We show how an initially prepared quantum state of a radiation mode in a cavity can be preserved for a long time using a feedback scheme based on the injection of appropriately prepared atoms. We present a feedback scheme both for optical cavities, which can be continuously monitored by a photodetector, and for microwave cavities, which can be monitored only indirectly via the detection of atoms that have interacted with the cavity field. We also discuss the possibility of applying these methods for decoherence control in quantum information processing.

An excited atom in a photonic band-gap crystal can emit a photon that is then reabsorbed and reemitted by the atom, to form a dressed atom-field state. We discuss the existence conditions for such a trapped photon dressed state with a general multilevel atom driven by coherent fields and show that more than one such trapped state can exist. The field characteristics of the trapped field and the population trapped in the excited state are studied for various atomic systems. It is found that the driving field can determine the spread and energy of the trapped field, as well as the number of possible trapped dressed states.

We study the photon production in a one-dimensional cavity whose left and right walls oscillate with the frequency ${\Omega }_L$ and ${\Omega }_R,$ respectively. For ${\Omega }_L\ne {\Omega }_R,$ the number of generated photons by the parametric resonance is the sum of the photon numbers produced when the left and the right wall oscillates separately. However, for ${\Omega }_L={\Omega }_R,$ the interference term proportional to $\cos \phi$ is found additionally, where $\phi$ is the phase difference between two oscillations of the walls.

The stability of four-body systems ${(m}_a^{+}{m}_b^{+}{m}_1^{-}{m}_2^{-})$ in three (3D) and two dimensions (2D) is discussed using accurate numerical results obtained by means of diffusion Monte Carlo calculations. In 3D, we extend our proof of the stability for the class of systems ${(m}_a^{+}{m}_b^{+}{m}_1^{-}{m}_1^{-})$, showing that they are stable against the dissociation in ${(m}_a^{+}{m}_1^{-})$ and ${(m}_b^{+}{m}_1^{-})$ for any value of the mass ratio $m_a^{+}{/m}_b^{+}$. In 2D, using the ground-state energy of the dipositronium, it is possible to prove that the stability of four-body systems follows the same scenario. We also give upper and lower bounds to the binding energies for the class ${(M}^{+}{M}^{+}{m}^{-}{m}^{-})$ in 2D, useful to discuss the relative stability of biexciton molecules in semiconductors.

We have measured the atomic transition probabilities of two weak $E1\mathrm{}$ multiplets in neutral oxygen for which advanced atomic structure calculations have encountered problems due to severe cancellation of positive and negative parts of the transition integrand. Measurements were made in emission from a wall-stabilized arc. Our results differ significantly from those of all recent calculations.

The ground states of all atoms (through $Z=94)$ and their first and second ions have been calculated using the Kirschbaum-Wilets (KW) quasiclassical, many-body atomic model [Phys. Rev. A 21, 834 (1980)]. The present calculations allow the atoms to be in any spin state. The results essentially agree with previous calculations, done for atoms and their first ions with $Z=1\mathrm{}$ to 38 [Phys. Rev. A 51, 266 (1995)], which constrained the atoms to minimal-total-spin configurations $(S=0\mathrm{}$ or $\frac{1}{2}$). The KW model is shown to be similarly valid for the heavier atoms. However, it is found that a number of the heavier atoms and ions have KW ground states with nonminimal total spins $(S=1\mathrm{}$ or $\frac{3}{2}$).

Multiconfiguration Dirac-Fock calculations for the ${2s}^{21}{S}_0–2s2p{}^3{P}_1$ intercombination transition in C III are revisited. To improve the accuracy, the orbital sets for the initial- and final-state wave functions were not restricted to be the same, but were optimized independently. The calculated $2s2p{}^{3}P$ fine-structure splitting and the ${2s}^{21}{S}_0–{2s2p}^3{P}_1$ transition energy are in good agreement with experiment. The predicted transition rate $A=102.9\mathrm{}\pm 1.5{\mathrm{s}}^{\mathrm{-}1}$ is in agreement with a recent storage-ring measurement by Doerfert et al. [Phys. Rev. Lett. 78, 4355 (1997)]. Results are also presented for the allowed ${2s}^{21}{S}_0–2s2p{}^1{P}_1$ transition and the ${2s}^{21}{S}_0–2s2p{}^3{P}_2$ magnetic quadrupole transition.

The oxygen $1s$ photoelectron spectrum for ${\mathrm{CO}}_2$ has been calculated using ab initio electronic structure theory, a localized hole simulated by the equivalent-cores approximation, and the harmonic-oscillator approximation. The predicted spectrum is found to be in good agreement with a recently reported experimental spectrum, and is in better agreement with the experimental spectrum than is an earlier prediction by Domcke and Cederbaum. The equivalence of the approach taken here and that used by Cederbaum and Domcke is discussed, as well as the reasons for the differences in the calculated results.

We study the hyperfine splitting (hfs) of an electron in hydrogenlike ${}^{209}{\mathrm{Bi}}^{82+}.$ It is found that the hfs energy splitting can be explained well by considering the $g$-factor reduction due to the binding effect of a bound electron. We determine the experimental value of the magnetic moment of a tightly bound electron.

The properties of the ground state of the muonic helium atom $e\mu {}^3{\mathrm{He}}^{2+}$ have been calculated nonvariationally, using the correlation function hyperspherical harmonic method utilizing a nonlinear parametrization of the correlation function. The parametrization is similar to the one used in an earlier paper for $e\mu {}^4{\mathrm{He}}^{2+}$ but the differences in the convergence were found to be important for the choice of optimal parameters. The parametrization is especially suited to accelerate the convergence of singular operators. As a result, the obtained expectation values of the $\delta \left({\mathbf{r}}_k\right)$ operators have error margins smaller than the differences in the literature. The lowest-order hyperfine splitting, which depends on the fine structure constant and on the magnetic moment of the ${}^3{\mathrm{He}}^{2+}$ nucleus, is compared with values in the literature.

Generalized oscillator strengths of krypton for ${4p}^6{(}^1{S}_0)\to {4p}^5{(}^2{P}_{3/2,1/2})5s$ are calculated in the first Born approximation at a self-consistent-field level with Dirac-Fock, Dirac-Slater, Hartree-Fock, and Hartree-Slater wave functions. Strong relativistic effects are noticed. The estimate of relaxation on a single-electron matrix element suggests the necessity of a Dirac-Fock multiconfiguration calculation for the excitations.

We report measured pressure ratios for mixtures of neon with other rare gases that lead to a cancellation of normal light-induced drift of potassium atoms excited on the $D_1$ line by a narrow-band laser. These data can provide a stringent test for potassium–rare-gas potential curves. To a good approximation, the pressure ratios that cancel normal light-induced drift are directly related to ratios of cross-section differences. Corrections for collisions that cause fine-structure transitions without randomizing the velocity of the potassium atom are discussed.

We calculate integral cross sections for the electronic excitation to the ${}^3{T}_2$ states of $X{\mathrm{H}}_4$ $(X=\mathrm{C}$,Si,Ge,Sn,Pb) by electron impact. This is the lowest-lying excited state of these molecules. Our results were obtained with the Schwinger multichannel method with pseudopotentials at the two-state level of approximation. In the case of ${\mathrm{CH}}_4$ we compare our results with previous results of an all-electron calculation obtained at the same level of approximation, in which case we found an excellent agreement between the two calculations. Though these molecules are very similar, after discarding the cores, as the pseudopotential technique does, the inelastic cross sections are very distinctive and do not have a monotonic behavior with increasing proton number $Z$ of the central atom.

The collision strengths associated with excitation of the Be-like O ion by electron collision are calculated by the relativistic distorted-wave method. The so-called generalized occupation numbers are used to determine the distorted potential. We investigate in detail the influence of the target states on the collision processes. It is found that the calculated results are sensitive to the amount of valence-valence correlation included, especially for the resonance excitations to the states ${2p}^2{\hspace{0.5em}}^1{D}_2$ and ${}^1{S}_0$, in which the resulting collision strengths are enhanced by several times.

A general scheme is presented for obtaining systematic relativistic corrections to quantum-mechanical sum rules for various calculations in atomic physics. The single-particle picture and the Foldy-Wouthuysen transformation [Phys. Rev. 78, 29 (1950)] are adopted to derive such corrections. Results are obtained for the “generalized Bethe sums,” which include all higher moments of the energy transfer, and explicit results are given to the lowest order of relativistic corrections.

Vibrationally state-selective complete population transfer from the ground electronic state to the excited electronic state is demonstrated on a femtosecond time scale for the $X^2\Pi$ and $A^2{\Sigma }^{+}$ states of the OH molecule by means of computer simulation within the Schrödinger wave-function formalism. State-selective population transfer within the excited electronic state $A^2{\Sigma }^{+}$ of OH is demonstrated as well. These processes are controlled by shaped linearly polarized femtosecond laser pulses in the ultraviolet and in the infrared region, correspondingly, with the probability of the population transfer being close to 100%.

We present a generalization of the Keldysh-Faisal-Reiss model that includes the second-order $S$-matrix element and thus the effects of rescattering. On the example of a screened Coulomb potential we show that using our model it is possible to reproduce all main features of the recent experimental results, mainly, the appearance of a plateau, the position of the cutoff, and the side lobes. We also show that the difference between the direct ionization probability and the rescattering plateau for a long-range Coulomb-like potential (with a small screening parameter) is very large, which is in discrepancy with the experimental results. This indicates that the effective potential that an electron feels during the rescattering is not a pure Coulomb potential. We show that this potential can be approximated by a screened Coulomb potential with a properly chosen screening parameter.

We report improved frequency stability of the dual-noble-gas maser. This recently developed device can measure very small changes in the Zeeman transition frequencies of cohabitating ensembles of ${}^{129}\mathrm{Xe}$ and ${}^3\mathrm{He}$ atoms, and thus may be useful for symmetry tests and precision measurements such as a search for a permanent electric dipole moment of the ${}^{129}\mathrm{Xe}$ atom. Using the dual-noble-gas maser, we measured the frequency stability (i.e., the Allan deviation) of the ${}^3\mathrm{He}$ Zeeman transition to be approximately 100 nHz in 6000 s of data acquisition. This Zeeman frequency stability is an order of magnitude improvement over our previous report [R. E. Stoner et al., Phys. Rev. Lett. 77, 3971 (1996)].

The probability that an atom of hydrogen, initially in the $5g$ $\mathrm{}(m=4\mathrm{})\mathrm{}$ state, survives a 90-fs pulse of 620-nm wavelength is calculated both by direct integration of the time-dependent Schrödinger equation and by a Floquet calculation. The two methods give virtually identical results. The survival probability calculated for a one-electron model of neon, for the same initial state, pulse duration, and wavelength, is in fair quantitative agreement with the experimental data of van Druten et al. [Phys. Rev. A 55, 622 (1997)].

We present a method for reconstructing the photon-number distribution from the homodyne statistics based on maximization of the likelihood function derived from the exact statistical description of a homodyne experiment. This method incorporates in a natural way the physical constraints on the reconstructed quantities, and the compensation for the nonunit detection efficiency.