We describe a set of experiments that investigate excitation transfer among the barium 6s5d
3DJ metastable levels due to collisions with argon perturber atoms, and quenching of atoms in these levels by nitrogen perturber molecules. The metastable levels were populated through optical pumping of the 6
s2 1S0\ensuremath{\rightarrow}6s6p
3P1o intercombination transition with a pulsed laser, followed by stimulated emission or stimulated Raman scattering into the 6s5d
3D1,2 levels. Collisional mixing then distributed population throughout the
3DJ levels. Time-dependent absorption coefficients were obtained by measuring the transmission of a weak cw probe laser beam (after the pulsed laser fired) at several perturber number densities. Argon buffer gas was used to investigate fine-structure mixing, and nitrogen was used to study quenching. The time-dependent absorption coefficients were fitted to appropriate fitting functions, obtained by solving the set of coupled rate equations used to model the time evolution of the metastable level populations. The buildup rates of the initially unpopulated
3D3 and underpopulated
3D1 levels were used to determine the collisional mixing rates, while the measured decay rates of the
3DJ level populations yielded the quenching rates. From plots of buildup rates versus argon number density for the
3D1,3 states, we were able to determine four of the six fine-structure mixing rate coefficients
kij.The results are
k12=(1.71\ifmmode\pm\else\textpm\fi{}0.18),
k21=(1.39\ifmmode\pm\else\textpm\fi{}0.13),
k23=(1.42\ifmmode\pm\else\textpm\fi{}0.13), and
k32=(1.92\ifmmode\pm\else\textpm\fi{}0.16)\ifmmode\times\else\texttimes\fi{}${10}^{\mathrm{\ensuremath{-}}13}$
cm3 ${\mathrm{s}}^{\mathrm{\ensuremath{-}}1}$. The experiment is not sensitive to the values of
k13 and
k31. Estimated rate coefficients for excitation transfer among the
3DJ levels due to collisions with barium atoms are
k21Ba=1.3\ifmmode\times\else\texttimes\fi{}${10}^{\mathrm{\ensuremath{-}}10}$
cm3 ${\mathrm{s}}^{\mathrm{\ensuremath{-}}1}$ and
k23Ba=5.8\ifmmode\times\else\texttimes\fi{}${10}^{\mathrm{\ensuremath{-}}10}$
cm3 ${\mathrm{s}}^{\mathrm{\ensuremath{-}}1}$. A plot of the population decay rates versus
N2 number density for each of the three metastable levels yielded the average nitrogen quenching rate coefficient
kq=(3.14\ifmmode\pm\else\textpm\fi{}0.19)\ifmmode\times\else\texttimes\fi{}${10}^{\mathrm{\ensuremath{-}}11}$
cm3 ${\mathrm{s}}^{\mathrm{\ensuremath{-}}1}$.
NaRb molecules in a thermal cell in a mixture of Na-Rb were excited to the D 1 Π state by means of a 514.5 nm beam from an Ar + laser. The natural lifetime for the NaRb D 1 Π, v' = 0, J' = 44 state, populated by the selective laser excitation, was found to be τ = 22.6 ± 1.8 ns. The kinetics of the molecular fluorescence lines as a function of the Rb atom concentration were also studied. The rate constant for collisional quenching of this molecular level by Rb atoms could be determined to k q = 2.1 · 10 −9 cm 3 s −1 . The experiments at steady state conditions demonstrate that energy transfer to the atomic 4 2 D levels of Rb dominates the collisional quenching with a rate constant of k ET = 1.2 · 10 −9 cm 3 s −1 . The results suggest that such molecule-atom transfer of electronic energy can be used to efficiently populate 2 D levels of alkali metal atoms.
We have examined fine-structure mixing between the rubidium
52P3/2 and
52P1/2 states along with quenching of these states due to collisions with methane gas. Measurements are carried out using ultrafast laser-pulse excitation to populate one of the Rb
52P states, with the fluorescence produced through collisional excitation transfer observed using time-correlated single-photon counting. Fine-structure mixing rates and quenching rates are determined by the time dependence of this fluorescence. As
Rb(52P) collisional excitation transfer is relatively fast in methane gas, measurements were performed at methane pressures of
2.5−−25 Torr, resulting in a collisional transfer cross section ($5\phantom{\rule{0.16em}{0ex}}^{2}P_{3/2}\ensuremath{\rightarrow}5\phantom{\rule{0.16em}{0ex}}^{2}P_{1/2}$) of $(4.23\ifmmode\pm\else\textpm\fi{}0.13)\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}15}\phantom{\rule{4pt}{0ex}}{\mathrm{cm}}^{2}$. Quenching rates were found to be much slower and were performed over methane pressures of
50−−4000 Torr, resulting in a quenching cross section of $(7.52\ifmmode\pm\else\textpm\fi{}0.10)\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}19}\phantom{\rule{4pt}{0ex}}{\mathrm{cm}}^{2}$. These results represent a significant increase in precision compared to previous work, and also resolve a discrepancy in previous quenching measurements.