Table of contents

We present high-resolution momentum distributions of ions and electrons created in single ionization of He, Ne and Ar targets by intense (0.15–2 PW cm⁻²) short-pulsed (25 fs) linearly polarized laser fields in the direction perpendicular to the polarization. Instead of a Gaussian shape predicted by standard tunnelling theory, the experimental data exhibit a sharp cusp-like peak at zero transverse momentum. The comparison of experimental data with (i) calculations performed within the strong-field approximation employing a Coulomb–Volkov wavefunction to model the final electron state and (ii) results of recent semiclassical calculations, shows that the 'cusp' appears due to the long-range Coulomb interaction between the emitted electron and the remaining ion. A similar structure was previously observed for ion–atom collisions.

This paper reviews our recent progress on the angle-resolved photoion-yield spectroscopic (ARPIS) technique applied to the K-shell excitations of the simple linear molecules, N₂, CO, C₂H₂, O₂, CO₂, N₂O, OCS and CS₂. This spectroscopic technique enables us to distinguish symmetries of electronic states; e.g., the ΔΛ = 0 parallel and ΔΛ = ±1 perpendicular transitions for the diatomic molecules can be distinguishable. We call this technique symmetry-resolved inner-shell excitation spectroscopy. The complete symmetry resolution is rationalized by the axial-recoil fragmentation. Moreover, the angular distribution of fragment ions is directly related to the molecular orientation upon the photoabsorption and the vibrations; therefore, the ARPIS measurement can be used to reveal complicated vibronic couplings in polyatomic molecules in terms of anisotropy in fragmentation. We have studied Rydberg-valence mixings, vibronically induced 1s → Rydberg excited states, and strong bending-mode couplings in Renner–Teller split 1s →π* states for the linear molecules through the ARPIS spectra.

We present an ab initio calculation of the line shape of the photoabsorption spectrum of Na₈ and Na⁺₉ clusters at zero temperature, including the polarization effect driven by the electron–plasmon coupling, together with the electron–phonon coupling. The electron–plasmon coupling is found to yield a significant lowering of the energy centroid of the plasmons, whereas the main effect the electron–phonon coupling has is to smooth the spectrum.

Following from our work on triply excited hollow resonances in three-electron atomic systems, a density functional theory (DFT)-based formalism is employed to investigate similar resonances in the Li-isoelectronic series (Z = 4–10). A combination of the work-function-based local nonvariational exchange potential and the popular gradient plus Laplacian-included Lee–Yang–Parr correlation energy functional is used. The generalized pseudospectral method provides nonuniform and optimal spatial discretization of the radial Kohn–Sham equation. First, all the eight n = 2 intrashell states of B²⁺, N⁴⁺ and F⁶⁺ are presented, which are relatively less studied in the literature compared to the remaining four members. Then calculations are performed for the eight 2l2l'nl'' (3 ⩽ n ⩽ 6) hollow resonance series, namely 2s²ns ²S^e, 2s²np ²P^o, 2s²nd ²D^e, 2s2pns ⁴P^o, 2s2pnp ⁴D^e, 2p²ns ⁴P^e, 2p²np ⁴D^o and 2p²ns ²D^e, of all the seven positive ions. Next, as an illustration, higher resonance positions of the 2s²ns ²S^e series are calculated for all the ions with a maximum of n = 25. The calculated excitation energies are in excellent agreement with the available literature data (for the n = 2 intrashell states the deviation is within 0.125% and excepting only one case, the same for the resonance series is well below 0.5%). With an increase in Z, the deviations tend to decrease. Radial densities are also presented for some of the selected states. The only result available in the literature for the lower resonances (corresponding to a maximum of n = 17) have been reported very recently. The n > 16 (17 for F⁶⁺) resonances are examined here for the first time. This gives a promising viable and general DFT scheme for the accurate calculation of these and other hollow resonances in many-electron atoms.

A procedure for the ab initio study of electron–molecule collisions at intermediate energies is presented in detail. The molecular R-matrix with pseudostates method is based on the inclusion of discretized continuum states in the close-coupling expansion. This method allows, for the first time, the calculation of totally ab initio cross sections for electron impact ionization of molecules as well as for electronic excitation above the first ionization threshold. The method is general and can be applied to multielectron targets. Results for collisions with H⁺₃ and H₂ are presented. Numerical considerations necessary for performing a successful calculation are detailed.

Argon atoms in a pulsed supersonic expansion are prepared in selected Stark components of Rydberg states with effective principal quantum number in the range n* = 15–25. When traversing regions of inhomogeneous electric fields, these atoms get accelerated or decelerated depending on whether the Stark states are low- or high-field seeking states. Using a compact electrode design, which enables the application of highly inhomogeneous and time-dependent electric fields, the Rydberg atoms experience kinetic energy changes of up to 1.2 × 10⁻²¹ J (i.e. 60 cm⁻¹ in spectroscopic units) in a single acceleration/deceleration stage of 3 mm length. The resulting differences in the velocities of the low- and high-field seeking states are large enough that the corresponding distributions of times of flight to the Rydberg particle detector are fully separated. As a result, efficient spectral searches of the Rydberg states best suited for acceleration/deceleration experiments are possible. Numerical simulations of the particle trajectories are used to analyse the time-of-flight distributions and to optimize the time dependence of the inhomogeneous electric fields. The decay of the Rydberg states by fluorescence, collisions and transitions induced by black-body radiation takes place on a timescale long enough not to interfere significantly with the deceleration during the first ∼5 µs.

We explore from a theoretical point of view pump and probe (P&P) analysis for fission of metal cluster where probe pulses are generalized to allow for scanning various frequencies. We show that it is possible to measure the time the system needs to develop to scission. This is achieved by a proper choice of both delay and frequency of the probe pulse. A more detailed analysis even allows us to access the various intermediate stages of the fission process.

Absolute cross sections for electron-impact dissociation of D¹³CO⁺ ions have been measured over a collision energy range from 4 to 100 eV with a crossed electron–ion beams method. Total experimental uncertainties are about 16% near the cross section peak. The role of resonant and direct dissociative excitation for energies below 21 eV is discussed in light of the energy levels and photo-dissociation cross sections of the HCO⁺ formyl cation predicted by ab initio multi-reference configuration interaction calculations.

Differential cross sections for photoionization into the 5d⁹6s^{2 2}D_5/2 ionic state of mercury have been obtained over the photon energy range from 15 eV to 17 eV. By measuring the photoelectron emission at the angles of 0^o and 90^o with respect to the major axis of the polarization ellipse the variation of the asymmetry β-parameter for the ²D_5/2 ionic state of mercury has been determined as a continuous function of photon energy over the above range. This energy range contains three series of autoionizing states converging to the ²D_3/2 ionization limit. The measurements have been performed using the 5 m McPherson normal incidence monochromator at the Daresbury Synchrotron Radiation Source with a photon energy spread of typically 3 meV.

In calculating collision strengths and excitation rates for electron impact on moderately ionized iron peak elements, one might question whether the Breit–Pauli R-matrix method is sufficiently accurate as compared with the Dirac R-matrix method. We test this for Fe¹⁴⁺ by removing as far as possible any variation in algorithmic features, such as the energy mesh and target state expansion, as opposed to genuine differences between the two approaches. We find the average difference between the Breit–Pauli and Dirac R-matrix effective collision strengths is only 6%, which confirms the hypothesis that if one gets the Dirac and Breit–Pauli target states close, and resolves the resonances adequately (we use up to 384 101 points), then the Dirac and Breit–Pauli collision strengths are in good agreement. We finally tabulate the best converged effective collision strengths for T = 10⁵–10⁷ K for all transitions involving the lowest 10 levels of Fe¹⁴⁺.

Absolute differential elastic and vibrational excitation cross sections have been measured for CH₄ at 90° and from 0.1 to 1.5 eV, with the resolution of all four vibrational modes. Threshold peaks were observed in the excitation of the ν₁, ν₃ and ν₄ modes, but not in the ν₂ mode. These results are in good agreement with the recent calculation of Nishimura and Gianturco (2002 J. Phys. B: At. Mol. Opt. Phys.35 2873). No near-threshold structures were observed.

The Mott insulator state created by loading an atomic Bose–Einstein condensate (BEC) into an optical lattice may be used as a means to prepare a register of atomic qubits in a quantum computer. Such architecture requires a lattice commensurately filled with atoms, which corresponds to the insulator state only in the limit of zero inter-well tunnelling. We show that a lattice with spatial inhomogeneity created by a quadratic magnetic trapping potential can be used to isolate a subspace in the centre which is impervious to hole-hoping. Components of the wavefunction with more than one atom in any well can be projected out by selective measurement on a molecular photo-associative transition. Maintaining the molecular coupling can sustain a commensurately filled register for the duration of a quantum computation.

The first complete multi-state CDW close coupling calculations which use a fully normalized basis set are performed. The results obtained at impact energies in the region of 10 keV for total and n = 2 capture cross sections are in reasonably good accord with experiment despite the fact that only the ground states of both species and the n = 2 states of the projectile are incorporated into the model. The theory has significant advantages over other atomic and molecular expansions which may require extensive bases to obtain similar accuracy.

Following an electrostatic interpretation of the force F_xc = −∂V_xc/∂r associated with the exchange-correlation potential V_xc(r), we present both analytical and numerical results for F_xc(r) in the Ne atom. The basic input is an existing quantum Monte Carlo (QMC) calculation of the ground-state electron density ρ(r) in this atom. The analytic form of ∂V_xc/∂r is in terms of the number of electrons Q(r) enclosed in a sphere of radius r centred on the nucleus, plus two phases needed to characterize the radial wavefunctions of density functional theory. Eigenvalue equations are presented for these phases, and used numerically. A brief discussion is added on the result of replacing the QMC ground-state density by its Hartree–Fock counterpart.

We investigate the fluorescence spectrum in a nearly degenerate atomic system of a F_e = 0 → F_g = 1 transition by analytically solving Schrödinger equations. An ultranarrow fluorescence spectral line in between the two coherent population trapping windows has been found. Our analytic solutions clearly show the origin of the ultranarrow spectral line. Due to quantum interference effects between two coherent population trapping states, the width and intensity of the central spectral line can be controlled by an external magnetic field. Such an effect may be used to detect a magnetic field.

The proton energy spectrum from photodissociation of the hydrogen molecular ion by short intense pulses of infrared light is calculated. The time-dependent Schrödinger equation is discretized and integrated. For few-cycle pulses one can resolve vibrational structure, arising from the experimental preparation of the molecular ion. We calculate the corresponding energy spectrum and analyse the dependence on the pulse time delay, pulse length and intensity of the laser for λ ∼ 790 nm. We conclude that the proton spectrum is a sensitive probe of both the vibrational populations and phases, and allows us to distinguish between adiabatic and nonadiabatic dissociation. Furthermore, the sensitivity of the proton spectrum from H⁺₂ is a practical means of calibrating the pulse. Our results are compared with recent measurements of the proton spectrum for 65 fs pulses using a Ti:Sapphire laser (λ ∼ 790 nm) including molecular orientation and focal-volume averaging. Integrating over the laser focal volume, for the intensity I ∼ 3 × 10¹⁵ W cm⁻², we find our results are in excellent agreement with these experiments.

A theory of electron collisions with van der Waals clusters is developed which is capable of describing the vibrational Feshbach resonances (VFRs) recently observed in electron attachment to CO₂ clusters. VFRs appear below vibrational excitation thresholds of one molecular unit and their energy decreases with the increase of the cluster size. The selectivity observed in vibrational excitation of individual components of a Fermi dyad is substantially reduced in the attachment spectra. We also calculate elastic electron scattering and vibrational excitation (VE) of one molecular unit in a cluster environment and demonstrate a strong influence of the VFRs on the VE cross sections. Theoretically predicted VFRs are much narrower than observed. We explain the VFR broadening by contribution of several neutral precursors, contribution of different cluster conformations and the dependence of the resonance energy on the position of the molecular unit in the cluster. A model is developed to describe the latter effect quantitatively.

The quantum mechanical lifetimes of atomic hydrogenic states are shown to follow a universal curve when plotted against a simple function of their quantum numbers n and l. This universal curve is found to agree with a result derived from the correspondence principle. A simple formula which approximates the universal curve can be used to easily calculate lifetimes for all states n, l ⩾ 1 to an accuracy of 400 parts per million or better. The formula is especially useful for high-n states, where the full quantum calculation is extremely difficult or even impossible to perform.

The theory of electron loss from projectile-ions in relativistic ion–atom collisions is extended to the case of collisions with excited atoms. The main feature of such collisions is a resonance which can emerge between electron transitions in the ion and atom. The resonance becomes possible due to the Doppler effect and has a well-defined impact energy threshold. In the resonance case, the ion–atom interaction is transmitted by the radiation field and the range of this interaction becomes extremely long. Because of this the presence of other atoms in the target medium and the size of the space occupied by the medium have to be taken into account and it turns out that microscopic loss cross sections may be strongly dependent on such macroscopic parameters as the target density, temperature and size. We consider both the total and differential loss cross sections and show that the resonance can have a strong impact on the angular and energy distributions of electrons emitted from the projectiles and the total number of electron loss events.

Switching anisotropic molecules from strongly absorbing to strongly amplifying through a transparent state is shown to be possible by the application of dc or ac control electric fields without the requirement of population inversion. It is based on decoupling of the lower level molecules from the resonant light while the excited ones remain emitting due to their state-dependent alignment. The amplification index may become dependent only on a number of excited molecules and not on the population of the lower state. A suitable class of molecules and applications in optoelectronics, fibreoptics and nanophotonics are outlined.

In a two-dimensional Bose–Einstein condensate, the reduction in dimensionality fundamentally influences collisions between the atoms. In the crossover regime from three to two dimensions, several scattering parameters have been considered. However, finite temperature results are more difficult to obtain. In this work, we present the many-body T-matrix at finite temperatures within a gapless Hartree–Fock–Bogoliubov approach and compare to zero and finite temperature results obtained using different approaches. A semi-classical renormalization method is used to remove the ultraviolet divergence of the anomalous average.

A stochastic model of associative ionization in collisions of Rydberg atoms with ground-state atoms is presented. The conventional Duman–Shmatov–Mihajlov–Janev (DSMJ) model treats the ionization as excitation of Rydberg electron to the continuum by the electric-dipole field generated by exchange interaction within the quasi-molecular ion. The stochastic model essentially extends this treatment by taking into account redistribution of population over a range of Rydberg states prior to ionization, which is caused by non-adiabatic processes in overlapping multiple level crossings of quasi-molecular Rydberg states. The redistribution is modelled as diffusion of electrons in the Rydberg energy spectrum using a Fokker–Planck-type equation. The process of l-mixing of Rydberg states at large internuclear distances and twisting of the collision trajectories on attractive potentials are taken into account. The choice of the collision velocity distribution is also shown to be important. Associative ionization rates have been calculated for Na**(nl) + Na collisions with n = 5–25 and l = 0, 1, 2, and compared with the available experimental data and the calculations performed using the nonlinear DSMJ model. At relatively low n the stochastic model yields a substantially better agreement with the experimental data than the DSMJ model, while the results of both models converge at large n.