Table of contents

The understanding of the magnetism of the atom, in the low-field limit, has recently been greatly improved through the discovery of an adiabatic invariant associated with the diamagnetic interaction. This was achieved using either classical perturbation theory or group theoretical methods on the Fock hypersphere. The authors present a new and simple way of establishing its expression based on symmetry considerations and only involving elementary calculations. The result looks like a generalised vectorial model for the diamagnetic interaction. Indeed, the success of the present approach lies in its close connection with the non-invariance algebra of the Coulomb problem thereby supplying the authors with an interesting starting point for tackling the complete problem.

Recently observed states in He^- in which two electrons reach high levels of excitation are shown to fit a Rydberg sequence of bound-state energy levels in a six-dimensional Coulomb potential (1/R). Thereby, these states provide conclusive evidence for a new class of quantised energy levels residing at the 'Wannier saddle point' of the two-electron system. In the vicinity of this saddle point of the potential surface, the system reduces to that of a single particle in an attractive Coulomb potential in the six-dimensional hyperspherical radius R=(r₁²+r₂²)¹2/. Energy positions, angular momentum aspects and intensity ratios of the observed states are all in conformity with this picture.

The photoionisation cross section of the ground state of neutral sodium is calculated with the close-coupling method. Polarisation of the 2p⁶ core by the valence electron is included in an ab initio manner by the use of polarised pseudo-states. The contribution to the radiative dipole matrix element due to core polarisation is shown to increase the cross section significantly at energies above the minimum. Good agreement is found with the experimental results if the latter are scaled by a factor of about 0.7, which suggests that the experimental curve should be renormalised. Oscillator strengths are also presented for 3s-np transitions are are compared with other theoretical results.

The angular distributions of photoelectrons resulting from six- and seven-photon ionisation (h(cross) omega =2.34 eV) of xenon have been investigated. They show pronounced structure. The results indicate that the seven-photon transition cannot simply be explained by a six-photon ionisation process followed by a continuum-continuum transition.

A verified theory of two-photon absorption is computationally investigated to study the dependence of the exponential index of multiphoton processes, K, on incoherent relaxation bandwidths as a function of detuning.

The screening effect of the projectile electrons in cases of 1s-2p excitation and L-shell ionisation is investigated in the framework of the semiclassical approximation. This approach allows the authors to take into consideration the change of the screening with the impact parameter and with the position of the projectile along its path. The dependence of the screening effect on the projectile energy and ionic charge is discussed.

Impact-parameter-dependent vacancy production probabilities for target K-shell excitation were measured for Sn to Cu and Xe to Ni collisions at 1.4 MeV amu^-1, up to centre-of-mass scattering angles of 143 degrees . In addition to direct excitation from MO near the united-atom limit found previously for light-on-heavy collisions, a peak due to the kinematic part of the rotational coupling was found at the smallest impact parameters. However, practically no signs of the adiabatic part of rotational coupling were found.

Cross sections for electron capture via resonant excitation of O⁵⁺ target ions are estimated for the intermediate states 1s²2pnl (2s to 2p, Delta n=0) involving high Rydberg states (nl). This result is combined with the previous calculation of the B²⁺ and C³⁺ cases to examine the effective charge dependence of the cross section.

A novel version of the classical model of post-collision interaction is proposed and extended to the cases when the energies of all the electrons involved in the process are of the same order of magnitude. Allowance is made for the finite time of motion of the Auger electron between the ion and the slower electron. The use of the model considerably improves the agreement between calculations and experimental data available on Auger line shifts at large excess energies.

Two approximate methods for evaluating the hyperfine splitting in the ground state of muonic helium are extended to states in which the muon (but not the electron) is excited. Simple, analytic expressions are given correct to first order in the electron-muon mass ratio, for any state and for both ⁴He and³He. The result agrees with the value given by Amusia et al. (1983) for the ⁴He 2s state.

The Auger spectrum of N₂ has been calculated by the Green's function method. Strong electron correlation effects have been found to be visible in this spectrum, corresponding to a breakdown of the two-quasi-hole picture for the final states of the Auger processes. Connections with the satellite structure in the photoelectron spectrum of N₂ have been established.

Nineteen ground and excited states have been calculated using a HF pseudopotential, a valence CI, and a core-valence perturbation. This work provides extremely precise adiabatic potential curves that will facilitate the interpretation of the experimental data. Three currently employed calculational methods (model potential, frozen-core all-electron, HF pseudopotential) are also discussed.

The hyperfine splitting and the isotope shift of the 542.2 nm line of Lu have been investigated by laser spectroscopy and accurate values of the A and B factors of the corresponding states have been determined for ¹⁷⁵Lu and ¹⁷⁶Lu. In addition the accuracy of the previous data for the 573.7 nm and the 605.5 nm lines of Lu has been improved.

Assuming the shake-off process and Coster-Kronig theory of the origin of the satellites accompanying L alpha ₁, the separations and relative intensities of these satellites are calculated for ₇₇Ir, ₇₉Au and ₉₂U. Contours of the satellite structure are obtained and compared with the results of Richtmyer and Ramberg for gold.

The L-shell X-ray intensity ratios I(L_alpha )/I(L_l), L(L_alpha )/I(L_beta ) and I(L_alpha )/I(L_gamma ) in Au, Pb, Th and U at various photon energies 15<or=E<or=60 keV have been measured and their energy dependence is studied. A comparison of the experimental values is made with those calculated using the X-ray emission rates and subshell photoelectric cross sections from Schofield (1973), subshell fluorescence yields and Coster-Kronig transition probabilities from Chen and Crasemann (1981), Krause et al. (1978), McGuire (1971) and Rao et al. (1969, 1971). Fairly good agreement is observed between the present experimental results and the calculations, within the uncertainties of the determinations. To the best of the authors' knowledge, the results in the elements Pb, Th and U, at the present experimental energies, are reported for the first time.

Diamagnetic splitting of Rydberg atomic states in weak magnetic fields is studied. The zero-order perturbation equation is solved by the discrete WKB method, the quantum defect is taken into account through the boundary condition. Effects of tunnelling are considered. A semiclassical quantisation condition is obtained determining eigenvalues as functions of quantum numbers, field and quantum defects of the atom in question.

Photoelectron spectra of atomic barium have been recorded at several photon energies in the range 20<or=hv<or=29 eV. The variations of the Ba⁺5p⁶nl (nl=6s,5d,6p,7s,6d) photoelectron peak intensities were measured in the 20-21 eV autoionisation region. The results indicate that each autoionising state decays to the various states of the ion in a characteristic way. Asymmetry parameters measured for the 6s, 5d and 6p channels show strikingly different values that follow the general trend beta (6s)> beta (6p)> beta (5d). Above 21 eV, 5p ionisation begins to dominate, and the Auger spectra were used to monitor the production of the various 5p hole states. As the photon energy is scanned across the autoionising resonances, these Auger spectra indicate that excitation of 5p hole states decaying by two-step autoionisation is the dominant mode in the production of very-low-energy photoelectrons (<or=2 eV) and their corresponding high-energy Auger electrons. Calculations suggest that a heretofore unobserved 5p⁵6p²²P₃2/5d autoionising level is responsible for the Auger distribution measured at hv=28.9 eV.

An exactly soluble model is used to describe transitions, induced by two strong laser beams, proceeding from the initial discrete state to the continuum via two channels. In the first channel the transition occurs via an intermediate, near-resonant discrete state located close to the continuum edge. In the second channel the intermediate state belongs to the continuum. Population trapping occurs due to an interference between the two channels. The model can be used to describe the photodetachment of electrons from negative ions or the photodissociation of molecules.

Absolute absorption coefficients for the Sr resonance line at 4607 AA perturbed by He, Ne and Kr below 700 Torr have been measured at about 1000K in the spectral range of about 1 to 100 cm^-1 on both sides of the line centre. The profile of Sr-Kr shows a strong asymmetry and has a red satellite at about 35 cm^-1 from the line centre. The profiles of Sr-He and Sr-Ne show, on the other hand, rather symmetric features and have a power dependence of approximately -2 on the separation from the line centre. An empirical interaction potential has been deduced from the red satellite of Sr-Kr using the unified Franck-Condon (UFC) method of Szudy and Baylis (1975). It has a well of depth 40 cm^-1 at an interatomic distance of 6.6 AA. Optical cross sections have been deduced from the profiles of Sr-He and Sr-Ne using a classical impact analysis. The results are 117 and 135*10^-16 cm² respectively. Finally the measured absorption profiles are compared with a recent experiment on collisionally redistributed radiation.

The structure of rotational rainbows for the excitation of homonuclear molecules out of the ground state by atoms is examined. The authors assume that the energy sudden approximation (ESA) is valid but not the centrifugal sudden approximation (CSA). In these circumstances the classical differential cross section is known to have a step at the maximum value of the classically accessible angular momentum transfer, and logarithmic singularities elsewhere, rather than a single square root singularity as when the CSA is also valid. The authors have investigated how quantal effects alter this picture and in particular have tried to ascertain when the effects due to the breakdown of the CSA become apparent. They conclude that when analysing such cross sections it is important to use a quantisation axis along the apse, (P_f-P_i), where P_i(P_f) is the initial (final) momentum of the incident atom. They find that both interference effects and transitions out of the m=0 state play important roles, and that when the m=2 state is classically accessible the CSA is no longer valid.

The authors present further measurements of cross sections for ionisation of atomic hydrogen using a crossed-beam coincidence technique previously developed in this laboratory. Cross sections for ionisation by 400-9756 keV Ar^(3-9)+ and 1000-4770 keV C^(2-6)+ and O^(2-6)+ ions have been determined and considered in terms of cross section scaling predictions. The authors have adapted the experimental technique to permit studies of one-electron capture by multiply charged ions; previous measurements have relied entirely on the furnace target method. The effectiveness of this alternative approach has been demonstrated by measurements of the cross section for electron capture by 140-4000 keV Ar^(3-6)+ ions. The cross sections for the combined processes of charge transfer and ionisation have been considered in terms of classical scaling rules.

The Au L X-radiation induced by He, Li, Be, C, O, Ne, Si and S ion impact has been investigated for impact energies ranging from 0.23 to 6.7 MeV u^-1. From the intensities of the L alpha , L gamma ₁ and L gamma _2,3,6 lines the L-subshell ionisation cross section ratios have been derived. The results are compared with measurements of other groups. Since cross section results depend critically on the procedure used in the data analysis, data from other groups were re-evaluated, if necessary, in order to apply a consistent procedure. After that, reasonable agreement among the various sets of experimental cross sections is found. Comparisons with state-of-the-art theories reveal systematic discrepancies. These are attributed to systematic errors in the atomic parameters used in the data analysis as well as to a failure of the ionisation theories employed.

Differential and total cross sections for the elastic scattering of electrons and positrons by magnesium atoms in the intermediate energy region from 10 to 500 eV are calculated using a real optical potential and are compared with the available theoretical and experimental results. The agreement with the experimental data is only moderate but that with the theoretical values of Gregory and Fink (1974) is fairly good.

Simplified versions of the second-order Born and Faddeev-Watson approximations are applied to the excitation of the n=2 levels of atomic hydrogen by the impact of 54.4 eV electrons. The theories are compared with the measurement of differential cross sections and angular correlation parameters. The results indicate that the Born approximation is better at low angles of scattering but that the Faddeev-Watson approximation is better at high angles. The importance of the phases of the two-body T matrices in the Faddeev-Watson approximation is illustrated.

The second-order distorted-wave amplitude is evaluated in a high-energy approximation. In this approximation, the non-local second-order distorting potential is replaced by a local second-order potential obtained in the closure approximation. It is shown that for the case of electron excitation of the 2¹P state of helium, the second-order effects are fairly large and improve agreement with experimental data. It is also noted, however, that the best agreement with experimental data may be obtained through a hybrid first-order calculation. Results for positron excitation of helium are also given and discussed.

The polarised orbital method for the adiabatic description of low-energy electron/positron-atom scattering is reformulated in terms of a polarised density. The polarised density is expanded into multipoles and the radial functions are treated variationally. This procedure is shown to provide an accurate dipole polarisation potential for e⁺/e^--H scattering. For many-electron targets the quantum statistical Thomas-Fermi energy functional with gradient corrections is used to calculate dipole and quadrupole polarisation potentials. Total elastic cross sections for positron energies below excitation thresholds agree with experiment and with polarised orbital Hartree-Fock calculations for e⁺-Ne scattering, while the agreement is less satisfactory for e⁺-Ar scattering.

Detector geometries may be chosen to determine the total cross section sigma _tot, the asymmetry parameter beta and the photo polarisation p while achieving good experimental statistics by using large acceptance angles in gas-phase photoemission experiments, in which the emitted electron current distribution is described by sigma _tot(1+ beta P₂(cos gamma ))/4 pi , where gamma is the angle between the emitted electron and the direction of one of the polarisation components of the photon beam. When a cylindrical mirror analyser (CMA) with variable acceptance geometry is used, it is possible to determine sigma _tot, beta and the incident light polarisation p by measuring the flux with several different restricted (vignetted) annular rings whose polar and azimuthal angular acceptance ranges are appropriately chosen. The mathematical structure of this problem precludes determination of sigma _tot, beta and p from the measured photocurrent using a CMA having its aperture vignetted only in the azimuthal or polar angle.

The intensity correlation behaviour of the resonance fluorescence from a vibronic system is studied, taking into account the effect of the pump radiation on the vibrational relaxation process. It is found that the correlation behaviour changes essentially with the strength of the pump, even if no saturation of the electronic transition occurs.

Methyl alcohol (CH₃OH) molecules pumped with the 9R(16) CW CO₂ laser line generate two new far-infrared laser lines at wavelengths of 33.4 and 36.4 mu m. The 33.4 mu m line has the shortest wavelength among the sub-mm lines reported so far from CH₃OH and its isotopic species pumped with CW CO₂ laser lines. Stark splitting is observed on the 36.4 mu m line as well as on the 39.9 and 77.4 mu m lines pumped with the 9P(34) and 9R(8) CO₂ lines, respectively. It is pointed out on the 39.9 and 77.4 mu m lines that the Stark components in each of the Delta M=+or-1 sequences of the far-infrared transition are accidentally degenerate to give a frequency shift independent of the magnetic quantum number M.

A method is presented for averaging the ionisation cross section over the electron energy distribution with due allowance for the curvature of the trajectory which becomes an important factor in strong electric fields whether or not the distribution is Maxwellian. The ionisation rate is then found to depend on the electron acceleration, mean velocity, and temperature as independent parameters. This theory has proved to be essential to an exact solution of the equations for electron wave breakdown, and can be successfully applied to the data on swarms in strong fields.