Configuration-interaction calculations with energy-optimized basis sets and an empirical estimate of all sources of truncation-energy errors are used to evaluate the nonrelativistic electronic energy of the Li ground state. Our result ${\mathit{E}}_{\mathrm{nr}}$=-7.478 062 4(7) a.u. (Li) is 3 μhartree below an upper bound recently obtained by King using Hylleraas-type calculations, and 11 μhartree above an old estimate based on experimental data supplemented by relativistic, radiative, and mass-polarization corrections, thus suggesting a reassessment of several related issues.

We have calculated the bond lengths and bond energies of ${\mathrm{Cu}}_2$, ${\mathrm{Ag}}_2$, and CuAg using nonlocal-density-functional theory with the generalized gradient approximation (GGA), in order to investigate the validity of GGA to metal molecules. In GGA, an exchange-correlation energy density is approximated by a function of electron density ρ(r), and its coordinate differences ∂ρ(r), which are not considered in the local-spin-density approximation (LSDA). We performed fully self-consistent calculations, considering all electrons: the molecular orbitals were expanded by a Slater-type atomic basis, and a nonrelativistic approximation was used. The calculated bond energies of ${\mathrm{Cu}}_2$, ${\mathrm{Ag}}_2$, and CuAg in GGA are 0.4 eV (15%), 0.34 eV (17%), and 0.38 eV (17%) smaller than those results in LSDA, so that GGA removes about 2/3 of the overestimated value of bond energies in LSDA.

The recently proposed local-scaling density-functional theory provides us with a practical method for the direct variational determination of the electron density function ρ(r). The structure of ‘‘orbits,’’ which ensures the one-to-one correspondence between the electron density ρ(r) and the N-electron wave function Ψ({${\mathbf{r}}_{\mathit{k}}$}), is studied in detail. For the realization of the local-scaling density-functional calculations, procedures for intraorbit and interorbit optimizations of the electron density function are proposed. These procedures are numerically illustrated for the helium atom in its ground state at the beyond-Hartree-Fock level.

The method of hyperspherical coordinates is used in obtaining very highly excited two-electron adiabatic potential curves up to the n=12 hydrogenic threshold. The present numerical effort exploits the analytic nature of the solutions to the Schrödinger equation at small and large distances to diagonalize the hyperspherical Hamiltonian in the combined basis set. Numerical deficiencies resulting from a linear dependency of the total nonorthogonal basis set are overcome in an automatic fashion. A high degree of diabaticity is observed which limits the channel interaction to within a select set of hyperspherical channels. The validity of this quasiconstant of motion as reported by H. R. Sadeghpour and C. H. Greene (SG) [Phys. Rev. Lett. 65, 313 (1990)] for ${\mathrm{H}}^{\mathrm{-}}$ is investigated for the photoabsorption of He and higher-z members of the He isoelectronic sequence. A generalized two-electron Rydberg formula capable of predicting the positions of the dominant He (ridge and valley) resonances is given. A new feature of the very-highly-excited resonance structure for ${\mathrm{H}}^{\mathrm{-}}$ and He [observed only recently by Domke et al., Phys. Rev. Lett. 66, 1306 (1991)] is discussed. Semiempirical multichannel-quantum-defect fits are also made to the photodetachment spectra of ${\mathrm{H}}^{\mathrm{-}}$, supporting the conclusions reached in SG.

An alternative method is proposed for the calculation of dynamic polarizability: one that combines the usual development over the excited states with additional terms obtained by multiplying the zeroth-order wave function with a dipole-moment factor (i.e., a polynomial of first degree in the electronic coordinates). The zeroth-order wave function includes correlation effects accounted for by a configuration-interaction perturbation subspace iterative method. The dipole-moment factor makes extrapolation possible if the number of states is limited. Applications are presented for ${\mathrm{Li}}_2$, LiH, and CO. The obtained values are in excellent agreement with the experimental ones. The analysis of the calculations shows that the extrapolation is not necessary for ${\mathrm{Li}}_2$ and LiH, the first excited states of which are of valence character, contributing strongly to the polarizability. In the case of the parallel component of the polarizability of CO, extrapolation is necessary because the first excited state of valence character appears in the seventh ${}_{}{}^1{}_{}{}^{}\mathrm{\Sigma }_{}^{+}{}_{}{}^{}$ state. In addition, we have calculated transition energies and moments for the few low-lying excited states of ${}_{}{}^1{}_{}{}^{}\mathrm{\Sigma }_{}^{+}{}_{}{}^{}$ and ${}_{}{}^1{}_{}{}^{}\mathrm{\Pi }$ symmetry for each molecule.

We have calculated the bound-state sticking fractions of the tdμ and ddμ molecular ions using basis sets of random-tempered Slater geminal functions. These values show that states with the same angular momentum produce sticking fractions that are almost identical. In addition, we have investigated the effect of different particle masses on the sticking fraction and have computed the energy of the very weakly bound tdμ(11) state to an accuracy of 1 μeV.

We consider the contribution of nuclear polarization to the Lamb shift of K- and L-shell electrons in heavy atoms and quasiatoms. Our formal approach is based on the concept of effective photon propagators with nuclear-polarization insertions treating effects of nuclear polarization on the same footing as usual QED radiative corrections. We explicitly derive the modification of the photon propagator for various collective nuclear excitations and calculate the corresponding effective self-energy shift perturbatively. The energy shift of the 1${\mathit{s}}_{1/2}$ state in ${}_{92}{}^{238}{}_{}{}^{}\mathrm{U}$ due to virtual excitation of nuclear rotational states is shown to be a considerable correction for atomic high-precision experiments. In contrast to this, nuclear-polarization effects are of minor importance for Lamb-shift studies in ${}_{82}{}^{208}{}_{}{}^{}\mathrm{Pb}$.

Differential cross sections and generalized oscillator strengths have been measured for two optically allowed transitions 5${\mathit{p}}^6$${(}^1$${\mathit{S}}_0$)→5${\mathit{p}}^5$${(}^2$${\mathit{P}}_{1/2}$)6s and 5${\mathit{p}}^5$${(}^2$${\mathit{P}}_{3/2}$)6s in Xe by means of electron-energy-loss spectroscopy. These measurements are carried out for electron kinetic energies of 100, 400, and 500 eV at small scattering angles (θ=1.4°–14.6°). The generalized oscillator strengths are extrapolated to zero momentum transfer to get the optical oscillator strengths and they are compared with those obtained by optical methods and theoretical calculations. The optical oscillator strengths obtained by this work for the 5${\mathit{p}}^5$${(}^2$${\mathit{P}}_{1/2}$)6s and 5${\mathit{p}}^5$${(}^2$${\mathit{P}}_{3/2}$)6s states are 0.158±0.019 and 0.222±0.027, respectively. Integral cross sections have been also determined for each impact energy. The errors are estimated to be less than 12%.

By starting with the distorted-wave form of the exact transition amplitude and introducing a second distorting potential, a generalized distorted-wave Born theory of electron capture is developed that is applicable to ion-ion collisions. The theory reduces to the distorted-wave Born theory developed by Taulbjerg and Briggs [J. Phys. B 14, 3523 (1983)] in the ion-atom case. Other theories, including the distorted strong-potential Born theory, the boundary-corrected first- and second-order Born theories, and the Coulomb Born theory, are seen to be approximations to the new amplitude.

A coupled-channel optical method for electron–atomic-hydrogen scattering is presented in a form that treats both the projectile and the target electrons symmetrically. Elastic differential cross sections are calculated at a range of energies from 0.5 to 30 eV and are found to be in complete agreement with the absolute measurements of Williams [J. Phys. B 8, 1683 (1975); 8, 2191 (1975)]. Total and total ionization cross sections are also presented.

We have calculated the electron-impact excitation-rate coefficients for the 3p-3d transitions of Ar-like ions with Z=42, 54, and 63. The direct-excitation rates were evaluated with relativistic multiconfiguration target wave functions in a distorted-wave approximation. The contributions from the autoionizing resonances were calculated by the multiconfiguration Dirac-Fock method. For most transitions, autoionizing resonances are found to enhance the rate coefficients by factors of 2–3 for electron temperatures below the temperature at which peak excitation occurs. The enhancement is much less at higher temperatures.

The cross sections for one- and two-photon detachment of ${\mathrm{H}}^{\mathrm{-}}$ with excitation of the degenerate H(2s) and H(2p) levels are calculated within an adiabatic hyperspherical representation. Both the partial cross sections, σ(2s) and σ(2p), and the photoelectron angular distribution asymmetry parameters are obtained. Our one-photon detachment results are dominated by the ${}_{}{}^1{}_{}{}^{}$${\mathit{P}}^{\mathit{o}}$ shape resonance feature above threshold, which is well known from prior theoretical and experimental work, with which our results are compared in detail. Our two-photon detachment cross sections and angular distribution asymmetry parameters exhibit a half-cycle of dipole-field-induced oscillation in the energy region from the H(n=2) threshold to approximately 34 meV above threshold. The occurrence of a half-cycle of oscillation with significant amplitude over a 34-meV energy region above the H(n=2) threshold is due to two circumstances unique to this two-photon process: first, in contrast to the strongly suppressed oscillations in the ${}_{}{}^1{}_{}{}^{}$${\mathit{S}}^{\mathit{e}}$ and ${}_{}{}^1{}_{}{}^{}$${\mathit{P}}^{\mathit{o}}$ attractive dipole-field channels for the H(n=2)-${\mathit{e}}^{\mathrm{-}}$ system, the amplitude of oscillation in the channel having ${}_{}{}^1{}_{}{}^{}$${\mathit{D}}^{\mathit{e}}$ symmetry is not strongly suppressed above threshold; second, electric dipole selection rules for the two-photon process do not permit population of the intense ${}_{}{}^1{}_{}{}^{}$${\mathit{P}}^{\mathit{o}}$ shape resonance located above the H(n=2) threshold, which would otherwise obscure the half-cycle of dipole-field-induced oscillation. For these reasons, the process of two-photon detachment of ${\mathrm{H}}^{\mathrm{-}}$ is a favorable one for observing this dipole-field-induced oscillation above threshold, which has yet to be observed experimentally in any process.

We examine the effect of a molecular hydrogen target, as contrasted with an atomic target, on one-electron-projectile electron loss within the framework of the plane-wave Born approximation. Two different approaches are explored: (1) the use of an approximate molecular wave function due to Weinbaum, and (2) the use of a modified form factor for the two atoms in the ${\mathrm{H}}_2$ molecule. The resulting cross-section expressions are compared with experimental cross sections for electron loss by various one-electron projectiles. We find that for a projectile atomic number larger than 5, molecular effects are negligible in total cross sections and then the electron-loss cross section per molecule is just equal to twice the total cross section per atom.

A quantum-mechanical approximation is developed for ionization of one-electron targets by charged-ion impact. The model is based on the nonrelativistic distorted-wave formalism valid for impact velocities larger than the electron orbital velocity in the initial state. The exact impulse wave function is used to describe the initial state, thus incorporating the projectile potential to all orders. The final state is represented by a product of continuum Coulomb wave functions around both centers, providing the correct asymptotic conditions and the projectile and target cusps. The theory is thought to be valid for large projectile charge, even larger than the ion velocity. The impulse approximation developed here is expensive in computing time, but it is probably one of the few models to deal with high projectile charges. Double-differential cross sections are computable in the forward and backward ejection angles. Comparisons with the experiments in different regions of interest are presented, including the binary sphere, capture to the continuum cusp, ridge electrons, and backward ejection angles. The theory proves to be quite successful, and it does not seem to deteriorate with increasing projectile charge.

Auger electrons emitted from doubly excited states of ${\mathrm{C}}^{4+}$(2l2l’,2l3l’,2l4l’,2l5l’) formed by transfer excitation in collisions of ${\mathrm{C}}^{5+}$(4–10 MeV) with ${\mathrm{H}}_2$ and He were studied using high-resolution zero-degree Auger electron spectroscopy. The KLL, KLM, KLN, and KLO resonant-transfer excitation (RTE) cross sections were measured. The impulse approximation theory with alignment and interference effects taken into consideration was found to be in reasonable agreement with the experimental data. The theory was normalized by a typical factor of 0.65 to obtain the best fits to the data. The RTEA resonance strengths were evaluated to investigate the n dependence (n=2,3,4,5) of RTEA, which is RTE followed by Auger electron emission.

This paper presents calculations of electron-loss cross sections for fast H(1s) projectiles incident on different atomic targets (He, Ne, Ar, Kr, Xe, H, N, or O) for a broad kinetic-energy range (90 keV/amu to 1.4 GeV/amu), and compares the calculations with experimental measurements where possible. The calculations are based on the semiclassical free-collision model. Theoretical differential cross sections for elastic electron scattering, and theoretical or semiempirical differential cross sections for inelastic electron scattering, are used in making the calculations. The results are in good agreement with experimental data. We apply a simple model based on the H(1s) results to obtain single- and double-electron-loss cross sections for ${\mathrm{H}}^{\mathrm{-}}$ projectiles incident on the same atomic targets. Again the results are in good agreement with experiment. Further calculations have yielded results for electron loss of H(1s) incident on molecular ${\mathrm{H}}_2$, ${\mathrm{N}}_2$, and ${\mathrm{O}}_2$ targets.

We have calculated the cross sections for vibrational excitation and dissociative attachment in ${\mathrm{H}}_2$ below 5 eV scattering energy. This completely ab initio calculation uses the frame-transformation method of Greene and Jungen [Adv. At. Mol. Phys. 21, 51 (1985)] for electron-vibronic coupling in resonant scattering from a neutral molecule. We found it necessary to modify their method to obtain good agreement with previous theory and experiment for v=0→${\mathit{v}}_{\mathit{f}}$ with ${\mathit{v}}_{\mathit{f}}$=1–3; for larger ${\mathit{v}}_{\mathit{f}}$ and for dissociative attachment we obtained good agreement with previous theory and qualitative agreement with experiment. The fixed-nuclei phase shifts were derived from a fully ab initio calculation in prolate spheroidal coordinates and then transformed to spherical l=1 phase shifts. The vibrational structure of ${\mathrm{H}}_2^{\mathrm{-}}$ becomes evident for excitation from higher vibrational states of ${\mathrm{H}}_2$ as well as for larger Δv’s, confirming previous theory and experiment.

Using the wave functions of the multichannel-quantum-defect theory, a semiclassical theory for both elastic and inelastic collisional cross sections of Rydberg atoms, in which there can be more than one valence electron, with ground-state rare-gas atoms is obtained. Especially the noble-gas-induced broadening cross sections of Rydberg series in alkaline-earth atoms can be calculated. An application of theory is shown to explain the irregular broadening rates of Sr 5sns ${}_{}{}^1{}_{}{}^{}$${\mathit{S}}_0$ series due to collision with Xe, and a good agreement between theory and experiment is achieved.

Excitation of ${\mathrm{He}}^{+}$ by electron impact is studied in the energy range 4.4–14.71 Ry using a six-state close-coupling expansion together with an optical potential representing higher bound and continuum states. An 18-state basis set emphasizing long-range pseudostates is used for calculations below 100 eV, and a 17-state basis containing more compact pseudostates is used at higher energies. Total excitation cross sections for transitions from the 1s, 2s, and 2p states are also calculated using the optical theorem. A detailed comparison is made with a recent 15-state close-coupling calculation and with experiment, where possible. In general, inclusion of the optical potential significantly reduces the excitation cross sections.

Measurements of the dissociative recombination of deexcited ${\mathrm{D}}_3^{+}$ ions yield cross sections similar at low energies to those found for ${\mathrm{H}}_3^{+}$ [H. Hus et al., Phys. Rev. A 38, 658 (1988)]. Previous measurements [J. B. A. Mitchell et al., J. Phys. B 17, L909 (1984)] performed with excited ions indicated that there was an isotope effect giving rise to a larger cross section for ${\mathrm{H}}_3^{+}$ than for ${\mathrm{D}}_3^{+}$. This effect can in fact be explained in terms of differing vibrational-state populations in those studies rather than by molecular-dynamics arguments. Dissociative excitation measurements for ${\mathrm{D}}_3^{+}$ are found to contain a wealth of resonance structure.