High-charge-state thorium (up to 80+) and uranium (up to 70+) ions have been extracted from an electron-beam ion trap (EBIT). The ions were produced after injecting low-charge-state thorium and uranium ions initially from a metallic vapor vacuum arc ion source into the trap. The extracted ions from the EBIT provide microsecond-wide ion-beam pulses of about ${10}^4$ ions per second. The production and loss rates are estimated from modeling calculations and are compared to the measured extracted ion rates.

Resonance exchange of excitation (REE) between gas molecules can reduce the magnitude of various macroscopic kinetic phenomena produced in the gas by velocity-selective excitation. Considerable reduction will occur whenever the rate of REE exceeds both the rate of elastic velocity-randomizing collisions and the rate of the spontaneous decay of the excited state. Expressions given here are necessary for the evaluation of the change in magnitude of the kinetic effects as a result of REE.

We develop the band theory of a two-level atom moving in a standing light wave. The energy bands allow for a simple discussion of various scattering resonances that play an important role in atom optics.

We report a quantitative determination of the v=1 polarizability in ${\mathrm{H}}_2$ by measuring the optical Stark shift and splitting of the Q(0) and Q(1) vibrational Raman transitions. For the Q(1) transition, the ${\mathit{M}}_{\mathit{J}}$=±1,0 components are resolved, thereby allowing a determination of the polarizability anisotropy for the v=1 state. The measured optical Stark shift of the Q(0) transition at 1060 nm is 23 MHz ${\mathrm{GW}}^{\mathrm{-}1}$ ${\mathrm{cm}}^2$, giving a polarizability difference between the v=1 and 0 states of Δα=0.49±0.03 a.u., in good agreement with ab initio theory.

The spectral distribution of light scattered by a two-level atom in a laser beam with phase modulation has been calculated. The spectrum has been obtained via numerical solution of the optical Bloch equations using the algorithm of Wilkens and Rza̧żewski [Phys. Rev. A 40, 3164 (1989)]. The modulation produces additional sidebands in both coherent and incoherent parts of the spectrum; the location of the added components depends on the modulation depth. In the case of two-frequency phase modulation with incommensurate frequencies, wild, ‘‘chaotic’’ behavior of the spectrum has been observed, similar to the case of bichromatic amplitude modulation. The intensity distribution is practically insensitive to the chosen constant value of the phase of modulation.

Quasinormal modes are characterized by complex energy eigenvalues and by eigenfunctions that are outgoing waves at infinity, and are therefore not normalizable in the usual sense. In the space of such wave functions, the operators of quantum mechanics are not Hermitian, and the familiar tools break down. We reformulate the time-independent perturbation theory for quasinormal modes and express the corrections to the complex energy in terms of matrix elements. One main feature of the results is the appearance of an alternative definition for the norm, which is finite for quasinormal modes. Explicit formulas are given for the real and imaginary parts of the first-order correction and for the imaginary part of the second-order correction, the latter in the approximation where the original quasinormal mode is extremely narrow. This restricted (but physically most useful) part of the perturbative formalism depends only on the first-order matrix elements, and the simple form of the second-order correction can be interpreted in terms of a generalized golden rule.

The observation of radiation trapping effects in time-resolved, stepwise-excited, electron-photon coincidence signals is reported. Hg atoms were excited by electron impact to the 6 ${}_{}{}^1{}_{}{}^{}$${\mathit{P}}_1$ state and then by laser excitation to the 6 ${}_{}{}^1{}_{}{}^{}$${\mathit{D}}_2$ state. Coincidence signals were observed from the 6 ${}_{}{}^1{}_{}{}^{}$${\mathit{D}}_2$–6 ${}_{}{}^3{}_{}{}^{}$${\mathit{P}}_1$ (313-nm) transition. The decay time of these signals was found to depend upon radiation trapping of the electron-impact-excited 6 ${}_{}{}^1{}_{}{}^{}$${\mathit{P}}_1$–6 ${}_{}{}^1{}_{}{}^{}$${\mathit{S}}_0$ (185-nm) resonance line. An explanation of this effect is given in terms of the laser excitation coupling the 313- and 185-nm transitions.

The nonrelativistic version of the authors’ theory of multiphoton ionization [Phys. Rev. 40, 4997 (1989)] is compared with other forms of scattering-theoretical approaches that employ different boundary conditions. These methods, which are expressed in terms of formal time-independent scattering theory, are shown to be equivalent in the large-photon-number limit to use of the semiclassical time-dependent S-matrix theory. In the authors’ treatment it is assumed that the photoelectron has escaped from both the electromagnetic and atomic fields; a perturbation expansion in the photon-electron interaction is obtained that correctly accounts for the atomic final-state interactions of the photoelectron. The case of multiphoton detachment of ${\mathrm{I}}^{\mathrm{-}}$ is treated, for which the influence of the final-state atomic potential on the transition rate can be ignored. It is shown both analytically and numerically for the Nd–yttrium aluminum garnet wavelength that the transition rates obtained for this case from the various approaches are distinctly different. The present approach is shown to be consistent with an abrupt switching off of a spatially unlimited monomode field, in contrast with Keldysh-type approaches, which correspond to a long switch-off time if the ad hoc assumption of ponderomotive acceleration is made. Above-threshold-ionization spectra calculated from the present approach exhibit interference effects and are redshifted with respect to what is expected for slow switch-off. It is demonstrated, using the semiclassical time-dependent approach, that the corresponding final off-field scattering-state wave function only exists if the ponderomotive potential per unit photon energy is an integer, in accordance with our earlier time-independent analysis. The origin of this curious feature is made clear.

The minimal-coupling Hamiltonian for a nonrelativistic hydrogenic atom interacting with a photon field in the α-Lorentz gauge is canonically gauge transformed to one with four arbitrary degrees of gauge freedom. The theory, nonrelativistic in the charge dynamics, is wholly covariant in its treatment of the scalar, transverse, and longitudinal electromagnetic fields. One degree of gauge freedom of the transformed Hamiltonian comes from the arbitrary scalar parameter contained within the α-Lorentz gauge, while the others come from arbitrary parameters governing the gauge transformation. A Jaynes-Cummings-type model is constructed by suitable restrictions to the gauge freedom, and found to display an admixture of transverse and scalar interactions but no longitudinal component. The model gives the Lamb shift, correct to the second order of the electronic charge consistent with relativistic quantum electrodynamics.

The dynamical behavior of counterpropagating light beams interacting with a nonlinear medium is studied numerically. When the incident intensities are increased, this system becomes unstable and exhibits oscillations undergoing complex behavior, including quasiperiodic motion and chaos. We analyze the output fields by using Fourier spectra, phase portraits, and autocorrelation functions. In the chaotic regime, the Lyapunov exponents and dimensions are calculated. A physical interpretation of the quasiperiodic motion is also given by an exact calculation of the absorption spectrum of a two-level atom.

Expressions are derived for determining higher-order correlation functions of a fluctuating signal partially masked by additive noise when the contribution of noise in the signal is not negligible. A way of computing higher-order correlation functions equivalent to cumulants by subtracting the contributions of all lower-order correlation functions is also described. The relations will be useful for the experimentalists wishing to obtain useful information, especially from noisy signals.

We have investigated the statistical properties of stimulated Brillouin scattering (SBS) in a single-mode optical fiber. Gain narrowing of the Stokes spectrum is observed as the input laser power is increased, and large stochastic fluctuations are observed in the Stokes output intensity. The experimental results can be described well by a theoretical model that includes the spontaneous nature of the initiation of SBS.

A simple model of coupled electronic and collective nuclear motion is used to study energy-transfer processes in the presence of intense short-wavelength laser fields. The complete Hamiltonian of the model includes distinct nuclear, electronic, and laser-field terms. A semiclassical method of extracting spectral information from the generated trajectories of the Hamiltonian is used to calculate laser-electronic-nuclear energy-transfer rates. The nuclear term describes a deformed heavy nucleus; it sufficiently treats the collective nuclear motion, which is lower in frequency than that which would be given by an individual-particle nucleon model. We specifically report on the dynamics of this model, and on the energy transfers that ensue, in the presence of an x-ray laser field, as a function of the electron radial distance. This general problem is of basic interest in the studies of laser-driven nuclear excitation.

We have measured the power spectrum of the intensity fluctuations of light transmitted by a Fabry-Pérot interferometer when the input field is the real Gaussian field. The real Gaussian field is a field characterized by real, random (Gaussian) amplitude fluctuations. The bandwidth of the real Gaussian field was varied, taking on values less than that of the interferometer, as well as greater. Comparisons of the measured spectra with calculated spectra are quite satisfactory. Of special interest is a feature in the spectra centered at the laser-interferometer detuning frequency.

We present a general theory for intensity noise reduction when one beam of light is used to control another. We use a semiclassical input-output formalism and determine the conditions necessary for noise reduction below the shot-noise level. Configurations are examined where the intensity of one beam is monitored and control is effected directly to the other beam (feedforward), or indirectly through the beam source (feedback). The results are applied to the case of twin beams generated by an optical parametric oscillator. We detail for this system the implementation of an electro-optic control channel. Finally, a comparison is made between theory and experiment.

Two-time intensity correlations in a laser with multiplicative white noise have been studied experimentally. Multiplicative noise was introduced by applying a Gaussian-noise voltage to an intracavity acousto-optic modulator in a He:Ne laser operating near threshold. We observe both a shift and a broadening of the peak in the correlation-time versus mean-light-intensity curve. These observations are in agreement with the previous measurements of the steady-state properties of the laser near threshold in the presence of multiplicative noise.

We present comprehensive results for the ionization of the one-dimensional model atom with one electron bound by an attractive δ-function potential in time-dependent external fields. We employ a high-order numerical method, allowing for high-resolution scans of the (essentially exact) ionization probability as a function of the external field parameters within reasonable computer time. We show that, for a sinusoidal external field with small frequency (multiphoton regime), the ionization depends sensitively on the field strength. This phenomenon is a generic multiphoton effect: The dominating peaks in the ionization rate as a function of the ponderomotive energy are separated by the photon energy. This structure is not sensitive to the pulse turn-on time, and, contrary to earlier beliefs, it also persists in the regime where tunneling is the dominant ionization process. (We obtain the tunneling background as the cycle-averaged dc rate.) The sensitive intensity dependence should be observable in photodetachment measurements with high resolution. A comparison with results recently presented by Greenwood and Eberly [Phys. Rev. A 43, 525 (1991)] for a one-dimensional finite-range model potential shows detailed agreement, demonstrating a surprising independence of the ionization on the short-range binding potential. Concerning time dependence, we report drastic deviations from the (usually excellently fulfilled) exponential decay law for the bound-state probability for cases where the latter becomes sufficiently small. In certain regions this function shows a sensitive dependence on time. In addition, we present evidence for exact zeros of the bound-state probability. Finally, we find the ionization rates in dc fields to agree excellently with those resulting from time-independent methods. Moreover, they yield a threshold field strength of some 80% of the experimental value for ${\mathrm{H}}^{\mathrm{-}}$.

We use a lithium iodate nonlinear crystal as a source of two quantum correlated beams at about 830 nm. A liquid-crystal cell is placed in one of the beams and is electrically driven to produce a very small modulation of its transmission coefficient. Low-noise high-efficiency photodiodes detect the light in each beam and the spectrum of the photocurrents, and the spectrum of the difference between the photocurrents, are measured. The difference spectrum shows a reduction of the noise below the shot-noise level of 4 dB compared with the classical difference technique, where the shot noise in the two detectors simply adds. However, the classical noise in one beam is low enough to make a single-beam shot-noise-limited measurement. Using an optimized difference technique, we show a reduction of the noise level compared to the single-beam shot noise of 2 dB.

A variational procedure is developed for the calculation of the multiphoton-ionization rate of an atom interacting with a superposition of two time-dependent electric fields, one weak and the other intense and slowly varying. Trial functions are chosen that take into account the effect of both the atomic binding potential and the external fields. The variational expression contains, as its leading term, an approximation for the transition amplitude that was derived previously. The correction term is carefully analyzed here with regard to its dependence on suitably defined frequency and intensity parameters. The analysis provides firmly based estimates of the order, relative to the leading term, of the error in the approximation. A sum rule is derived which, owing to the cancellation of higher-order correction terms, allows for an accurate representation, in a relatively simple form, of the total transition rate summed over the net number of photons exchanged with the low-frequency field. A modified final-state trial function is presented which may be used, without loss of accuracy, to account for the existence of scattering resonances with energies close to that of the emergent electron. The calculational procedure is tested numerically on a one-dimensional model problem; the results are consistent with the more formal analysis of the relative order of the correction term and the accuracy of the sum rule.

We derive the exact solution of the evolution equation for the density matrix of the field in a lossless micromaser cavity with the injected atoms prepared in their ground state. The time development of photon statistics and of the fluctuations in the field quadratures is analytically and numerically investigated. It is shown that a squeezed state with sub-Poissonian photon statistics can be created at the onset of evolution.