Measurement of the radiative lifetime of the cesium 5D5/2 level using pulsed excitation and delayed probe absorption

Abstract

Abstract.The 5D5/z level of cesium is populated by pulsed 540 nm photodissociation of Csz, and selectively detected via the 6P3/2 population that results by radiative decay. It is monitored by absorption of a delayed probe pulse. Its time evolution yields a value of 1260 + 80 ns for the radiative lifetime of the Cs 5D5/2 level, in good agreement with the calculation by Theodosiou [15]. PACS: 32.70.FwRadiative lifetimes are important data in themselves. They are also useful as checks of ab initio calculations, or as inputs for semi-empirical calculations, as is the case with cesium. Indeed cesium is the atom of choice for atomic parity-violation experiments [ 1,2] and as such has been the subject of a great deal of theoretical work (see in particular the recent ab initio calculation of [3]). Atomic theory has now reached a point where the calculated lifetimes or allowed dipole matrix elements of the cesium S and P states fit accurately the most precise measurements [4]. Up to now however, the 5D levels have not been as extensively studied. Experimentally, the difficulty comes from the absence of an El-allowed transition from the ground state. Two lifetime measurements have been published. In the first one by Marek [5], the 7 P3/2 level of cesium was excited by tunable laser pulses in the presence of xenon, and emission bands were observed around 7100 A. These bands were attributed to the fluorescence of Xe-Cs molecular states formed with the 5D atoms that result from the decay of the 7 P3/2 level. The time constant 890 4-90 ns of the fluorescence decrease was taken as a measure of the 5D lifetime. In the second 5D lifetime measurement by Sasso et al. [6], the observed signal was the atomic 5D~6S fluorescence of one 5 Dj level tbllowing a pulsed laser excitation of the other 5D j, level ("sensitized fluorescence") or of the predissociative E level * Laboratoire associ6 au CNRS (URA 18), g l'Ecole Normale Sup6rieure et ~ l'Universit6 Pierre et Marie Curie of Cs 2. The 5D--*6S fluorescence has the advantage of being directly proportional to the 5D population, but has the drawback of being an E 1-forbidden process, with the consequence of low counting rates, typically one photon for 100 laser pulses at the Cs pressures of one or several Torr used for the experiment. At these pressures, the 5D radiative decay rates were small compared to the cotlisional quenching rates, and the value of T D (1250 _+ 150 ns) could be obtained only after extrapolation to zero Cs density. In addition, both measurements involved the populations of the two fine structure levels 5D5/2 and 5D3/2, with unknown relative contributions. This point is of importance since the calculations of the lifetimes give values that differ by roughly 35% fYom 5D5/2 to