Polaritons are promising constituents of both synthetic quantum matter1 and quantum information processors2, whose properties emerge from their components: from light, polaritons draw fast dynamics and ease of transport; from matter, they inherit the ability to collide with one another. Cavity polaritons are particularly promising as they may be confined and subjected to synthetic magnetic fields controlled by cavity geometry3, and furthermore they benefit from increased robustness due to the cavity enhancement in light–matter coupling. Nonetheless, until now, cavity polaritons have operated only in a weakly interacting mean-field regime4,5. Here we demonstrate strong interactions between individual cavity polaritons enabled by employing highly excited Rydberg atoms as the matter component of the polaritons. We assemble a quantum dot composed of approximately 150 strongly interacting Rydberg-dressed 87Rb atoms in a cavity, and observe blockaded transport of photons through it. We further observe coherent photon tunnelling oscillations, demonstrating that the dot is zero-dimensional. This work establishes the cavity Rydberg polariton as a candidate qubit in a photonic information processor and, by employing multiple resonator modes as the spatial degrees of freedom of a photonic particle, the primary ingredient to form photonic quantum matter6.


Recent years have seen a growing effort to create matter from light1,7. The speed of the photon may be employed to realize reduced-dimensional models with near-arbitrary single-particle dynamics, and recent breakthroughs have even enabled the creation of synthetic gauge fields that impose a Lorentz-like force on the photon8,9,10. While a number of groundbreaking experiments have probed the physics of photon condensation4,11, emergent crystallinity from long-range interactions12,13,14, and opto-mechanics15, the challenge of achieving strong interactions between individual photons for exploration of entangled photonic quantum matter persists. The demonstration of a single-photon nonlinearity mediated by a single atom coupled to a Fabry–Pérot cavity16 spurred an outpouring of theoretical work exploring Hubbard physics in coupled nonlinear resonators17,18, but the severe technical challenges associated with the deterministic preparation of individual atoms in resonator arrays remain.

Two major efforts have begun to address the technical challenges of mediating strong photonic interactions using atoms. In the first, researchers successfully demonstrated the loading of an individual atom into the near-field of a photonic crystal cavity19, and this capability was subsequently extended to the coupling of atomic ensembles to waveguides20. In the second, the challenges associated with trapping and cooling single atoms within optical resonators are eliminated by using Rydberg electromagnetically induced transparency (EIT)21. With Rydberg EIT, an entire thermal gas of atoms may act as a single ‘super atom’ that has the cross-section of the ensemble and the nonlinearity of a single atom22,23,24,25,26. Nonetheless, the propagation of these free-space excitations complicates their application to quantum information processors or materials compared to their cavity-confined counterparts.

Here, we present an approach combining Rydberg EIT and optical resonators27,28, which benefits from the best aspects of each previous approach: Rydberg EIT provides a robust atomic-ensemble-mediated single-photon nonlinearity, while the resonator enhances the light–matter coupling and allows us to work at substantially lower atomic densities while maintaining a high optical depth per blockade volume (ODB). This quantity is crucial, because ODB determines the number of collisions per polariton lifetime in a synthetic quantum material and the gate fidelity in a photonic quantum information processor29 (see Supplementary Information B). Perhaps most significantly, in the future, a quantum nonlinear resonator may be tuned near a mode-degeneracy point where interactions couple the transverse modes and the system behaves as a continuum-equivalent of an array of tunnel-coupled resonators. This approach would enable the creation of a strongly interacting quantum fluid within an individual, multi-mode cavity. Furthermore, the fluid can be trapped and subjected to strong magnetic fields3, providing access to the physics of photon crystals and Laughlin puddles6.

Here, we use this platform to demonstrate the first strongly interacting polaritonic quantum dot. We create a discrete mode spectrum by confinement of the photon in a cavity and strong interactions by tight confinement of the assembled atom cloud. In what follows, we describe the quantum dot in more detail and show that it exhibits the characteristic features of strong nonlinearity at the single-photon level, in contrast to related works exhibiting only weak nonlinearities5,30. First, we demonstrate strong light–matter coupling by measuring a well-resolved dark-polariton resonance in the transmission spectrum. We next explore transport via the statistics of photons tunnelling through the cavity, observing deep anti-bunching indicative of blockade due to strong interactions between individual polaritons. We further demonstrate that our polaritonic dot is zero-dimensional by detecting coherent Rabi oscillations between an empty dot and a dot containing a single polariton, as well as saturation of the intracavity dark polariton number at 0.47(3). We conclude with a discussion of applications in quantum information and strongly correlated, topological phases of matter.

Our quantum dot is formed from a sample of 2,600(500) atoms loaded into the 12 μm × 14 μm waist of a high-finesse single-mode optical resonator (Fig. 1a,c, and Supplementary Information D for details), at a peak density of \(1\times 1{0}^{11}\) cm−3 (see Methods in Supplementary Information A). These atoms are distributed over a 35 μm r.m.s. axial length that may be reduced to 10 μm r.m.s. by spatially selective optical depumping, or ‘slicing’ (see Supplementary Information C). Due to strong light–matter coupling (\({{\rm{OD}}}_{{\rm{B}}}\approx 13\)), the modes of the system hybridize, forming polaritons (Fig. 1b): composite states of a resonator photon with an atomic excitation. Two of the hybridized modes are ‘bright’ polaritons, composed primarily of an admixture of a resonator photon and a short-lived P-state excitation; they are weakly interacting due to their small Rydberg component and short lifetime. The third polariton, called the ‘dark’ polariton, is composed of a coherent superposition of a Rydberg excitation and a resonator photon, with the mixing ratio set by θ, the dark-state rotation angle31. For our typical conditions, \({\rm{\tan }}\theta =\sqrt{N}g/\Omega =2.9(5)\), where N, g and Ω are the atom number, single-atom Rabi frequency and control-field Rabi frquency, respectively. The Rydberg component of the dark polariton enables it to strongly repel other dark polaritons within a surrounding ‘blockade volume’, which scales with \({n}^{11/2}\) (for van der Waals interactions), where n is the Rydberg level’s principal quantum number. It is for this reason that we typically employ the \(n=100\) Rydberg state, as the atoms' extreme properties make this blockade volume comparable to the sample size. Confining the region of overlap between the resonator mode and the atomic gas to be smaller than the blockade volume gives rise to a zero-dimensional, strongly interacting polaritonic quantum dot; like its solid-state counterpart, the electronic quantum dot32, it is expected to exhibit blockaded transport (Fig. 1d).

Fig. 1: Rydberg polariton blockade in an optical resonator.
Fig. 1

Photons in an optical resonator are hybridized with Rydberg excitations via EIT, inducing strong photon–photon interactions. The coupled atom–cavity system thus becomes a single two-level emitter. a, As photons do not naturally interact with one another, we employ atoms to mediate interactions. An optical resonator couples 780 nm photons (red) to a sample of N ≈ 150 87Rb atoms, creating a collective \(\sqrt{N}g\) enhancement in the single atom–photon Rabi frequency, g. A 480 nm ‘control’ laser (blue) coherently couples excited atoms to the n = 100 Rydberg state with Rabi frequency Ω. Rydberg atoms interact strongly with one another, and the light–matter coupling imprints these interactions on the 780 nm photons. b, When we probe the spectrum of this strongly coupled light–matter system, we observe three distinct peaks corresponding to the three quasi-particle eigenstates, called polaritons: two broad ‘bright’ polaritons, with a linewidth set by the excited-state spontaneous decay rate Γ, and one narrow, tall ‘dark’ polariton in the middle, with a loss rate γD independent of Γ. The dark polariton is the longest-lived and most Rydberg-like quasi-particle, so it provides the best platform for mediating interactions between photons. c, Atoms are loaded into the smallest waist of the resonator and ‘sliced’ (see Supplementary Information C) so that the Rydberg components of the polaritons all reside within a single blockade volume and thereby strongly repel each other. Signatures of this interaction are then observed in the 780 nm photons that leak out through one of the resonator mirrors, pass through a 50:50 beamsplitter (BS) and are measured by single-photon detectors (D1 and D2). d, The system behaves as a zero-dimensional strongly interacting quantum dot. The weak probe field is in a coherent state \(| \alpha \rangle\), represented by the red curve, which consists primarily of zero photons, occasionally one photon and infrequently two photons, when counts are binned into a time \(1/{\gamma }_{{\rm{D}}}\). Each photon enters the cavity as a dark polariton, and strong Rydberg–Rydberg interactions cause the first polariton to broaden and shift the dark polariton resonance sufficiently to preclude simultaneous injection of a second polariton; accordingly, only isolated photons emerge from the cavity.

To investigate the polaritonic excitations of our quantum dot, we first probe the transmission spectrum of the laser-dressed cavity–atom system. In Fig. 1b, we plot the quantum dot’s transmission spectrum on the \(\{5{S}_{1/2},\,F=2\}\to \{5{P}_{3/2},\,F=3\}\to \{100{S}_{1/2}\}\) transition, without slicing the cloud. We observe two broad resonances that we identify as bright polaritons, whose spectral separation is an indication of strong light–matter coupling. The central feature is the spectrally narrow dark polariton. As we increase the probe power, the fractional transmission on the dark polariton resonance drops, as shown in Fig. 2, indicating that the repulsion between these polaritons suppresses their simultaneous injection into the dot.

Fig. 2: Nonlinear spectroscopy of a polaritonic quantum dot.
Fig. 2

The quantum dot’s transmission versus probe frequency is plotted for increasing incident photon rates. Here, the sample is unsliced and thus contains several (~3) blockade volumes, resulting in weak, nearly mean-field interactions. We thus observe smooth broadening and transmission suppression of the dark polariton peak (at \(\delta =0\) MHz). The bright polariton resonances at \(\delta =\pm 7\) MHz are not suppressed at a high probe rate, because bright polaritons have little Rydberg admixture and are thus essentially non-interacting. The data for the three lowest powers are Gaussian-filtered (with \({\sigma }_{{\rm{filter}}}=481\) kHz) outside the interval (−1.5 MHz, 1.5 MHz); all data are normalized to the bare-cavity transmission.

To demonstrate that individual dark polaritons interact strongly with one another within the quantum dot, we explore transport dynamics. We test transport by injecting photons into the resonator at the energy of the dark polariton, such that most photons become dark polaritons on entering. The strong repulsion between polaritons shifts the energy and reduces the lifetime of a second polariton in the dot, thereby precluding its injection. In an electronic quantum dot, the analogous Coulomb-blockade physics is typically ascertained from the dependence of transport on bias voltage32; by contrast, we directly observe suppression of simultaneous polariton transit by detecting when photons tunnel through the dot. We characterize the tunnelling times via the temporal intensity autocorrelation function, \({g}_{2}(\tau )\), where τ is the delay time, of the cavity transmission. This function compares the rate of photon pairs leaking out of the resonator separated by a time τ to a Poisson process with the same average rate. A reduction near zero delay (\(\tau =0\)) indicates interaction-driven suppression of double-occupancy of the dot, while the rise time back to steady state is dominated by the injection Rabi frequency.

After slicing the cloud so that it can hold only a single polariton at a time, we observe almost complete suppression (\({g}_{2}(0)=0.27(8)\)) of simultaneous tunnelling through the dot, indicative of transport blockade due to strong interactions (Fig. 3a). Indeed, strong anti-bunching at zero-time separation validates a model where a single intracavity polariton shifts and broadens the energy for the injection of the next polariton by more than the polariton linewidth, strongly suppressing its tunnelling into the resonator until the first polariton tunnels out. The observed \({g}_{2}(\tau )\) is consistent with a master-equation effective theory bootstrapped by a full model of interacting three-level atoms coupled to a single resonator mode (see Supplementary Information J). Note that the linear growth of \({g}_{2}(\tau )\) near zero delay, as opposed to the quadratic growth anticipated from a single-mode theory, arises from virtual excitation of bright polaritons.

Fig. 3: Transport blockade of cavity Rydberg polaritons.
Fig. 3

The photons transmitted through the cavity exhibit a strong anti-bunching/bunching effect, showing an interaction on the single quantum level. a, A smoking gun of strong interactions in a quantum dot is blockade: the suppression of near-simultaneous particle transmission. The temporal intensity autocorrelation function \({g}_{2}(\tau )\) on the dark-polariton resonance with a sliced atomic cloud reveals strong blockade: an ~5× suppression of two-photon events relative to the long-time value shown in the inset (blue curve, \({g}_{2}=0.27(8)\) near \(\tau =0\) compared to \({g}_{2}=1.3\) at long times). The green dashed line and band are the theoretically predicted minimum value and time dynamics, respectively (see Supplementary Information J–L). The unsliced cloud contains numerous blockade volumes; accordingly, it provides a weaker suppression of g2 near \(\tau =0\) (grey). In the inset, the time bins for \(| \tau | > 0.8\) μs are larger by a factor of 4 to reveal large τ trends. b, On the dark-polariton resonance of the 85S1/2 Rydberg state (blue), the suppression of g2 is reduced because multiple blockade volumes fit in the transverse plane. On the bright polariton resonance at 100S with a sliced cloud (orange), polaritons are very weakly interacting, resulting in uncorrelated transmission. c, We observe bunching (\({g}_{2}(0) > 0\)) when probing near a two-polariton resonance, achieved by shifting the cavity resonance by −5 MHz, increasing the control frequency by 5 MHz and probing −1 MHz from the single-polariton resonance. In each panel, the large-τ value of g2 exceeds 1 due to atom number fluctuations, which lead to a variation in the on-resonance transmission.

We verify that the observed transport blockade results from strongly interacting polaritons by comparing the observations to transport with weaker interactions. First, we test the unsliced cloud (Fig. 3a), which exhibits only weak suppression of g2 near \(\tau =0\) because the cloud is large enough to hold multiple polaritons along the resonator axis simultaneously. Similarly, a sliced sample with Rydberg atoms in the smaller 85S state can fit multiple excitations transversely and also exhibits weak suppression of g2 (Fig. 3b). Furthermore, we find that transport of bright polaritons is not blockaded at all (Fig. 3b), as these polaritons are essentially non-interacting33. The blockade of the dark polaritons can even be `reversed' by detuning from the atomic 5P state to make the polariton–polariton scattering more elastic and probing transport on the two-polariton resonance (Fig. 3c). In this case, we observe photon bunching, \({g}_{2} > 1\), reflecting the preferential transit of photon pairs34.

To investigate the dimensionality of the polaritonic dot, we probe the cavity with pulses of substantially higher intensity and observe the transmitted light as the intracavity field rings up and then down. A zero-dimensional system may be coherently excited and then de-excited by the drive field, exhibiting Rabi oscillations, while in a one-dimensional system the excitation would propagate away before it can be removed. Figure 4a shows the detuning dependence of the ring-up process, exhibiting rapid, detuned Rabi oscillations away from the dark-polariton resonance, and slower oscillations on-resonance. Figure 4b shows the ring-up dynamics on the dark-polariton resonance for various probe intensities. We observe Rabi oscillations between zero and one dark polaritons at the highest intensities, indicative of a strongly blockaded, zero-dimensional dot interacting with many photons within the polariton lifetime. The dark-polariton oscillations exhibit a Rabi frequency of \({\Omega }_{{\rm{polariton}}}=2{\rm{\pi }}\times 245\) kHz, in agreement with a first-principles calculation based on the probe power (see Supplementary Information G). Note that the fast ~100 ns oscillations arise from off-resonant excitation of non-interacting bright polaritons.

Fig. 4: Ring-up and ring-down dynamics of a polaritonic quantum dot.
Fig. 4

When the dot is simultaneously excited with many photons within the dark-polariton lifetime, these photons stimulate coherent tunnelling of a polariton into and out of the dot until the dot finally equilibrates. a, We drive the resonator near the dark-polariton resonance for 2.5 μs and record the transmitted intensity during the equilibration process. As the probe frequency is tuned away from the resonance, the oscillation frequency increases and the visibility decreases, resulting in a `chevron' of damped Rabi oscillations. b, With increasing probe strength, the overall transmitted power increases as well as the frequency of the occupation oscillation. The slow dynamics (microseconds timescale) reflect the tunnelling oscillations between zero and one dark polariton, while the fast dynamics (on a ~100 ns timescale) reflect off-resonant excitation of the broad bright polaritons. Bright polaritons are weakly interacting and do not saturate with increasing probe power. Accordingly, they produce a dominant background at the highest probe powers (nearly 50% of the observed signal). c, To isolate the physics of the dark-polariton saturation from the bright-polariton background, we explore the ring-down of the polaritons once the probe field is removed. The first ~100 ns of ring-down is dominated by the fast decay of bright polaritons. At later times, the cavity leakage is dominated entirely by the dark polaritons, with a lifetime substantially longer than the bare resonator (grey curve). The colour code is the same as in b; the relative transmission drops as the incident photon rate increases due to the blockade. d, We extract the steady-state dark-polariton number by extrapolating the slow decay of cavity emission to the beginning of the ring-down. The dark-polariton number saturates strongly at the highest probe power to \({n}_{{\rm{D}}}=0.47(3)\) (see Supplementary Information H).

Strong interactions should cause the number of dark polaritons in the dot to saturate at large probe powers. However, the off-resonant excitation of bright polaritons does not saturate, resulting in a high steady-state leakage rate of bright polaritons that overwhelms the dark-polariton signal at high probe powers. To disentangle the bright and dark contributions to the cavity transmission, Fig. 4c shows the ring-down of the dot once the probe beam is turned off. At each probe intensity, the ring-down consists of slow and fast exponential decays; the slow feature (with a time constant of 730(40) ns) reflects the dynamics of the dark polariton, while the fast feature reflects the decay of the interfering bright polaritons. We isolate the steady-state dark-polariton number, shown in Fig. 4d, by extrapolating the slowly decaying tail of the ring-down to its zero-time population. The strong saturation of the dark-polariton number with drive power again indicates blockade, while the saturation value of \({n}_{{\rm{D}}}=0.47(3)\) reflects the behaviour of a single zero-dimensional two-level emitter, as described by the optical Bloch equations. This stationary, strongly coupled, two-level emitter is an excellent candidate qubit for a photonic quantum information processor35. To verify saturation, we measure the g2 of the slowly decaying tail of the ring-down, after the bright polaritons have all decayed away. We find \({g}_{2}=0.31(7)\), confirming that the dot rarely hosts multiple dark polaritons. Moving forward, blockade-enhanced detection of dark polaritons will directly boost the dark-polariton detection signal-to-noise above the bright-polariton background36, while enhanced control-field and light–matter coupling will suppress the background (see Supplementary Information H), enabling the dot to act as an on-demand single-photon source.

Cavity Rydberg polaritons are now ripe for applications in synthetic quantum materials; recent experiments have demonstrated synthetic magnetic fields for resonator photons3 and in conjunction with the present work, there is now a clear path to topological many-photon states6. More broadly, the numerous proposals to study synthetic quantum matter in cavity arrays and continua1 can now be explored in the modes of an individual, multi-mode optical cavity, employing Rydberg atoms to mediate strong photon–photon interactions. Achieving stronger light–matter coupling will require only additional atomic density, rather than increasingly challenging advances in optical super-mirror technology necessary for single-emitter approaches. Indeed, densities routinely achieved in free-space experiments (\(\rho \approx 1{0}^{12}\) cm3)23 and finesses of single-atom experiments (\(F\approx 1{0}^{5}\)) would provide \({{\rm{OD}}}_{{\rm{B}}}\approx 88,000\), more than sufficient for few-particle synthetic materials6. Therefore, upcoming challenges centre on harnessing these developments to answer questions about dissipative preparation of many-body quantum states37,38 and about the exploration of the resulting phase diagrams39 using the quasi-particle manipulation and detection tools unique to quantum optics40,41.

Data availability

The data that support the plots within this paper and other findings of this study are available from the corresponding author upon reasonable request.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    Carusotto, I. & Ciuti, C. Quantum fluids of light. Rev. Mod. Phys. 85, 299–366 (2013).

  2. 2.

    Saffman, M., Walker, T. G. & Mølmer, K. Quantum information with Rydberg atoms. Rev. Mod. Phys. 82, 2313–2363 (2010).

  3. 3.

    Schine, N., Ryou, A., Gromov, A., Sommer, A. & Simon, J. Synthetic Landau levels for photons. Nature 534, 671–675 (2016).

  4. 4.

    Deng, H., Haug, H. & Yamamoto, Y. Exciton–polariton Bose–Einstein condensation. Rev. Mod. Phys. 82, 1489–1537 (2010).

  5. 5.

    Parigi, V. et al. Observation and measurement of interaction-induced dispersive optical nonlinearities in an ensemble of cold Rydberg atoms. Phys. Rev. Lett. 109, 233602 (2012).

  6. 6.

    A. Sommer, Büchler, H. P. & Simon, J. Quantum crystals and Laughlin droplets of cavity Rydberg polaritons. Preprint at https://arxiv.org/abs/1506.00341 (2015).

  7. 7.

    Douglas, J. S. et al. Quantum many-body models with cold atoms coupled to photonic crystals. Nat. Photon. 9, 326–331 (2015).

  8. 8.

    Haldane, F. D. M. & Raghu, S. Possible realization of directional optical waveguides in photonic crystals with broken time-reversal symmetry. Phys. Rev. Lett. 100, 013904 (2008).

  9. 9.

    Hafezi, M., Mittal, S., Fan, J., Migdall, A. & Taylor, J. M. Imaging topological edge states in silicon photonics. Nat. Photon. 7, 1001–1005 (2013).

  10. 10.

    Rechtsman, M. C. et al. Photonic Floquet topological insulators. Nature 496, 196–200 (2013).

  11. 11.

    Klaers, J., Schmitt, J., Vewinger, F. & Weitz, M. Bose–Einstein condensation of photons in an optical microcavity. Nature 468, 545–548 (2010).

  12. 12.

    Gopalakrishnan, S., Lev, B. L. & Goldbart, P. M. Emergent crystallinity and frustration with Bose–Einstein condensates in multimode cavities. Nat. Phys. 5, 845–850 (2009).

  13. 13.

    Baumann, K., Guerlin, C., Brennecke, F. & Esslinger, T. Dicke quantum phase transition with a superfluid gas in an optical cavity. Nature 464, 1301–1306 (2010).

  14. 14.

    Kollár, A. J. et al. Supermode-density-wave-polariton condensation with a Bose–Einstein condensate in a multimode cavity. Nat. Commun. 8, 14386 (2017).

  15. 15.

    Purdy, T. P. et al. Tunable cavity optomechanics with ultracold atoms. Phys. Rev. Lett. 105, 133602 (2010).

  16. 16.

    Birnbaum, K. M. et al. Photon blockade in an optical cavity with one trapped atom. Nature 436, 87–90 (2005).

  17. 17.

    Hartmann, M. J., Brandao, F. G. S. L. & Plenio, M. B. Strongly interacting polaritons in coupled arrays of cavities. Nat. Phys. 2, 849–855 (2006).

  18. 18.

    Angelakis, D. G., Santos, M. F. & Bose, S. Photon-blockade-induced Mott transitions and x y spin models in coupled cavity arrays. Phys. Rev. A 76, 031805 (2007).

  19. 19.

    Thompson, J. D. et al. Coupling a single trapped atom to a nanoscale optical cavity. Science 340, 1202–1205 (2013).

  20. 20.

    Goban, A. et al. Superradiance for atoms trapped along a photonic crystal waveguide. Phys. Rev. Lett. 115, 063601 (2015).

  21. 21.

    Mohapatra, A. K., Jackson, T. R. & Adams, C. S. Coherent optical detection of highly excited Rydberg states using electromagnetically induced transparency. Phys. Rev. Lett. 98, 113003 (2007).

  22. 22.

    Pritchard, J. D. et al. Cooperative atom–light interaction in a blockaded Rydberg ensemble. Phys. Rev. Lett. 105, 193603 (2010).

  23. 23.

    Peyronel, T. et al. Quantum nonlinear optics with single photons enabled by strongly interacting atoms. Nature 488, 57–60 (2012).

  24. 24.

    Dudin, Y. O. & Kuzmich, A. Strongly interacting Rydberg excitations of a cold atomic gas. Science 336, 887–889 (2012).

  25. 25.

    Tiarks, D., Baur, S., Schneider, K., Dürr, S. & Rempe, G. Single-photon transistor using a Förster resonance. Phys. Rev. Lett. 113, 053602 (2014).

  26. 26.

    Gorniaczyk, H., Tresp, C., Schmidt, J., Fedder, H. & Hofferberth, S. Single-photon transistor mediated by interstate Rydberg interactions. Phys. Rev. Lett. 113, 053601 (2014).

  27. 27.

    Guerlin, C., Brion, E., Esslinger, T. & Mølmer, K. Cavity quantum electrodynamics with a Rrydberg-blocked atomic ensemble. Phys. Rev. A 82, 053832 (2010).

  28. 28.

    Boddeda, R. et al. Rydberg-induced optical nonlinearities from a cold atomic ensemble trapped inside a cavity. J. Phys. B 49, 084005 (2016).

  29. 29.

    Gorshkov, A. V., Otterbach, J., Fleischhauer, M., Pohl, T. & Lukin, M. D. Photon–photon interactions via Rydberg blockade. Phys. Rev. Lett. 107, 133602 (2011).

  30. 30.

    Ningyuan, J. et al. Observation and characterization of cavity Rydberg polaritons. Phys. Rev. A 93, 041802 (2016).

  31. 31.

    Fleischhauer, M., Imamoglu, A. & Marangos, J. P. Electromagnetically induced transparency: Optics in coherent media. Rev. Mod. Phys. 77, 633–673 (2005).

  32. 32.

    Kouwenhoven, L. P. et al. in Mesoscopic Electron Transport (eds Sohn, L. L., Kouwenhoven, L. P. & Schön, G.) 105–214 (Springer, Dordrecht, 1997).

  33. 33.

    Shore, B. W. & Knight, P. L. The Jaynes–Cummings model. J. Mod. Opt. 40, 1195–1238 (1993).

  34. 34.

    Verger, A., Ciuti, C. & Carusotto, I. Polariton quantum blockade in a photonic dot. Phys. Rev. B 73, 193306 (2006).

  35. 35.

    Brion, E., Carlier, F., Akulin, V. M. & Mølmer, K. Quantum repeater with Rydberg-blocked atomic ensembles in fiber-coupled cavities. Phys. Rev. A 85, 042324 (2012).

  36. 36.

    Günter, G. et al. Interaction enhanced imaging of individual Rydberg atoms in dense gases. Phys. Rev. Lett. 108, 013002 (2012).

  37. 37.

    Hafezi, M., Adhikari, P. & Taylor, J. M. Chemical potential for light by parametric coupling. Phys. Rev. B 92, 174305 (2015).

  38. 38.

    Lebreuilly, J. et al. Stabilizing strongly correlated photon fluids with non-Markovian reservoirs. Phys. Rev. A 96, 033828 (2017).

  39. 39.

    Grusdt, F. & Fleischhauer, M. Fractional quantum Hall physics with ultracold Rydberg gases in artificial gauge fields. Phys. Rev. A 87, 043628 (2013).

  40. 40.

    Grusdt, F., Yao, N. Y., Abanin, D., Fleischhauer, M. & Demler, E. Interferometric measurements of many-body topological invariants using mobile impurities. Nat. Commun. 7, 11994 (2016).

  41. 41.

    Umucallar, R. O. & Carusotto, I. Many-body braiding phases in a rotating strongly correlated photon gas. Phys. Lett. A 377, 2074–2078 (2013).

Download references


We would like to thank M. Fleischhauer and H. P. Buechler for fruitful conversations. This work was supported by DOE grant DE-SC0010267 for apparatus construction, DARPA grant W911NF-15-1-0620 for modelling, and MURI grant FA9550-16-1-0323 for data collection and analysis. A.G. acknowledges support from the UChicago MRSEC grant NSF-DMR-MRSEC 1420709. A.R. acknowledges support from the NDSEG Fellowship.

Author information

Author notes

    • Albert Ryou

    Present address: Department of Electrical Engineering, University of Washington, Seattle, WA, USA

    • Ariel Sommer

    Present address: Department of Physics, Lehigh University, Bethlehem, PA, USA


  1. Department of Physics and James Franck Institute, University of Chicago, Chicago, IL, USA

    • Ningyuan Jia
    • , Nathan Schine
    • , Alexandros Georgakopoulos
    • , Albert Ryou
    • , Logan W. Clark
    • , Ariel Sommer
    •  & Jonathan Simon


  1. Search for Ningyuan Jia in:

  2. Search for Nathan Schine in:

  3. Search for Alexandros Georgakopoulos in:

  4. Search for Albert Ryou in:

  5. Search for Logan W. Clark in:

  6. Search for Ariel Sommer in:

  7. Search for Jonathan Simon in:


The experiment was designed and built by all authors. J.N., N.S., L.W.C. and J.S. collected and analysed the data. All authors contributed to the manuscript.

Competing interests

The authors declare no competing interests.

Corresponding author

Correspondence to Jonathan Simon.

Supplementary information

  1. Supplementary Information

    Supplementary notes, supplementary figures 1–9, supplementary references

About this article

Publication history