Table of contents

A simple approach is proposed to determine analytically the classically stable configurations of doubly excited states in atoms and Z ions, in the case when both electrons are located on one side of the nucleus. An analytical formula gives the ratio between the electron distances within the 1% accuracy of the existing numerical computations.

Anticrossings of the 1s5d and 1s6d ¹ Lambda and ³ Lambda Stark sublevels of He I were investigated using radiofrequency spectroscopy. Single and multiple photon transitions were induced and detected after excitation of thermal He atoms by 10-20 keV ion impact. The energy separations of closest approach of four 1s5d and two 1s6d anticrossings were deduced from the resonance signals.

Using a new approach the authors have determined the internal (vibrational) energy distribution of the molecular ions formed in the collisional associative ionization (AI) between thermal Na*(3p) atoms. The method is based on the wavelength dependent measurement of the Na⁺ ions formed in the photodissociation of the Na₂⁺ ions. The authors find that there is a substantial vibrational excitation and even a population inversion, with the upsilon =2 to 4 levels predominantly populated. These findings are in good agreement with previous experimental results obtained through electron detection, and with the detailed ab initio calculations by Dulieu et al (1991). The present method should prove itself suitable under conditions where other methods are difficult to apply, e.g. in collisions between cold atoms.

The dissociative and non-dissociative ionization of H₂ by electron impact has been measured up to 1 keV incident energy using a pulsed electron beam and pulsed ion extraction technique in conjunction with a time-of-flight spectrometer and relative flow technique. These data represent the first accurate measurement on the dissociative ionization of H₂ by electron impact over such a large energy range. It has been shown that the ratio, R*, of the dissociative ionization cross section to non-dissociative ionization cross section in H₂ serves as a useful parameter in probing the electron correlation effects. A comparison of the value of R* obtained by electron impact with those obtained by proton and antiproton impact shows strong dependence on the sign of the projectile charge.

We have carried out a generalization of the continuum distorted wave eikonal initial state (CDW-EIS) approximation for ion impact ionization in which the interaction of the ejected electron with the target is represented by a Hartree-Fock potential. We apply this theory to the ionization of Ar by 350 keV protons. Very good agreement is obtained with experiment for the angular and energy distribution of electrons ejected from the L and M shells.

We use unitarized first-order many-body theory (UFMB) to obtain differential cross sections for the electron impact excitation of the 2³S and 2³P states of helium in the intermediate energy range. Our results are in much better agreement with experiment than previous calculations using the weak-coupling approach.

When positronium formation in ordinary matter is enhanced through application of a sufficient electric field, determination of the increase in the triplet component, if the lifetime method is used, requires correcting the increase in the long component intensity by the factor ( lambda _L+ lambda _f- lambda _T)/(( lambda _L+ lambda _f).

A multiple-scattering method for non-muffin-tin potentials is presented. The new fully numerical method is implemented for self-consistent electronic structure calculations of molecules and clusters using the local-density approximation. No shape approximation of the potential is used. The method is general and can be applied to molecules of an arbitrary shape and is not limited to only diatomic molecules as previously reported fully numerical methods. The method is tested by calculating bound-state energies for H²⁺ and by self-consistent calculations for N₂ and CO using the local X alpha exchange-correlation potential.

Beam-foil excited spectra of nitrogen in the 40-120 nm wavelength range have been analysed with improved resolution. Among 122 new lines 65 have been classified. The charge states indicate that the remaining 57 lines should belong to N I/N II/N III/N IV as 13%/44%/30%/13%. A number of lifetimes and of relative intensities have been measured.

The potential value of pump-probe spectroscopy for the study of the spectral profiles of narrow gas-phase absorption lines has been investigated using the technique of concentration-modulated absorption spectroscopy (COMAS). The relationship of the probe laser 'gain' to the absorption line profile is established and the parameters of the Voigt line profile for the 3p²P₁2 /-3s²S₁2 / transition of atomic sodium in a hydrocarbon flame have been determined. The effects of a significant reduction in the Doppler contribution to the linewidth of this transition in the flame are clearly identified.

Auger satellites arising from transitions in the presence of 'spectatorY vacancies have minor intensity in electron impact excited spectra. They are, however, very efficiently produced if preceded by another Auger transition. We have measured the L₂₃-MM Auger spectrum of argon emitted after photoexcitation with broad-band synchrotron radiation of energies largely lying above the K-shell ionization potential. The spectrum exhibits a high degree of line overlap and is shown to be a superposition of at least six L₂₃-MM spectra belonging to various configurations of spectator vacancies and having comparable overall intensities. We made relativistic calculations in configuration average which reproduce the general appearance of the complex spectrum quite well. A number of lines in the L₂₃(L₂₃)-M₂₃M₂₃(L₂₃) and L₂₃(M₂₃²)-M₂₃M₂₃(M₂₃²) spectra have been analysed and five L₂₃^-1M₂₃^-2 states have been energetically located.

Radiative lifetimes have been measured for the 4p ²P^o and 5p ²P^o levels of Na I by time-resolved laser-induced fluorescence. The results are in excellent agreement with theoretical lifetimes derived from oscillator strengths calculated with the SUPERSTRUCTURE atomic structure program including detailed consideration of configuration interaction effects. Transition probabilities for all transitions depopulating these levels have been determined from a combination of the two sets of results.

For periodically forced systems we show that the scarred state corresponding to the unstable periodic orbit of the main resonance has a lower ionization rate than adjacent states because a representation exists in which coupling to other states is minimal. Our results explain why this scarred state has been observed in recent experiments on the microwave ionization of excited hydrogen atoms; other previously unexplained features are also elucidated.

Single photon double ionization of nitrogen (N₂), carbon monoxide (CO), nitric oxide (NO) and oxygen (O₂) has been studied in an electron-electron coincidence experiment. Threshold photoelectrons coincidence (TPEsCO) spectra have been obtained using a dual penetrating field technique, in which only electrons with nearly zero kinetic energy (<20 meV) are detected. The high sensitivity and energy resolution attainable have allowed observation of several states of N₂²⁺, CO²⁺, NO²⁺ and O₂²⁺, excited essentially at their respective thresholds, resolving vibrational energy levels for nearly all states where they exist. Accurate energies of the features observed in the TPEsCO spectra, and the molecular parameters derived from them, are presented, and the possible mechanisms of formation are discussed.

Total Auger rates for some nitrogen- and silicon-containing molecules have been calculated by two different methods. Using a direct approach, the rates for individual transitions have been calculated and then totalled. Alternatively, the total rate has been determined using a statistical approach based upon knowledge of the Mulliken atomic orbital populations in the molecule and the atomic Auger subshell rates. There is good agreement between the two approaches. Total Auger rates based on the statistical approach enable a first estimate of some core-hole inherent lifetimes in molecules to be determined.

New measurements of the absorption cross section of pure CF₃H in the spectral region of emission of CO₂ lasers are presented. Over the entire 9 mu m band, two distinct absorption regions were observed. At low laser fluences, the absorption cross section remained constant indicating that the absorption was linear. As the laser fluence was increased, a non-linear absorption process was observed. The fluences required for the onset of this non-linearity were found to be strongly dependent on the laser wavelength, varying from 50 J cm^-2 at 9P(8) to 5 J cm^-2 at 9R(26). Measurements were also taken in the 10 mu m region, but only weak absorption was observed over the fluence range studied (up to 100 J cm^-2).

The influence of collisions on magneto-optical rotation spectra is considered. By using a Rydberg state and nearby doubly-excited perturbers of the barium spectrum as probes, it is shown that, for a given spectral resolution, the broadening of the lines as a function of pressure can be determined in a sensitive way at lower pressures of the perturbing gases than in conventional photoabsorption experiments. The relative f-values of the lines (which do not change significantly in the pressure range investigated) are also accurately monitored by the same method. For the 5d8p doubly excited states in Ba and the Rydberg member which lies between them, the f-values are found to be constant to an accuracy of 5% as the pressure of rare gas is increased in the range up to 100 mb. We have also observed an intruder which can only be excited in the presence of the magnetic field.

We have analysed the capture-to-continuum aspects of the ionization process in ion-atom collisions in a channel-distorted, first-order approximation. In contrast to the standard Brinkman-Kramers approximation, we have found that the channel-distorted version correctly reproduces the ionization amplitude derived in a quasi-elastic scattering model for electrons in the binary-encounter-peak region of the ionized electron spectrum. Clearly, the first-order approach misses the double-scattering contribution, known to play a significant role at the cusp of electrons travelling with approximately the same velocity as the projectile, but our analysis is interesting in this region too, since it shows that the single-collision part of the cusp amplitude depends strongly upon the high-momentum tail of the Compton profile of the initial electron state, a feature that should receive more attention in future evaluations of higher-order theories. The present calculations are tested against experimental data in the binary-encounter region. Quantum effects due to binding in the initial state are isolated in a generalized off-energy-shell factor. This factor approaches unity at the centre of the binary peak, but appears to shift the position of the peak and to change its shape. The theoretical shift is in reasonable accord with the experimental data, but it is concluded that further work is needed, with special emphasis on the validity of the single-particle model and the corresponding representation of the initial state of the target atom.

The significance of Coriolis coupling in charge-transfer reactions during collisions of highly charged ions with atoms is studied under conditions where electron correlation effects may be ignored. Reaction amplitudes are expressed in closed form in the case of full Coriolis mixing and parametrized in simple terms when partial mixing is considered. It is shown that Coriolis mixing only plays a minor role in the diabatic wing of the reaction window, i.e. for capture into states with fairly small energy defects. The form of the reaction window is, on the other hand, strongly influenced by rotational coupling in the adiabatic wing corresponding to capture into more deeply lying final states.

A classical phase-space model of the hydrogen molecule which includes both electrons and nuclear centers is utilized to study state selective electron capture for collisions with alpha particles at intermediate impact energies (30-200 keV amu^-1). The calculations are performed within the framework of the classical trajectory Monte Carlo method. The direct one-electron transfer process and the transfer ionization reaction which becomes increasingly important at the higher impact energies are directly included to obtain line emission cross sections for capture to the He⁺(np) states, n=2-5, for energies from 16 to 60 keV amu^-1. Agreement between theory and available experimental data is good at energies above 40 keV amu^-1.

The spin polarization of electrons elastically scattered from Xe atoms was calculated for impact energies up to 10 eV. The calculations were carried out within a relativistic polarized orbital approximation. Our results are in good agreement with recent experimental data.

A variational R-matrix approach from first principles-the selected states R-matrix (SSRM)-is proposed for the study of electron scattering off polyatomic, non-linear molecules. Cartesian Gaussian-type orbitals are used exclusively to represent both valence and continuum orbitals. The two-electron integrals along with the nuclear-attraction integrals are evaluated over the entire space. In a subsequent step these integrals are effectively confined to the R-matrix sphere by subtracting their contributions from outside the sphere using a multipole expansion which approximates the projectile-target interaction in the outer region. This approach allows the use of efficient integral codes available for the calculation of molecular bound states. Correlation and polarization effects are only taken into account for selected, energetically low-lying (n+1)-particle CI roots, corresponding to the R-matrix poles. Higher-lying roots are approximated by a static exchange calculation. In order to test this method, electron scattering off N₂ is considered in detail for both resonant (² Pi _g) and non-resonant (² Sigma _g⁺) scattering symmetries. Good accordance is found between the eigenphase sums from the SSRM calculations and results obtained with a standard R-matrix package for linear molecules as well as with reference data from the literature.

A recently developed technique for calculating cross sections in electron-molecule collision processes based on variational R-matrix theory is discussed. One of the main advantages of this method is its applicability to polyatomic systems. As a first example, integral and partial cross sections for the e^-+CH₄ system below the ionization limit and in the energy range of the Ramsauer minimum are presented. In addition, the corresponding eigenphases are shown. All calculations are based on the fixed-nuclei approximation. The paper is concluded by an analysis of the Ramsauer minimum with respect to the angular dependency of the differential cross sections in that energy region.

Vibrationally elastic cross sections for electron collisions with CH₄ are calculated with an ab initio electrostatic potential. Effects of electron exchange and target polarization are taken into account approximately. Differential cross sections (DCS) are calculated for the electron energies 10-50 eV and compared with recent measurements. Comparing with the result of a spherical potential model, effects of anisotropy of the molecule are made clear. Integral and momentum-transfer cross sections are also calculated to see the validity of the extrapolation procedure taken in deriving those cross sections from experimental DCS.

Starting from the usual partial wave method, we present a theoretically exact formulation for the calculation of differential and total cross sections corresponding to electron impact excitation in atomic systems. For transitions involving an s state the derived formulae are very simple, free of Racah or Clebsch-Gordan coefficients, so that for computational purposes they are more convenient than the usual expressions. In particular, we have considered monopole, dipole and quadrupole transitions from (or to) an s state. In comparison with the usual formulation, the present one also has the advantage that it allows us to separate out the scattering amplitudes corresponding to excitation of individual magnetic substates: electron-photon coincidence parameters for dipole and quadrupole transitions from an s state can be obtained directly. For the total differential cross section an alternative formulation is derived and compared with the previous one. Both formulations given here are very general because they can be employed using any non-relativistic T-matrix, and are valid for both neutral and charged targets. In the weak coupling approximation and for alkali-like systems, the resulting formulae are compared with others of the literature; as an illustration the Coulomb Bethe approach is described. Finally, expressions for the integrated cross sections corresponding to excitation of individual magnetic substates are also given and it is shown how, for charged targets, the Coulomb phase must be included in all calculations except in those which give the total integrated cross section.

Using configuration interaction wavefunctions collision strengths Omega are calculated for transitions among the lowest 12 excited states of the (1s²2s²2p⁶) 3s², 3p², 3s3p, 3s3d, 3s4s and 3s4p configuration of Al II. The standard and no-exchange R-matrix programs are employed for computing the contributions of Omega for L<or=12 and L>12 respectively. Pseudo-resonances have been removed and the results are listed at 10 energies in the range 1.5 to 10 Ryd. A comparison of present results shows the earlier available collision strengths to be in error by over an order of 3 for some of the transitions. This error arose mainly in the computation of the contribution of Omega for L>6 in the earlier results. However, the differences are discussed and the accuracy of our present collision strengths is assessed.

Measurements of the differential cross section for electron impact excitation of the b ³Sigma_u⁺ continuum of molecular hydrogen are presented. The data were taken at incident electron energies of 9.2, 10.2, 12.2, 15.2, 17.2 and 20.2 eV, at scattering angles in the range of 20 degrees to 120 degrees . Integral cross sections are derived from these differential cross sections and the data are compared to available experimental and theoretical cross sections.

Electron beam transmission experiments have been performed, in the energy range 10 meV to 175 meV, with a magnetically collimated electron beam formed in a synchrotron radiation photoionization source using SuperACO, LURE. The apparatus has been calibrated with He and absolute backward scattering cross sections have been measured for the target gases H₂, N₂ and O₂. A relationship, involving s and p partial waves, has been established between the backward scattering cross sections ( sigma _B) and the momentum transfer cross sections ( sigma _M). This has been used to check the accuracy of experimental data and the consistency of values of sigma _B, sigma _M and total scattering cross sections. Experimental data and theory for H₂ are in good agreement, whereas for N₂ experimental values of sigma _B and sigma _M conflict below 80 meV and agreement with theories is mixed. For O₂, discrepancies are greater than for N₂ both in experimental data and between theory and experiment, which may differ by up to a factor of 5 for the total scattering cross section at the lowest energies.

The first absolute cross section measurements For double ionization of C⁺, N⁺, O⁺ and Ne⁺ ions by electron impact are reported. The animated crossed beams method has been employed in the energy range from below ionization thresholds to approximately 2500 eV. The classical binary encounter approximation overestimates measured cross sections by almost two orders of magnitude. Along the sequence, the cross section maximum does not follow classical scaling laws. A simple scaling law based on an electron pair ejection model is proposed for the prediction of direct double ionization. Inner-shell ionization followed by autoionization is seen to play a dominant role for C⁺ only.

Using a recently developed, unified version of both R-matrix and distorted-wave methods, we have carried out new calculations for the electron-impact ionization of Ti³⁺ and found that the previous calculations of Griffin et al. (Phys. Rev. A, vol. 25, p.1374-82, 1982) are incorrect. Contrary to one of the major conclusions of that paper, coupling between the three autoionizing and the sixteen bound LS terms of the 3p⁵3d² configuration is not important. However, the other major conclusion of that paper that the inclusion of singly-excited states is necessary to provide an alternate decay path for resonances remains valid. Further calculations including configuration-interaction (CI) mixing with correlation configurations formed by the double-electron promotions 3p² to 3d² were necessary to achieve reasonable agreement with experiment. A larger close-coupling calculation which included 3p⁶4s and 3p⁵3d4s target states and the corresponding 3p² to 3d² correlation configurations provided further improvements in the results. Potential problems with pseudo-resonances are found to be effectively diagnosed by comparison of distorted-wave and R-matrix results.

Energy distributions of scattered positrons (e⁺) and ejected electrons (e^-) resulting from 100 eV e⁺-Ar collisions have been determined at 30 degrees and 45^degrees in coincidence with the remnant Ar⁺ ions. A calibration method has been employed to convert the measured intensities into absolute cross sections. The experimental data are compared with values calculated by a classical-trajectory Monte Carlo method. The agreement is good for the secondary e^- but poor for the scattered e⁺. The ratios of the measured secondary e^- intensities at 15 eV have been determined for e^- and e⁺ impact and compared with existing experimental and theoretical values.

Based on the classical relativistic electrodynamics and taking account of the acceleration mechanism driven by the longitudinal electric field component, this paper indicates with both analytical and numerical calculation approaches that no noticeable energy exchange can occur in a vacuum between a charged particle and focused laser beams or laser pulses if the interaction length is unlimited.