Quasi-continuous sub-$μ$K strontium source without a high-finesse cavity stabilized laser

Abstract

We demonstrate a quasi-continuous sub-$μ$K strontium source achieved without the use of a high-finesse cavity-locked laser. Our frequency reference is based on a dispersion-optimized, fiber-based frequency comb that enables sub-kHz linewidths. The long-term stability of the comb is defined by an external RF reference: either a 10 MHz RF signal from the Dutch Metrology Institute (VSL), or a tunable RF source whose long-term stability is maintained by monitoring and stabilizing the position of a narrow-line magneto-optical trap (MOT). The comb-stabilized system is benchmarked against a conventional cavity-locked laser and achieves comparable performance in broadband and single-frequency MOTs using the narrow $^1$S$_0$ $\rightarrow$ $^3$P$_1$ laser cooling transition. We generate high-flux, sub-$μ$K samples of all three bosonic strontium isotopes and demonstrate quasi-continuous outcoupling from the MOT. These results highlight the system's suitability for compact, robust, and field-deployable continuous cold atom devices.

Figures (8)

• Figure 1: Measured and theoretical sensitivity of the repetition rate $\frac{df_{\mathrm{r}}}{dP}$ (left axis) and carrier-envelope offset frequency $\frac{df_{\mathrm{CEO}}}{dP}$ (right axis) as a function of pump power. Experimental data for $\frac{df_{\mathrm{r}}}{dP}$ (red points) is compared with theoretical contributions from spectral shift (dashed), third-order dispersion (TOD, dot-dashed), resonant gain (doted), and self-steepening (SSD, dot-dot-dashed), with their total shown as a solid red line. Measured $\frac{df_{\mathrm{CEO}}}{dP}$ values are shown as blue triangle. The theoretical repetition rate sensitivity crosses zero at a pump power of 160.27 mW (vertical red dashed line), while the theoretical carrier-envelope offset sensitivity crosses zero at 160.21 mW (vertical blue dashed line).
• Figure 2: Calculated and experimental results for the pump-induced fixed point frequency (left axis, orange curve) and optical linewidth at 689 nm (right axis, purple points) as a function of pump power. The orange curve shows the calculated fixed point frequency, while the purple points represent the experimentally measured linewidth values, averaged over 10 independent measurements per point. The purple line corresponds to the theoretical linewidth prediction. An optimal operating point at 161 mW can be found where the fixed point (horizontal dashed line at 435 THz) aligns with the atomic transition frequency, and the linewidth reaches a minimum below 1 kHz.
• Figure 3: (a) Experimental schematic. Strontium atoms exit the oven and propagate in the -$\hat{x}$ direction. Transverse cooling beams operating on the $^1$S$_0\,\rightarrow\,^1$P$_1$ transition collimate the atomic beam along the $\hat{y}$ and $\hat{z}$ axes. The atoms are slowed by a Zeeman slower before arriving in a 2D blue MOT, both operating on the $^1$S$_0\,\rightarrow\,^1$P$_1$ transition. Atoms then fall from the blue MOT down into a lower chamber due to gravity, where they are captured in a 3D MOT operating on the $^1$S$_0\,\rightarrow\,^3$P$_1$ transition. Push beams on the $^1$S$_0\,\rightarrow\,^1$P$_1$ transition and red molasses beams addressing the $^1$S$_0\,\rightarrow\,^3$P$_1$ transition help to minimize loss of atoms between the blue and red MOTs. Atoms are captured in the BB (zoomed image, bottom) and SF (zoomed image, top) MOT simultaneously in a "dual-position MOT" (see Section \ref{['sec:exp_setup:quasi-continuous']}). (b) Energy level diagram for strontium showing the blue and red cooling transitions.
• Figure 4: Stability of the position of the BB red MOT over 900 s for the cavity-stabilized system long-term referenced to hot vapor spectroscopy (dark red dots) vs. the comb-stabilized system referenced to a 10 MHz signal from VSL (orange dots) and to the MOT position (light orange dots). The standard position deviation for the three stabilization methods over 900 s are $\sigma _{\mathrm{cavity}}=27.1~\mu$m, $\sigma _{\mathrm{VSL}}=60.73~\mu$m, and $\sigma _{\mathrm{active}}=83.6~\mu$m, respectively. All measurements are taken at a small magnetic field gradient ($0.39$ G/cm) to magnify the differences.
• Figure 5: Atom number over time during loading of a BB MOT for the three bosonic isotopes, $^{88}$Sr (blue stars), $^{86}$Sr (green crosses), and $^{84}$Sr (orange triangles), using the comb-stabilized laser system with the 10 MHz reference from VSL. The loading data for the cavity-stabilized system and $^{88}$Sr is shown as a reference (red dots). All data is fitted to the following loading curve: $N(t)=N_{\mathrm{sat}}(1-e^{-t/\tau})$ and plotted with a solid (dashed) line for the comb (cavity)-stabilized system.