Sub-Doppler cooling, state preparation, and optical trapping of a triel atom

TL;DR

This work establishes indium, a triel element, as a viable platform for ultracold quantum science by achieving sub-Doppler cooling to $15\,\mu\mathrm{K}$, spin-polarized quantum-state preparation with ~90% purity for both the $5\mathrm{P}_{3/2}$ and $5\mathrm{P}_{1/2}$ manifolds, and stable optical trapping in a $1064\,\mathrm{nm}$ lattice with multi-second lifetimes. Sub-Doppler cooling relies on polarization gradient cooling (PGC) optimized for large hyperfine splittings, with residual magnetic fields quantified and canceled via microwave spectroscopy of a ground-state hyperfine transition. State preparation combines optical pumping and magnetic trapping to isolate single $m_F$ sublevels, achieving high purity and substantial increases in polarized populations. Optical trapping leverages intensity modulation to load a 1D lattice, accounting for the polarizability landscape of indium’s states, and revealing state-dependent loss processes. Collectively, these results position indium as a promising candidate for future quantum simulations, clock-based probes, and spin-orbit-coupled many-body systems in optical lattices or tweezer arrays.

Abstract

Ultracold gases of atoms from Main Group III (Group 13) of the Periodic Table, also known as "triel elements," have great potential for a new generation of quantum matter experiments. The first magneto-optical trap of a triel element (indium) was recently realized, but more progress is needed before a triel is ready for modern ultracold quantum science experiments in optical traps. Reaching this regime typically requires atoms that are cooled to the 10 uK level or below, prepared in pure quantum states, and confined in a laser field. Here we report the achievement of all three of these milestones in atomic indium. First, we perform polarization gradient cooling of an indium gas to 15 uK. Second, we spin polarize the gas into a single hyperfine sublevel of either the $5P_{1/2}$ indium ground state or the $5P_{3/2}$ metastable state. Third, we trap indium in a 1064 nm optical lattice, achieving a 3 s trap lifetime. With these results, indium is now a candidate for a next generation quantum research platform.

Figures (6)

• Figure 1: (a) Indium energy levels used for laser cooling. This scheme is presented in more detail in our previous work Yu2022b. The laser cooling transition is $\ket{5P_{3/2}, F=6} \rightarrow \ket{5D_{5/2}, F=7}$ throughout this manuscript. The frequencies in the figure are the natural linewidths of the respective transitions. (b) Atomic transitions used for spin polarization. Two spin polarization configurations are possible, allowing us to polarize atoms into either $\ket{5P_{1/2}, F=4, m_F=4}$ or $\ket{5P_{3/2}, F=6, m_F=6}$. The purity of spin polarization is determined with microwave spectroscopy.
• Figure 2: Left: Typical microwave spectra with decreasing external fields. Panel (c) shows the smallest Zeeman splitting we observed. A theoretical model of the spectrum explains the broadening of lines with larger $\abs{m_F}$ (the features that are farthest detuned from zero) as a result of a dynamic residual field supplemental-material. Right: The residual field (blue circles) and the measured temperature of the atoms (red triangles). The residual field was determined by the frequency difference between microwave resonances. The difference in applied field minima between the residual field and the temperature is because the two curves were measured at different dynamic residual field values supplemental-material.
• Figure 3: Characterization of polarization gradient cooling. Left: Temperature vs. detuning, indicating sub-Doppler scaling in the temperature. The data is fit to the sub-Doppler cooling law $T = T_0 + T_1 (\Gamma/|\Delta|)$, where $T_0 = 7.0 \ \mathrm{\mu K}$ and $T_1 = 65.8 \ \mathrm{\mu K}$. In practice, the lowest temperature achieved is $15\,\mathrm{\mu K}$. Middle: Temperature vs. PGC pulse duration, which optimizes at $5\,\mathord{\mathrm{ms}}$. Right: Temperature vs. laser intensity, which shows the best performance at our maximum power.
• Figure 4: Left: Atom number as a function of the field gradient magnitude along the direction of gravity. Green data (left y-ais) was taken with optical pumping; blue data (right y-axis) was taken without optical pumping. The step-like features McCarron2011 occur when the force of gravity is equal to the confining force of the gradient for a given spin component. The atom number in the polarized component ($m_F=6$) is $1.3\times10^5$ ($6.8\times10^3$) with (without) optical pumping, so optical pumping boosts polarized atom number by a factor of 19. The annotations (a) through (d) indicate the points at which we measured the four plots [also labeled (a) through (d)] in the right panel. Right: In (a), several oscillation frequencies are apparent from the $m_F = 3,4,5,6$ spin components. In (d), only one oscillation frequency remains, indicating a spin polarized gas in the $m_F = 6$ state. The figure numbers (a) through (d) correspond to the measurement values (both field gradient and atom number) indicated in the left panel. The color bar represents the atomic density integrated along the $x$ and $y$ directions.
• Figure 5: LEFT: Microwave spectra at different stages of the $5P_{1/2}$ spin polarization sequence. The top plot shows the microwave spectrum before spin polarization is applied, indicating many spin components. The second plot down shows the $5P_{1/2}$ population after spin polarization using the $\ket{5P_{3/2},F=6} \rightarrow \ket{5D_{5/2},F=6}$ transition. The third plot down shows the effect of holding atoms in the quadrupole trap to remove spin components with $m_F \leq 0$. The bottom plot shows the microwave spectrum after spin polarization with a $\sigma^+$-polarized 410 nm laser. The three prominent peaks are expected when the atoms are polarized in the target state, and they imply quantum state purity at the 90% level. RIGHT: The transitions corresponding to those labeled in the observed microwave spectra.