Radiative lifetimes, collisional mixing, and quenching of the cesium 5DJ levels.

Abstract

We report the results of a series of pulsed and cw laser experiments which measure spontaneous and effective lifetimes of the Cs(5DJ) levels and investigate excitation-transfer collisions involving Cs(5DJ) atoms and ground-state-cesium perturbers. With cw excitation of the dipole-forbidden but electric-quadrupole-allowed 6S1/2\ensuremath{\rightarrow}5D5/2 transition, we monitor the ratio of sensitized to direct fluorescence, i.e., I5D3/2\ensuremath{\rightarrow}6P1/2/I5D5/2\ensuremath{\rightarrow}6P3/2. A rate-equation analysis of these data yields values for the Cs(5D5/2)+Cs(6S)\ensuremath{\rightarrow}Cs(5D3/2)+Cs(6S) excitation-transfer rate, and for the rate of quenching of Cs(5D) by ground-state perturbers. The role of out-of-multiplet quenching is discussed at length, and we have demonstrated that quenching by ground-state atoms dominates over that by cesium dimers under these conditions. In the pulsed-laser experiments, the temporal evolution of the 5DJ\ensuremath{\rightarrow}6S and cascade 6PJ\ensuremath{'}\ensuremath{\rightarrow}6S fluorescence was observed following either direct forbidden-line pumping of one fine-structure level, or indirect excitation of both fine-structure levels (i.e., molecular excitation followed by predissociation). Analysis of the buildup and decay rates of the various levels as a function of cesium density yields values for the natural lifetimes of the Cs(5DJ) levels, as well as for the excitation transfer and quenching rates. Best values for the lifetimes and cross sections for Cs-Cs collisions obtained from these combined experiments are \ensuremath{\tau}5D=1250\ifmmode\pm\else\textpm\fi{}115 ns, \ensuremath{\sigma}5/2\ensuremath{\rightarrow}3/2=36\ifmmode\pm\else\textpm\fi{}8 \AA{}2, and \ensuremath{\sigma}5D=30\ifmmode\pm\else\textpm\fi{}3 \AA{}2. In addition, numbers for the 6P level-quenching cross sections due to collisions with Cs atoms and Cs2 molecules, respectively, were obtained: \ensuremath{\sigma}6P(Cs)=6.6\ifmmode\pm\else\textpm\fi{}3.0 \AA{}2 and \ensuremath{\sigma}6P(Cs2)=863\ifmmode\pm\else\textpm\fi{}260 \AA{}2. In the discussion, we show how these new values can be used to reconcile the seemingly discrepant results of several previous studies.