Efficient all-optical production of large ${}^6\mathrm{Li}$ quantum gases using $D_1$ gray-molasses cooling

We use a gray molasses operating on the $D_1$ atomic transition to produce degenerate quantum gases of ${}^6\mathrm{Li}$ with a large number of atoms. This sub-Doppler cooling phase allows us to lower the initial temperature of $10{}^9$ atoms from 500 to 40 $\mu \mathrm{K}$ in 2 ms. We observe that $D_1$ cooling remains effective into a high-intensity infrared dipole trap where two-state mixtures are evaporated to reach the degenerate regime. We produce molecular Bose-Einstein condensates of up to 5 $\times$ $10{}^5$ molecules and weakly interacting degenerate Fermi gases of $7\times {10}^5$ atoms at $T/T_F<0.1$ with a typical experimental duty cycle of 11 s.

Figure 1

Level scheme (not to scale) for ${}^6\mathrm{Li}$ showing the $D_1$ hyperfine structure and the transitions used for the gray molasses. The laser detunings of the repumper and cooling lights are ${\delta }_1$ and ${\delta }_2$, while their relative detuning is $\delta$.

Figure 2

(a) Temperature $T$ (blue squares) and cooled fraction $N/N_0$ (red circles) after 2 ms of gray molasses vs the relative detuning $\delta$ in units of $\Gamma$ with ${\delta }_2=5.4\Gamma$, $I_{\text{cool}}=2.7I_S$, and $I_{\text{rep}}=0.5I_S$. (b) $N/N_0$ vs the initial cloud temperature $T_0$ at $\delta =0$. Clouds with $T_0<200$ $\mu \mathrm{K}$ are produced after 2 ms of gray molasses with parameters away from the Raman condition. The dotted line is only a guide for the eye. (c) $T$ (blue squares) and $N/N_0$ (red circles) vs the absolute detuning ${\delta }_2$, keeping $\delta =0$. (d) Same as (a), for a 300-$\mu \mathrm{s}$ molasses stage applied to atoms loaded into the optical trap at $P=120$ W (see text for details). In (a) and (d) the dotted line indicates the value $\delta =0$. In all the plots, the error bars are one standard deviation of five independent measurements.

Figure 3

(a) Total atom number $N$ (red circles) and temperature $T$ (blue squares) of the atoms during forced evaporation in the optical trap at $B=800$ G. The dotted line marks the evaporation time for which $T/T_c=1$. (b) Total atom number $N$ (blue squares) and temperature $T$ (red circles) of the atoms during forced evaporation in the optical trap at $B=300$ G. The dotted line indicates the evaporation time for which $T/T_F=1$. Above this value, the temperature is measured after expansion close to the zero crossing, while $T_F$ is determined by the mean value of N and the measured trap frequencies. For $T/T_F<1$ a surface fit to a polylog function is used to determine $T/T_F$. (c) $T/T_F$ vs $N$ approaching the end of the evaporation ramp, where $T/T_F=0.06\left(1\right)$. (d) Comparison between a Gaussian (black line) and the finite-temperature Fermi distribution fit (gray line) for a typical experimental profile corresponding to $T/T_F=0.06\left(1\right)$. In (a)–(c) error bars account for both statistical (five independent measurements) and $10\%$ systematic uncertainties.