Excited State Distributions and Fluorescence Signals

N. Dimarcq, V. Giordano, P. Cérez

Gerald J. Diebold

When a homonuclear diatomic molecule is photodissociated to give one ground state and one electronically excited atom, the wave function of the two-atom systems must reflect an uncertainty as to which atom is excited. Consequently, the wave function of the photon for a photodetector viewing the resulting fluorescence must be written as a superposition of states. Detection of such photons is shown to give rise to an interference signal in the detector voltage at a frequency depending on the Doppler shift from the atomic recoil. Photodissociation of a random distribution of molecules gives rise to a detector voltage whose power spectral density is determined by the angular distribution of photofragments.

John V. James

We have observed fluorescence and ionization signals resulting from two-photon resonant excitation of thallium in an atomic beam. The output of a pulsed dye laser is used to excite the resonance between the ground state (6('2)P(,1/2)) and an excited n('2)P state of thallium. Both two-photon absorption and three-photon absorption occur simultaneously; the latter results in photoionization of the atom. The excited states populated by two-photon absorption decay via cascade fluorescence back to the ground state; one or more of these spectral components constitute our fluorescence signal. The lifetimes of the excited n('2)P states of thallium have been measured directly by observing the exponential decay of this fluorescence. Calculations are included to ascertain the time required for the fluorescence signal to reach within 2% of a simple exponential decay. This decay is a measure of the lifetime of the initial state, under the condition that none of the fluorescence components arise from states which are populated by cascade through longer-lived states. The lifetimes of the n('2)P(,1/2) and n('2)P(,3/2) states of thallium for n = 7 to 11 are determined in this manner and compared with the values deduced from semi-empirical calculations of oscillator strengths by various authors. For calculations employing the effects of core polarizability in the electron wavefunctions and in the transition matrix elements, the agreement is very good--to within our experimental uncertainty, which is typically(, )<5%. Our measurements are also compared to the four other known experimental values for such excited states of thallium. Some of these previous experiments and calculations were carried out in conjunction with parity non-conservation experiments in thallium. Our measurements may provide useful checks on the accuracy of some of the calculations performed in the evaluation of such experiments. The ionization and fluorescence signals are also studied under conditions of strong saturation of the two-photon resonance with the 7('2)P states. The manifestation of saturation in the ionization signal is a broadening of the excitation spectrum, which is found to greatly exceed that of the fluorescence signal. Furthermore, the ratio of the ionization signal to the fluorescence signal at the peak of the resonance is found to vary linearly at intensities below about 10('9) W/cm('2); this too saturates at higher intensities. Various models are employed to simulate the observed behavior, encompassing both rate equations as well as density matrix formulations. The numerical solutions to these models are compared with the observed behavior; the agreement is found to depend strongly on the assumed experi- mental conditions to which the ensemble of atoms is subjected.

Miguel Orszag

Scott Bergeson, A. Denning, Mary Lyon +1 more

We report measurements and simulations of disorder-induced heating in ultracold neutral plasmas. Fluorescence from plasma ions is excited using a detuned probe laser beam while the plasma relaxes from its initially disordered nonequilibrium state. This method probes the wings of the ion velocity distribution. The simulations yield information on time-evolving plasma parameters that are difficult to measure directly and make it possible to connect the fluorescence signal to the rms velocity distribution. The disorder-induced heating signal can be used to estimate the electron and ion temperatures $~$100 ns after the plasma is created. This is particularly interesting for plasmas in which the electron and ion temperatures are not known.