Devices known as magneto-optical traps have long been used to cool and confine atoms, but not molecules — until now. This new ability should enable many studies and applications of the physics of ultracold molecules. See Letter p.286
On page 286 of this issue, Barry et al.1 report the first demonstration that diatomic molecules can be caught in a three-dimensional magneto-optical trap — a device that combines the effects of lasers and of a magnetic field to capture and cool particles. By using this approach, the authors reach the lowest temperature yet achieved for molecules using direct-cooling methods. This opens the way to the study of a plethora of fascinating phenomena in quantum physics, and might enable applications ranging from quantum information processing and simulations, to precision measurements and ultra-cold chemistry2.
Molecules — even simple diatomic molecules — are many times more complex than atoms. They possess a large number of internal states because of their electronic, vibrational and rotational degrees of freedom. Diatomic molecules can also carry an electric dipole moment and can undergo involved reactive processes, adding to their complexity. But it is precisely this complexity that makes molecules so interesting.
Tremendous progress has been made in cooling and slowing ensembles of molecules2 to study their quantum behaviour. However, despite intense experimental and theoretical studies, developing a versatile strategy to simultaneously deeply cool and trap molecules of various kinds remains a major challenge. With the approach of the 20th anniversary of the first realization of an atomic Bose–Einstein condensate3 — an ensemble of ultracold atoms that exhibits collective quantum behaviour — now is a good time to ask why molecules have been so much more difficult to capture and manipulate in cooling experiments.
The rapid development of ultracold-atom physics is certainly due in part to the success and robustness of laser cooling4,5. When it was first proved that massless photons could exert a radiation-pressure force on massive particles, such as sodium atoms, causing the atoms to slow down (that is, to be cooled), the full implications for atomic, molecular and optical physics were no doubt unseen. But in less than half a century, physicists have used this phenomenon in methods to manipulate, trap and cool an ever-increasing number of atomic species, and have also trapped ions to ultra-cold temperatures. Nowadays, every standard experiment with ultracold atoms employs a magneto-optical trap (MOT) to capture and cool atoms down to temperatures of roughly tens of microkelvins.
The MOT technique was originally proposed by the physicist Jean Dalibard in the 1980s (ref. 6), and was demonstrated in the laboratory soon after7. A typical MOT is realized using three pairs of counter-propagating light beams and a static quadrupole magnetic field. Its working principle is based on laser cooling, which involves an exchange of momentum between an atom and a photon in repeated cycles of photon absorption and emission. During a cycle, the atom is first excited to a high-energy state and then spontaneously decays back to its initial state. Although the cooling effect is extremely powerful, it does not provide spatial confinement. But in a magnetic-field gradient, the light field becomes position dependent, and generates a restoring force that gathers atoms around the zero-magnetic-field position.
So how does this technique apply to the cooling and trapping of molecules, with their complicated internal states? The resulting complexity is so great that finding transitions between states suitable for absorption–emission cycles seems, at first glance, to be a hopeless task. This prejudice has been partially broken down by the experimental demonstration of direct laser cooling of some simple diatomic molecules8,9,10, including two-dimensional transversal MOT cooling of a molecular beam9, and cooling of polyatomic molecules through a mechanism known as the opto-electric Sisyphus effect11.
Barry and colleagues' report of magneto-optical trapping of strontium monofluoride molecules represents a major advance in molecular cooling. Using their approach, complex molecules can be treated like atoms — in other words, the standard three-dimensional configuration of an atomic MOT can be applied to molecules.
The researchers' optical-cooling cycle uses a remarkably simple recycling scheme that involves just a few excited molecular states12 and only four light wavelengths. The trick is to use a manifold of rotational excited states that has fewer energy levels than the ground-state manifold. The molecules therefore have relatively few escape routes from the cooling cycle, so that, in this case, only three light wavelengths are needed to pump escaped atoms back into the cycle. The price paid for this approach is a weaker net radiation-pressure force than that in ordinary atomic MOTs, and thus fewer molecules collected in the trap. However, schemes and ideas of how to circumvent this problem have been proposed9.
Barry et al. report that, after 106 cycles of absorption and emission, their MOT contains about 300 molecules at a temperature of 2.5 millikelvins. These numbers, although certainly below the performance of ordinary MOTs of atoms (which typically trap 106 to 109 atoms at temperatures of tens to hundreds of microkelvins) prove for the first time that molecules can be cooled and trapped. Cooling decreases the range of velocities of molecules in an ensemble, whereas trapping clusters the molecules together. This combination provides a route to high-density molecular ensembles.
The magneto-optical trapping of molecules might have the same tremendous impact as its atomic counterpart, revolutionizing the field of molecular cooling. Barry and co-workers' strategy of applying atomic approaches to molecules is only the beginning. As noted earlier, molecules possess a feature absent in atoms: strong electric dipole moments. It is time to use this property at the lowest temperatures achievable in MOTs to trigger dipolar scattering, or to create a reservoir of electrically trapped molecules in metastable states to refill the MOT, as has been achieved for strongly magnetic atoms by exploiting their magnetic dipoles13.
Barry, J. F., McCarron, D. J., Norrgard, E. B., Steinecker, M. H. & DeMille, D. Nature 512, 286–289 (2014).
Carr, L., DeMille, D., Krems, R. & Ye, J. New J. Phys. 11, 055049 (2009).
Anderson, M. H., Ensher, J. R., Matthews, M. R., Wieman, C. E. & Cornell, E. A. Science 269, 198–201 (1995).
Hänsch, T. W. & Schawlow, A. Opt. Commun. 13, 68–69 (1975).
Cohen-Tannoudji, C. & Guéry-Odelin, D. Advances in Atomic Physics: An Overview (World Scientific, 2011).
Raab, E. L., Prentiss, M., Cable, A., Chu, S. & Pritchard, D. E. Phys. Rev. Lett. 59, 2631 (1987).
Shuman, E. S., Barry, J. F. & DeMille, D. Nature 467, 820–823 (2010).
Hummon, M. T. et al. Phys. Rev. Lett. 110, 143001 (2013).
Zhelyazkova, V. Phys. Rev. A 89, 053416 (2014).
Zeppenfeld, M. et al. Nature 491, 570–573 (2012).
Stuhl, B. K., Sawyer, B. C., Wang, D. & Ye, J. Phys. Rev. Lett. 101, 243002 (2008).
McClelland, J. & Hanssen, J. Phys. Rev. Lett. 96, 143005 (2006).