Collimation of dense atomic beams by Swept Velocity Shelving

Technical Summary: Collimation of Dense Atomic Beams by Swept Velocity Shelving

Problem Statement
The generation of continuous, high-flux, and collimated atomic beams is a critical resource for metrology, quantum simulation, and material deposition. While optical molasses cooling has long been the standard for narrowing atomic velocity distributions, it encounters fundamental limitations when applied to dense atomic beams. In high-flux regimes, the high optical density of the atomic cloud leads to significant absorption of the counter-propagating cooling beams. This absorption creates an intensity imbalance between the beams, resulting in a force imbalance that degrades cooling efficiency. Consequently, the velocity distribution becomes broadened and shifted. Attempts to compensate by increasing laser intensity introduce power broadening, which raises the minimum achievable Doppler temperature and velocity spread, creating a self-defeating cycle. Furthermore, conventional molasses requires atoms to continuously scatter photons, leading to photon budget inefficiencies and recoil heating.

Methodology: Swept Velocity Shelving (SVS)
To overcome these limitations, the authors introduce a collimation scheme called Swept Velocity Shelving (SVS). Unlike optical molasses, which relies on balanced radiation pressure forces, SVS employs a one-sided force combined with conditional velocity selection. The method utilizes three distinct laser transitions on strontium atoms ($^{88}\text{Sr}$):

1. Sweeping Beam: A broadband, near-resonant transition ($^1S_0 \to {}^1P_1$) is used to exert a unidirectional force, decelerating atoms and sweeping their velocity distribution toward a target range.
2. Selection Beam: A narrowband transition ($^1S_0 \to {}^3P_1$) is used to selectively address atoms within a specific, narrow velocity class.
3. Shelving Beam: A third transition ($^3P_1 \to {}^3S_1$) optically pumps the selected atoms into a long-lived metastable reservoir state ($^3P_2$ or $^3P_0$).

The process operates in a cyclic, interleaved sequence. Atoms in the ground state are swept toward the target velocity. When an atom's velocity falls within the selection window, the selection beam excites it, and the shelving beam rapidly transfers it to the metastable reservoir. Once shelved, the atom no longer interacts with the laser fields, effectively removing it from the optical cycle. This "one-way wall" mechanism prevents re-absorption and power broadening of the cycling transition, as the selected atoms are immediately removed from the interaction.

Key Contributions

• Theoretical Framework: The authors developed a theoretical model and numerical simulations demonstrating that SVS can achieve velocity compression without suffering from absorption-induced force imbalances. The simulations show that the final velocity spread is determined primarily by the selection transition's linewidth and power broadening, rather than the atomic density.
• Experimental Implementation: The scheme was experimentally realized using a thermal strontium beam coupled to an optical cavity. The setup included a "cleaning" beam to remove residual ground-state atoms and a "push" beam to optimize the initial velocity distribution for the one-sided sweep.
• Characterization: The performance of the SVS scheme was quantitatively characterized using cavity normal-mode spectroscopy. By measuring the collective coupling strength and the effective atomic linewidth, the authors extracted the number of interacting atoms and the residual Doppler broadening along the cavity axis.

Results

• Velocity Narrowing: The SVS scheme successfully produced a collimated atomic beam with a programmable mean velocity and a narrow velocity spread. The experiment achieved a minimum effective Doppler linewidth of $2\pi \times (1.37 \pm 0.05)$ MHz at an optimal compromise where the interacting atom number was reduced by less than 15% compared to its maximum. This corresponds to a residual longitudinal velocity spread estimated at $0.7 \pm 0.1$ m/s. This result is consistent with the power-broadened linewidth of the selection transition.
• Density Independence: Unlike optical molasses, the SVS performance remained effective at high atomic densities ($n \approx 2 \times 10^{11} \text{ cm}^{-3}$). The simulations and experimental data confirmed that the velocity compression mechanism does not degrade due to absorption-induced force imbalances.
• Comparison with Molasses: When compared to previous results using conventional blue-detuned molasses cooling in a similar system, the SVS scheme produced a symmetric velocity distribution centered at zero velocity with no observable tilt-induced broadening. While the number of collectively coupled atoms in the SVS experiment ($N_C \approx 1.4 \times 10^5$) was lower than in the molasses case ($N_C \approx 4 \times 10^5$), the authors attribute this to the reduced interaction length in the current geometry rather than a fundamental limitation of the method. The SVS approach offers a scalable route to high-flux, narrow-velocity beams at higher oven temperatures where molasses cooling fails.

Significance and Claims
The paper claims that SVS provides a fundamentally different strategy for atomic beam collimation that circumvents the absorption limits of balanced-force methods. By decoupling the momentum control (sweeping) from the velocity selection and removal (shelving), the method achieves a state-heralded, collimated beam with high photon efficiency. The authors assert that this approach is broadly applicable to alkaline-earth and alkaline-earth-like atoms, offering a pathway toward narrow-velocity, high-flux atomic sources essential for continuous cavity-QED platforms, steady-state superradiant light sources, and precision metrology. The work demonstrates that high phase-space density can be achieved in dense beams without the trade-offs inherent in traditional Doppler cooling.