Quantum gases show flashes of a supersolid

Supersolids are highly sought-after structures whose atoms can simultaneously support frictionless flow and form a crystal. Hallmarks of a supersolid have now been observed in three experiments that involve quantum gases of dipolar atoms.
Lode Pollet is in the Faculty of Physics, Ludwig Maximilian University of Munich, 80333 Munich, Germany.

Search for this author in:

Sixty years ago, the theoretical physicist Eugene Gross suggested that a substance could have properties of both a solid and a liquid at the same time, provided that the liquid is a superfluid1. A superfluid is a state of matter that can flow without friction and is known to exist2,3 in helium-4 at temperatures below 2 kelvin. Gross’s putative substance was called a supersolid. But despite this theoretical simplicity, supersolids in the purest sense of the term have evaded experimental detection4. Now, Tanzi et al.5, writing in Physical Review Letters, and Böttcher et al.6 and Chomaz et al.7, writing in Physical Review X, report transient signatures of supersolidity in quantum gases of atoms that have strong magnetic dipole moments.

In Gross’s proposal, a supersolid is pictured as the superposition of a liquid and a periodic density variation. In other words, a supersolid comprises liquid droplets that consist of many atoms and that form a periodic structure (Fig. 1). Each droplet can be described by its number of atoms and a property known as a quantum-mechanical phase. In a supersolid, unlike in an ordinary solid, each droplet retains the same phase. Such phase rigidity requires the exchange of atoms between the droplets and is possible only if the droplets are sufficiently close to each other.

Figure 1 | Density distributions of three states of matter. a, A superfluid is a substance that can flow without friction. The density of a superfluid is uniform, apart from small fluctuations (as shown in this snapshot). b, Three papers57 report experimental evidence for a supersolid — a spatially ordered material that has superfluid properties. A supersolid comprises droplets that contain many atoms and that are coupled to each other. c, In these experiments, an ordinary solid consists of isolated droplets. The colours in ac represent the atomic density from low (light blue) to high (dark red).

Historically, supersolidity was sought in solid helium-4 using an apparently different, but formally equivalent, concept8. In this picture, a supersolid is a mostly crystalline substance in which certain defects enable a flow of adjacent atoms, which in turn sets the neighbours of these atoms in motion. This process continues until the whole crystal develops a fluid component. Despite some initial excitement9,10, pure supersolidity is not observed in solid helium-4. However, in this substance, related phenomena such as giant quantum plasticity11 are measurable and there is mounting evidence of frictionless flow along line-type defects12 called dislocations, as was first proposed by theorists13.

Over the past decade, cold-atom systems have shifted the focus back to Gross’s picture, because of the controllability and lack of defects and impurities in these systems. When atoms are cooled to temperatures near 0 K, they can form a state of matter called a Bose–Einstein condensate, which gives rise to frictionless flow. The difficulty in producing a supersolid then lies in imposing a periodic density variation that is set by the intrinsic interactions of atoms in these extremely dilute systems. This imposition takes place through a mechanism known as roton softening.

A roton is a minimum in the energy–momentum spectrum of a superfluid’s excitations. This minimum is located at a value of the momentum that is equal to the inverse of the average spacing between atoms. When the energy of the roton hits zero, the superfluid becomes unstable and forms a structure that has a periodic density variation. This structure could be, for instance, a supersolid or an ordinary solid. By contrast, alternative approaches for observing signatures of supersolidity have depended on external perturbations from lasers1416, rather than intrinsic properties of the system.

In the current experiments, Tanzi et al. and Böttcher et al. used dysprosium-162 atoms, whereas Chomaz et al. used dysprosium-164 and erbium-166 atoms. All of these atoms have intrinsically strong magnetic dipole moments. The interactions of these atoms have the theoretically required ingredients for supersolidity: a repulsive, tunable short-range (contact) component and an attractive, long-range (dipolar) component. Previously, some of the authors of the Böttcher et al. paper and their colleagues succeeded in producing a periodic density variation in a system of dysprosium-164 atoms17,18. But the droplets in the resulting state were too distant from each other, leading to the loss of frictionless flow.

However, it has been worked out theoretically19 that, under certain conditions, there is a narrow window in the ratio of dipolar-interaction strength to contact-interaction strength for which the droplets are situated close enough to each other to retain phase rigidity. By tuning an external magnetic field, which changes the way in which atoms scatter when they collide, the authors of the current papers reduced the strength of the contact interaction, bringing all three experiments into the desired parameter regime. The researchers then released the gases from the traps in which they were formed and allowed the matter waves associated with the atoms to interfere with each other. The resulting interference patterns contained a double-peak structure that is a hallmark of supersolidity.

In all the experiments, the peaks were transient phenomena because of three-body losses — losses of atoms that occur when a pair of atoms forms a bound molecular state with the aid of a third collision partner. The lifetimes of the supersolid properties ranged from a few tens of milliseconds for dysprosium-162 and erbium-166 atoms to 150 ms for dysprosium-164 atoms. For the latter atoms, the contact-interaction strength is smaller than the dipole-interaction strength. This feature makes a technically advantageous cooling protocol possible that avoids unwanted excitations and dynamics.

Current limitations of the studies are that each of the experiments involves only a handful of droplets, as well as a complex interplay between the droplets and the axially elongated (cigar-shaped) traps. Future studies should address these issues, aim for a direct manifestation of phase rigidity and study the excitations of a supersolid. Another convincing proof of supersolidity would involve letting a superfluid flow through a prospective supersolid — a situation that is not possible for ordinary fluids and solids.

Nature 569, 494-495 (2019)

doi: 10.1038/d41586-019-01585-w


  1. 1.

    Gross, E. P. Phys. Rev. 106, 161–162 (1957).

  2. 2.

    Kapitza, P. Nature 141, 74 (1938).

  3. 3.

    Allen, J. F. & Misener, A. D. Nature 141, 75 (1938).

  4. 4.

    Boninsegni, M. & Prokof’ev, N. V. Rev. Mod. Phys. 84, 759–776 (2012).

  5. 5.

    Tanzi, L. et al. Phys. Rev. Lett. 122, 130405 (2019).

  6. 6.

    Böttcher, F. et al. Phys. Rev. X 9, 011051 (2019).

  7. 7.

    Chomaz, L. et al. Phys. Rev. X 9, 021012 (2019).

  8. 8.

    Andreev, A. F. & Lifshitz, I. M. Sov. Phys. JETP 29, 1107–1113 (1969).

  9. 9.

    Kim, E. & Chan, M. W. H. Nature 427, 225–227 (2004).

  10. 10.

    Kim, E. & Chan, M. W. H. Science 305, 1941–1944 (2004).

  11. 11.

    Haziot, A. et al. Phys. Rev. Lett. 110, 035301 (2013).

  12. 12.

    Ray, M. W. & Hallock, R. B. Phys. Rev. Lett. 100, 235301 (2008).

  13. 13.

    Boninsegni, M. et al. Phys. Rev. Lett. 99, 035301 (2007).

  14. 14.

    Baumann, K., Guerlin, C., Brennecke, F. & Esslinger, T. Nature 464, 1301–1306 (2010).

  15. 15.

    Léonard, J., Morales, A., Zupancic, P., Esslinger, T. & Donner, T. Nature 543, 87–90 (2017).

  16. 16.

    Li, J.-R. et al. Nature 543, 91–94 (2017).

  17. 17.

    Kadau, H. et al. Nature 530, 194–197 (2016).

  18. 18.

    Wenzel, M., Böttcher, F., Langen, T., Ferrier-Barbut, I. & Pfau, T. Phys. Rev. A 96, 053630 (2017).

  19. 19.

    Roccuzzo, A. M. & Ancilotto, F. Phys. Rev. A 99, 041601 (2019).

Download references

Nature Briefing

Sign up for the daily Nature Briefing email newsletter

Stay up to date with what matters in science and why, handpicked from Nature and other publications worldwide.

Sign Up