Observation of superfluorescent emissions from laser-cooled atoms

E. Paradis, B. Barrett, A. Kumarakrishnan, R. Zhang, and G. Raithel

Phys. Rev. A 77, 043419 – Published 25 April 2008

Abstract

We study superfluorescence (SF) from spherical and cigar-shaped clouds of laser-cooled Rubidium atoms from the $5 D_{5 / 2}$ level through the $6 P_{3 / 2}$ level to the $5 S_{1 / 2}$ ground level. The atomic system is initially excited to the $5 D_{5 / 2}$ level from the ground state via two-photon excitation through the intermediate $5 P_{3 / 2}$ level. The fluorescence on the $6 P − 5 S$ transition at 420 nm is recorded using time-resolved measurements. The time delays of the observed SF emission peaks typically scale as $\sim N^{- 1}$ , where $N$ is the atom number, and are much smaller than the time delay expected for uncorrelated cascade fluorescence. Since $N$ is significantly smaller than the threshold number for SF on the 420 nm transition, and larger than the threshold number for the $5 D − 6 P$ transition at $5.2 μ m$ , our observations suggest that the 420 nm SF emission is triggered by rapid deexcitation of the $5 D$ to the $6 P$ level via SF at $5.2 μ m$ . The observed SF time delays for 420 nm emission agree with SF time-delay estimates for the $5.2 μ m$ transition. For spherical clouds, the SF is isotropic. For cigar-shaped clouds, the SF is highly anisotropic. Along the long axis of cigar-shaped atom clouds, SF and incoherent cascade fluorescence produce temporally well-resolved peaks in the detected signal. In this case, the SF component of the signal is highly concentrated along a direction in between the directions of the two almost parallel excitation beams. The observed SF intensities scale as $N$ , suggesting that the $5 D$ level is regeneratively pumped during the SF decay.

Received 12 November 2007

DOI:https://doi.org/10.1103/PhysRevA.77.043419

Authors & Affiliations

E. Paradis, B. Barrett, and A. Kumarakrishnan

Department of Physics and Astronomy, York University Toronto, ON M3J 1P3 Toronto, Canada

R. Zhang and G. Raithel

Department of Physics, University of Michigan, Ann Arbor, Michigan 48109, USA

Article Text (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 77, Iss. 4 — April 2008

Reuse & Permissions

Access Options

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
(Color online) Rubidium 87 level diagram: the solid lines represent excitation pulses, and the dashed lines correspond to SF emissions pertaining to this work. The dotted lines represent the expected SF cascade for incoherent excitation. The radiative lifetimes of the $5 D_{5 / 2} \to 6 P_{3 / 2}$ , $6 P_{3 / 2} \to 5 S_{1 / 2}$ transitions are 690 and 357 ns, respectively. The total radiative lifetimes of the $5 D_{5 / 2}$ and $6 P_{3 / 2}$ states are 241 and 109 ns, respectively.Reuse & Permissions
Figure 2
(Color online) Diagram of experimental setup. SF emissions at 420 nm can be measured both perpendicular to and along the direction of the excitation pulses. ( $ν_{1}$ denotes the frequency of the $5 S_{1 / 2}, F = 2 \to 5 P_{3 / 2}, F = 3$ transition, and $ν_{2}$ that of the $5 P_{3 / 2}, F = 3 \to 5 D_{5 / 2}, F = 4$ transition.)Reuse & Permissions
Figure 3
Time-resolved SF detected along the long axis of cigar-shaped atom clouds for various ground-state atom numbers $N$ . The excitation pulses are overlapped, and time $t = 0$ corresponds to the center of the excitation pulses. The atom number $N$ increases from the bottom to the top curve. The dashed curve indicates the time-dependence of the excitation pulses. The inset shows the height of the two maxima observed in the time-resolved emission signals (peak 1 and peak 2) for a fixed number of atoms, as a function of the delay time between the two excitation pulses. (Delay time $< 0$ corresponds to STIRAP excitation pulse ordering, while delay time $> 0$ corresponds to sequential excitation pulse ordering.)Reuse & Permissions
Figure 4
Calculated $5 D$ excitation probability (solid) and $5 S − 5 D$ coherence (dashed) versus delay time between excitation pulses for typical experimental parameters. Negative delays correspond to STIRAP excitation.Reuse & Permissions
Figure 5
(Color online) Delay time of SF vs number of atoms. The delays marked peak 1 (solid diamonds) and peak 2 (empty diamonds) correspond to temporally resolved peaks, as shown in Fig. 3, measured along the long axis of the cigar-shaped cloud. Delays for the spherical trap are shown with triangles. For the case of the cigar and the sphere, the experimental data are fitted with functions of the type $a N^{- x} + b$ with fit parameters $a$ , $b$ , and $x$ , shown as solid lines. The fits show that the delays scale as the expected, with $x = (1.12 \pm 0.16)$ for the cigar and $(1.17 \pm 0.14)$ for the sphere, respectively. The offset $b$ is zero for the cigar and $\sim 70 ns$ for the sphere. The dashed line shows the predicted $5.2 μ m$ delay, calculated on the basis of Eq. (1) using measured cloud parameters. Numerical simulations for cascade fluorescence suggest that peak 2 should occur at 174 ns.Reuse & Permissions
Figure 6
(Color online) Peak intensity as a function of number of trapped atoms for the long axis of the cigar-shaped trap, short axis of the cigar-shaped trap, and for a spherical trap. The graph for the case of the long axis of the cigar-shaped trap has been scaled down by a factor of 10. The slopes $m$ based on linear fits are indicated.Reuse & Permissions

Physical Review A

covering atomic, molecular, and optical physics and quantum information