of the exact KS results, whereas both the local spin density (LSD) approximation and the Becke exchange only energy functional lead to deviations which on average are two orders of magnitude larger and significantly exceed the criterion for quantum chemistry accuracy. Finally, using the KLI method for orbital dependent Exc, we compare the KS results for I and A for Z ≤ 20 with the experimental values by employing various approximations for Exc[{φiσ] including: (1) HF exchange with LSD correlation with SIC, (2) LSD approximation for exchange and correlation with SIC, (3) the conventional spin density LSD approximation, and (4) the Becke exchange-energy functional with LSD correlation. In addition, we examine how closely the ionization theorem for ∊m, the energy eigenvalue for the highest occupied orbital, is satisfied in these approximations. © 1995 John Wiley & Sons, Inc.

Notes: First-principles local density functional approximation (LDA) calculations using the linear-response method are more efficient than are traditional supercell or dielectric matrix techniques for determining phonon spectra and dielectric constants. We have implemented this approach using a linearized-augmented-plane-wave (LAPW) basis, which facilitates such calculations for systems containing transition-metal and other atoms with localized orbitals. The accuracy of the method is illustrated with applications to some semiconductors and ferroelectrics. Theoretical work on the perovskite ferroelectrics has focused on the possible roles of disorder and soft-phonon behavior. A complete mapping in the Brillouin zone (BZ) of the ferroelectric instability of KNbO3 has been carried out, revealing a pronounced two-dimensional character. In real space, this instability corresponds to chains oriented along 〈100〉 directions, of displaced Nb atoms. Such instabilities are discussed in relation to the static chain structures in the eight-site order-disorder model introduced by Comes et al. © 1995 John Wiley & Sons, Inc.

Notes: Electrostatic and dispersion contributions of solute-solvent interactions have been considered in the self-consistent reaction field scheme and implemented in the LCGTO-DF (linear combination of Gaussian-type orbitals-density functional) method. Results for the tautomeric equilibrium of formamide-formamidic acid, for the cis-trans energy difference in dichlorodiammineplatinum(II), and for H2O—HF hydrogen-bond systems are in agreement with the available experimental and previous high-level ab initio data. The role of the dispersion energy is discussed for the different studied systems. © 1995 John Wiley & Sons, Inc.

Notes: Local (LSD) and nonlocal (NLSD) spin density calculations using different exchangecorrelation functionals have been performed to determine equilibrium geometries, harmonic vibrational frequencies (ωe), ionization potentials (IP), electron affinities (EA), dipole moments (μ), and singlet-triplet energy gaps (Δ EST) of SiH2, GeH2, and SnH2. Geometrical structures as well as vibrational frequencies are in agreement with the available experimental data and compare favorably with the most sophisticated postHartree-Fock computations performed until now. Both computed IPS (9.15 and 9.25 eV for SiH2 and GeH2, respectively) and EA of SiH2 (1.17 eV) compare favorably with experimental data (9.17, 9.21, and 1.2 eV). Accurate values are obtained also for singlet-triplet energy gaps. We report for the first time the electron affinities of all neutral systems and the spectroscopic constants of the cations and anions. © 1995 John Wiley & Sons, Inc.

Notes: The coupled clusters singles and doubles (CCSD) method for calculations of open-shell systems with the single restricted Hartree-Fock (ROHF) reference determinant is extended by the noniterative triples to give CCSD(T). Our approach profits from the fact that (a) single- and double-excitation amplitudes are spin-adapted, which directly leads to a computationally less demanding algorithm than are nonadapted procedures and produces the spin-adapted CCSD wave function and (b) triple excitations calculated from converged spin-adapted (SA) CCSD amplitudes are also obtained more effectively. Altogether, computer demands of our SA CCSD(T) approach, applicable to high-spin open-shell cases which are well represented by a single-determinant reference is comparable to that for closed-shell systems. Our approach is not based on semicanonical orbitals, applied by Bartlett's group. However, we compare some other possible choices of ROHF orbitals to this “standard.” Numerical results for a series of atoms and molecules demonstrate little sensitivity to this selection. © 1995 John Wiley & Sons, Inc.

Notes: Ab initio accurate all-electron relativistic molecular orbital Dirac-Fock self-consistent field calculations are reported for the linear symmetric XeF2 molecule at various internuclear distances with our recently developed relativistic universal Gaussian basis set. The nonrelativistic limit Hartree-Fock calculations were also performed for XeF2 at various internuclear distances. The relativistic correction to the electronic energy of XeF2 was calculated as ∼ -215 hartrees (-5850 eV) by using the Dirac-Fock method. The dominant magnetic part of the Breit interaction correction to the nonrelativistic interelectron Coulomb repulsion was included in our calculations by both the Dirac-Fock-Breit self-consistent field and perturbation methods. The calculated Breit correction is ∼6.5 hartrees (177 eV) for XeF2. The relativistic Dirac-Fock as well as the nonrelativistic HF wave functions predict XeF2 to be unbound, due to neglect of electron correlation effects. These effects were incorporated for XeF2 by using various ab initio post Hartree-Fock methods. The calculated dissociation energy obtained using the MP2(full) method with our extensive basis set of 313 primitive Gaussians that included d and f polarization functions on Xe and F is 2.77 eV, whereas the experimental dissociation energy is 2.78 eV. The calculated correlation energy is ∼ -2 hartrees (-54 eV) at the predicted internuclear distance of 1.986 Å, which is in excellent agreement with the experimental Xe - F distance of 1.979 Å in XeF2. In summary, electron correlation effects must be included in accurate ab initio calculations since it has been shown here that their inclusion is crucial for obtaining theoretical dissociation energy (De) close to experimental value for XeF2. Furthermore, relativistic effects have been shown to make an extremely significant contribution to the total energy and orbital binding energies of XeF2. © 1995 John Wiley & Sons, Inc.

Notes: Based on the EHMO approach, the band structures for the Y—Ba—Cu—O superconductors doped by La were calculated. The influence of the partial substitutions of La for Y and Ba in YBa2CU3Oy on its electronic structures was investigated. The results demonstrate that the La doping at the Ba site has a great effect on the electronic structures of the Y—Ba—Cu—O superconductors, whereas the change in the band structures caused by the La doping at the Y site is very small. The increase in the oxygen content caused by the La doping results in an increase in the densities of states at Ef, N(Ef), for La1+x Ba2-xCu3Oy, but the increase in N(Ef) cannot compensate the decrease caused by the La doping at the Ba site. In addition, the 2D Cu—O planes are much more sensitive to the change in N(Ef) than are the 1D Cu—O ribbons, which implies an important role of the 2D Cu—0 planes in the Y—Ba—Cu—O superconducting system, regardless of whether La substitutes for Y or for Ba. © 1995 John Wiley & Sons, Inc.

Notes: The proof of Parr and Chattaraj for the Principle of Maximum Chemical Hardness is very general and can be applied to many more observables. The case of the physical, or mechanical, hardness of a solid is taken as an example. It is shown that this also should be a maximum in an equilibrium system. Assuming the validity of the argument leads to new information about the compressibility of solids. A simple expression is also given for the Gruneisen constant. These maximum principles are necessary consequences of the equilibrium conditions of quantum mechanics and thermodynamics. © 1995 John Wiley & Sons, Inc.

Notes: The field theoretical background of relativistic density functional theory is emphasized and its consequences for relativistic Kohn-Sham equations are shown. The local density approximation for the exchange energy functional is reviewed and the importance of relativistic corrections for an accurate representation of the exchange functional is demonstrated. © 1995 John Wiley & Sons, Inc.

Notes: The dynamics of Jahn-Teller systems has recently been discussed in terms of generalized electronic charge and current densities in nuclear-coordinate space. The introduction of the electronic phase as a function of both electronic and nuclear coordinates, in addition to the electronic density, was a crucial component of this formulation. Here, a densitybased treatment of Born couplings is derived from first-principles quantum mechanics beyond the Born-Oppenheimer approximation. Because of the degenerate electronic configuration of a Jahn-Teller molecule, there are an infinite number of ways in which the charge distribution can be oriented for the same energy, leading to a vanishing bond hardness for the molecule in the symmetric nuclear configuration. Further, the moving nuclear framework serves as the perturbation necessary to define the orientation of the charge density, leading to unhindered rotation of the charge cloud. This leads to the dynamical Jahn-Teller problem, namely, the coupling of electronic and nuclear motions through the Born coupling terms. Applications to superconductivity theory are discussed. © 1995 John Wiley & Sons, Inc.