Table of contents

The ¹D-¹S quadrupole transition probabilities for Ca, Sr and Ba have been computed using extended GTO and STO valence basis sets and configuration-interaction wavefunctions that include the important core-valence correlation effects. For Ba and Sr, the relativistic contraction of the core orbitals was accounted for in the GTO calculations by a relativistic effective-core potential. The authors' computed Einstein coefficient for Ca of 39.6 s^-1 is in excellent agreement with the recent experimental value of 40+or-8 s^-1. Their best Einstein coefficients for Sr (44.7 s^-1) and Ba (2.98 s^-1) imply increasing quadrupole line strengths down the column. Relativistic effects substantially increase the quadrupole Einstein coefficient for Ba.

Optical emission from the (2p2p2p) ⁴S⁰ states in three-electron systems has been studied both experimentally and theoretically. The wavelengths and transition rates have been calculated for the (2p2p2p) ⁴S⁰-(1s2p2p) ⁴P transition for the lithium-like ions ranging from Z=2 to Z=9. In foil-excited fast beams this transition has been observed for Li I and Be II at 145.02+or-0.05 AA and 80.85+or-0.03 AA, respectively. The experimental and theoretical data for Li I and Be II agree closely.

A measurement of the 'quasi-Landau resonance' (QLR) spacing of lithium, using an atomic beam in a magnetic field of 5.6 T, has been made for comparison with earlier absorption observations, which reported an anomalous spacing of 0.5 omega _c. Data were obtained using hydrogenic m=-1 states near n=30 in an atomic beam directed axially down a solenoid magnet. In the absence of motional Stark effects the 0.5 omega _c resonance spacing no longer emerges near threshold. Comparisons are made with existing theories for diamagnetic effects.

The author describes Na(3s) to Na(3p) transition induced by collision with Na⁺ via two-state rotational coupling at E_K=37.5 eV in a semiclassical first-order Magnus approximation. Calculations are performed using several different data sets for the potential energies. It is found that recent pseudopotential values in which core polarisation is included reproduce the available experimental results best.

A theoretical consideration of collisional processes for excited atoms within an atomic beam is made. The results are presented of calculations of rate constants for the associative ionisation reaction (AI) when resonantly excited atoms of sodium collide. It was shown that the output of charged particles in AI reactions in atomic beam experiments may greatly differ from that in gas cell conditions due to the inherent features of the distribution function of the beam atoms over their relative velocities. The parameters of the model of the ionisation process are derived from a comparison of experimental data obtained in atomic beam and gas cell conditions.

The complex resonance poles of the S matrix for electron-atom or electron-molecule scattering are calculated directly using separable expansions of the potential (and thus of the T matrix) within the framework of the many body optical potential approach. The performance of the method is demonstrated for the well known 2.3 eV shape resonance in electron-N₂ scattering. The authors present the first direct calculation of this resonance pole at the equilibrium geometry of N₂ including both polarisation and target correlation.

An explicit, second-order accurate and stable finite difference algorithm for the solution of the time-dependent Schrodinger equation in full coordinate space including infinity, based on the leapfrog method, is presented. The problem of monopole polarisation (1+1 dimensions) is used in order to demonstrate the difficulties of the standard bound-state basis expansion method and of the variational method with bound-state trial functions at intermediate collision velocities.

General expressions are presented for the line profile produced by an emitting atom embedded in a sea of randomly distributed identical perturbers, on the assumption that the interaction between the emitter and a perturber is of the type 1/R^p, where p>or=3. The motion of each perturber relative to the emitter is treated classically and the trajectory of the perturber is assumed to be a straight line. Results are given for the case in which all emitter-perturber interactions are assumed to be additive; when the emitter is subject to the linear Stark effect, additional results are obtained that do not depend on the assumption of scalar additivity. The profile can be expressed in terms of the Fourier transform of a 'universal' function of a single dimensionless variable that may be evaluated once and for all, since it does not depend explicitly on any particular parameters associated with the spectral line or with the perturbing gas. The properties of this function are studied in detail, and for p>or=4, it is shown that it may be represented very accurately by a semi-empirical analytic function of relatively simple form. It is also shown that certain features of the line profile can depend appreciably on whether or not the correct average over a Maxwellian distribution of velocities of the perturbers is carried out. The use of the semi-empirical universal functions means that profiles can be generated extremely quickly for a particular case, and these can be used as a starting point for a more detailed assessment of the additional physical effects involved that are not included in the present model. Explicit results are given for the cases p=4 and 5; these are relevant to the analysis of the broadening of atomic and molecular lines by molecules.

Exact solutions to a one-dimensional, time-dependent Schrodinger equation with two colliding delta-function potentials of different strengths are presented. The equation can serve as a model for resonant and quasi-resonant ion-atom collisions. Electrons initially bound to one of the potentials may be transferred to the other potential or to a continuum state and probabilities for these processes are calculated. The model reproduces the damped oscillatory behaviour of the charge-transfer probability seen in quasi-resonant collisions when the collision energy is varied.

Quantum mechanical cross section calculations have been performed in order to interpret the energy dependence of absolute total cross sections for fine-structure transitions in the 4²P level of K, induced in collisions with Ar. The comparison between the experimental and theoretical data allows available interatomic potentials to be tested. The various adiabatic potential sets used in the calculations, along with a careful analysis of the polarisation in the 4²P level of K, allow the coupling mechanisms responsible for the fine-structure transitions in K(4²P)+Ar collisions to be analysed in some detail.

A method of producing an intense beam of H(3p) atoms is described. The 3p hydrogen atoms are formed by laser excitation of metastable H(2s) atoms and efficiencies of the order of 0.3 to 0.4 are easily obtained. The metastables themselves are formed by the charge exchange of protons in a Cs vapour. Cross sections for electron loss from H(3p) atoms and H(2s) atoms in collision with H₂ molecules and rare-gas atoms have been measured in the energy range 1260-3600 eV in the forward direction (10 mrad acceptance). The results are normalised to absolute cross sections obtained by Roussel et al. (1977) for electron loss from ground-state hydrogen atoms in collision with the same targets. Comparison with other measurements is only possible for electron loss from H(2s) atoms and a good agreement with other authors is found.

The capture of an electron from a hydrogenic ion by a fully stripped ion, having atomic numbers Z_A and Z_B respectively, is investigated using the relativistic second Born approximation. By expanding in powers of alpha Z_A and alpha Z_B and retaining the leading terms, alpha being the fine-structure constant, the authors have derived formulae for the scattering amplitude which are used to evaluate differential and total cross sections for electron capture between ground states. A closed formula for the relativistic total capture cross section, with full allowance for spin flips, is derived which tends to the non-relativistic second Born cross section first obtained by Drisco in the limit of non-relativistic but high impact energies. At ultra-high relativistic impact energies E it is found that the relativistic second Born capture cross section decays as E^-1 in accordance with the relativistic OBK approximation but with a reduced coefficient.

The authors have calculated the 1s-1s charge transfer probability for large projectile scattering angles in asymmetric ion-atom collisions, using the strong-potential Born approximation and including the effect of recoil on the projectile states non-perturbatively. A significant angular dependence is found. Numerical results for 1s-1s capture in 0.3-20 MeV proton impact on C and Ne are presented, showing good agreement with recent experimental results.

Making use of a semiclassical (JWKB) method which includes the quantum effect of the scattered waves, charge transfer processes in Li⁺-Li collisions have been investigated theoretically. The electronic wavefunctions of the colliding system were expanded in terms of molecular orbitals with an electron translation factor. Calculations for the differential cross sections of the processes were made in the incident energy range 0.5-1 keV (laboratory system). The general behaviour of the calculated results is much improved compared with the previous impact parameter calculation. Particularly, the relative magnitudes of the largest differential cross section peaks to others are found to be in good agreement with experimental results. These largest peaks are enhanced by the interference effect due to rainbow scattering. For resonant (elastic) charge transfer, Li⁺+Li(2s) to Li(2s)+Li⁺, the agreement is almost satisfactory. However in excitational charge transfer, Li⁺+Li(2s) to Li(2p)+Li⁺, the calculated peak and dip positions of the scattering angles shift a little toward greater scattering angles compared with experimental ones.

Measurements have been made of the Lyman- alpha intensity from Stark-mixed n=2 states of hydrogen, using the electron-photon coincidence technique at an incident electron energy of 350 eV. The data are related to a combination of excitation amplitudes, including a term which depends on the relative phase of S and P amplitudes. Field-free measurements of lambda and R have also been made at this energy. Comparison is made with several theoretical results.

The crossed ion and electron beams technique has been used to make an absolute measurement of the cross section for the ionisation process e+W⁺ to 2e+W²⁺ at energies ranging from below threshold up to 750 eV. The compounded random and systematic errors are estimated to be less than about 5% at energies in excess of 18 eV. The W⁺ ions are probably mainly in the ⁶D ground state but with a significant fraction in the ⁴D metastable state. Other metastable ions may account for a few per cent of the total. Comparison of the present data with the scaled Born prediction of McGuire taking into account both 6s outer electrons and 5d inner-shell electrons shows quite good agreement above about 50 eV, but the theory seriously underestimates the cross section at lower energies. The measured ionisation threshold for the ground-state ion is 16.1+or-1.0 eV, compared with the 17.7 eV given by Finkelnburg and Humbach (1955) from a study of screening constants. There are no other known measurements of the ionisation threshold.

The angular behaviour of the differential cross sections for vibrational excitation of N₂O by electron impact has been measured in the 1.4-3.1 eV energy range. These observations have been interpreted using angular correlation theory and lead one to propose that both Sigma and Pi resonances are present, each with only one contributing partial wave. Above 2.4 eV the Pi contribution would be twice that of Sigma and below 1.8 eV only the Sigma state would be present. These conclusions have been shown to be consistent with the angular distributions for O^- formation measured previously where a reinterpretation of these results would also indicate the presence of both resonances. Finally, an angular study of vibrational excitation through the 8 eV resonance indicates Sigma symmetry with a single p sigma wave.

Model polarisation potentials for use in calculating positron-molecule cross sections are often based on approximations in which terms that depend on the sign of the charge of the projectile are neglected. Thus, these potentials do not fully distinguish between electron and positron scattering. The validity of using such potentials for positron-H₂ collisions is investigated by performing scattering calculations using electron- and positron-H₂ model polarisation potentials. Results show that, because of differences in the way electrons and positrons distort the target charge cloud, sign-dependent terms in the adiabatic potential can significantly influence the total cross section. Low-energy cross sections are also shown to be rather sensitive to how non-adiabatic polarisation effects are taken into account. Using a cut-off fully adiabatic positron-H₂ polarisation potential, theoretical total cross sections in reasonably good agreement with recent measurements are obtained.

The density operator theory and method are used to extend the fully quantum-mechanical theory of the two-level single-mode laser advanced by Sargent, Scully and Lamb (1974) to the homogeneously broadened four-level atomic system so that the master equation of the three-mode laser operation for this system is obtained. The steady-state operation characteristics of the laser in different cases are discussed and compared. Especially, pictorial comparison and discussion of the quantum-statistical properties of the laser are carried out by means of numerical computation. Meanwhile some experimental results obtained by Otsuka (1978) are qualitatively explained.