The degree of alignment induced in excited levels by two-step excitation with short broadline optical pulses having perpendicular linear polarisations is considered for different sequences of angular momenta J 0 to J 1 to J 2 . The results are compared with those recently reported for the simpler case of two-step excitation with pulses having parallel linear polarisations. It is found that for certain sequences, such as J to J to J+1 and J to J+1 to J+1, for which the degree of alignment in the case of parallel polarisations is small or even zero, the degree of alignment in the perpendicular case is large. These results can be applied to determine the optimum excitation polarisations in Zeeman quantum beats experiments for a given two-step excitation sequence.
For pt.I see ibid., vol.20, p.251 (1987). Theoretical expressions for cascade light intensity as a function of time have been evaluated for the case of cascade Zeeman beats produced by two-step excitation with broad-line laser pulses, and the results are compared with experimentally observed cascade Zeeman beats in Fe I. The beats were detected in time-resolved fluorescence from intermediate levels populated by cascades from highly excited levels of Fe I, using three different stepwise-excitation-cascade sequences. In the case of coincident laser pulses, good agreement between theory and experiment is obtained over a range of magnetic fields when allowance is made for Larmor precessions during the excitation period, for the width and shape of the exciting laser pulses and for the contribution of impurity processes. The effects of precession during the excitation period are especially important for the excitation sequence a 5 D 4 to z 5 F degrees 5 to e 5 F 5 , for which short-pulse theories predict no quantum beats for the experimental geometry used (parallel linear polarisations in the two excitation steps). For non-coincident pulses a predicted oscillation in the amplitude of the cascade beats with magnetic field has also been observed.
A theory of stepwise pulsed-laser excitation of cascade Zeeman quantum beats has been developed for a four-level system. The laser pulses are treated in the broad-line approximation and the full optical-pumping equations for the atomic-density matrix of each level have been obtained. These can be used to describe saturation and light shift effects. The weak-pumping theory solutions to the optical pumping equations have been used to obtain expressions for the cascade quantum-beat intensity versus time curves. Larmor precession, spontaneous-emission decay and the populating of the final cascade level during the excitation pulse are all allowed for. A Wiener-Levy phase diffusion model with slowly varying amplitudes is used for the laser pulses, and the final theoretical expressions for the quantum beats depend on the time-dependent intensities of the laser pulses as well as on their polarisations. Hyperfine effects have been ignored. The theory applies to both coincident pulses and to pulses with a time delay between them. An illustrative application to the case of two very short but coincident pulses (during which Larmor precession, spontaneous emission etc. can be ignored) is given. In an accompanying paper the theoretical expressions are further evaluated and successfully applied to the interpretation of cascade Zeeman quantum beat experiments in Fe I.
Aligned atomic Rydberg states of sodium can be prepared using two-step excitation from the ground state by linearly polarized pulsed lasers. Information that is normally inaccessible, e.g., sublevel partial cross sections in charge-transfer experiments, can be obtained when aligned targets are used. The calculations of orbital alignment must carefully allow for fine and hyperfine structure, laser linewidths, pulse widths and delays, sublevel coherences, and other factors. In this paper we derive the orbital alignments and time-averaged d-state sublevel populations for 3
2S1/2\ensuremath{\rightarrow}3
2PJ1\ensuremath{\rightarrow}n
2D excitations in Na using angular-momentum and density-matrix methods. We consider both quadrupole alignment
A(2) and hexadecapole alignment
A(4), with excitation through either
J1=1/2 or 3/2 intermediate states considered on the same footing. We show sublevel populations for \ensuremath{\Vert}
ML\ensuremath{\Vert}=0, 1, and 2 analytically and graphically. Finally, we formulate the experimental design problem quantitatively in order to ascertain how to optimize the choice of polarizer angles for extraction of sublevel partial cross sections. Although perhaps the commonest instance, two-step excitation of Na(nd) is but one of a large number of interesting cases, and this study is further intended to illustrate and guide the application of these methods to other light atoms.
The induction of higher polarization moments-the octupole orientation and the hexadecapole alignment—in the case of two-step excitation of atoms through dipole transitions and their manifestation in the polarization of dipole emission under conditions of anisotropic collisions are studied. For all channels of the two-step excitation process J 0→J 1 J, the efficiencies of production of higher polarization moments with integer and half-integer angular momenta from J=2 to J=7 through the absorption of linearly or circularly polarized light are calculated. The rates of collisional relaxation and interconversion of higher polarization moments, for both orientation and alignment, as functions of frequency detuning of the laser line from the center of the Doppler profile are calculated. The theory is illustrated by an example of two channels of excitation of the atomic state J=2 by circularly polarized light: J 0=0→J 1=1J=2 and J 0=3→J 1=2 (or 3)→J=2. In the first case, the octupole orientation enhances or attenuates the signal of circular polarization depending on the laser frequency detuning. In the second case, it represents the only source producing orientation and circular polarization of the emitted light.
The effects of laser-induced anisotropy on the generation of the short-pulse laser-induced polarization spectroscopy (LIPS) signal is investigated by direct numerical integration (DNI) of the time-dependent density matrix equations. The calculations are performed in the short-pulse regime (laser pulse width τL less than characteristic collision time τC) to reduce the influence of collisions on the generation of medium anisotropies. The Zeeman-state structure of the upper and lower energy levels is included in the multistate formulation of the density matrix equations. For a P-branch transition when the isotropic ground-level population is pumped by a circularly polarized light, oriented anisotropy is mainly responsible for the LIPS signal generation; whereas, when the resonance is pumped by a linearly polarized light, aligned anisotropy is mainly responsible. For a Q-branch transition that is pumped by a circularly polarized light, the contributions to the LIPS signal from orientation and alignment are comparable. The effects of saturation on the induced anisotropy is also investigated. The magnitude of the LIPS signal increases by more than a factor of 14 for an initial right circularly polarized-oriented anisotropic distribution in the ground Zeeman states as opposed to an isotropic distribution. An understanding of the effects of anisotropy on the LIPS signal will aid the modeling of the LIPS signal-generation processes and the interpretation of experimental LIPS signals.