Dissipative time crystal in a strongly interacting Rydberg gas

The study explores emergent nonequilibrium many-body phases, focusing on time crystals—phases that spontaneously break time-translation symmetry. Continuous time crystals (CTCs) display self-sustained oscillations under continuous driving but are incompatible with equilibrium settings. Dissipation in open quantum systems offers a route to stabilize such phases. Prior observations of dissipative CTCs were made in atomic Bose–Einstein condensates in optical cavities, where atom loss limits long-time dynamics. Rydberg gases, with controllable long-range interactions and operation in room-temperature vapor cells without atom loss, provide a promising platform. The research question is whether a continuously driven, dissipative Rydberg gas can realize a CTC exhibiting persistent oscillations with long-range temporal order and robustness to temporal perturbations. The study demonstrates such a phase and elucidates its microscopic origin in the competition among multiple Rydberg levels.

The work situates itself within nonequilibrium many-body physics and time-crystal research. Foundational studies discuss time crystals and no-go theorems for equilibrium continuous time crystals, as well as dissipative mechanisms for time-translation symmetry breaking. Experimental observations include discrete time crystals in various platforms and a dissipative CTC in an atom–cavity system. Rydberg ensembles have shown nonequilibrium phase transitions, bistability, and self-organization phenomena in prior experiments and theory. The paper builds on mean-field treatments for thermal Rydberg gases where motion averages spatial correlations, enabling interaction-induced nonlinearities and bistability. It contrasts with reports attributing bistability in hot vapors to ion-induced interactions, noting that here attractive vdW interactions in nD states and experimental conditions differ (lower density and narrower spectra), while acknowledging possible ionic contributions.

Experimental platform: A room-temperature 85Rb vapor cell (7.5 cm) hosts atoms driven via a two-photon scheme: a 780 nm probe field (σ+ to couple |g⟩=5S1/2, F=3, mF=3 to |e⟩=5P3/2, F=4, mF=4) and a counterpropagating 480 nm coupling field to Rydberg nD states (e.g., |r⟩=nD5/2, mj=1/2; |s⟩=nD5/2, mj=5/2; |q⟩=nD3/2, mj=1/2). Polarizations are set by quarter-wave plates to control branching to different Rydberg Zeeman sublevels. A calcite beam displacer provides parallel probe and reference beams; differential transmission is read out on a balanced photodiode. Magnetic field control (homogeneous field parallel to beams) tunes Zeeman splittings to adjust level competition. Typical Rabi frequencies: Ωp/2π≈18–20 MHz for the probe, Ω/2π≈0.7–1.3 MHz for the coupling (for selected transitions), with intermediate-state detuning Δp/2π≈30–94 MHz. The principal quantum number n is varied (e.g., 55–85) to adjust interaction strengths and level separations.

Spectroscopy and phase mapping: Transmission spectra are recorded by slowly scanning the 480 nm coupling detuning around two-photon resonance at fixed intermediate-state detuning. By varying magnetic field and coupling polarization, the experiment isolates cases with one or multiple Rydberg states involved: oscillations appear only when at least two Rydberg states are simultaneously driven (elliptical polarization and small Zeeman splitting). The onset and extent of oscillations versus Rabi frequency and detuning are mapped; an oscillation contrast ratio C serves as an order parameter to extract a phase diagram of stationary versus limit-cycle (OSC) regimes.

Quench protocol and analysis: For time-crystal tests, the probe is abruptly turned on (≈2 μs) and then held constant up to 125 ms (coupling light on continuously); laser frequencies are cavity-locked (ULE, PDH). Single-shot transmission time traces are analyzed by discrete Fourier transform (DFT) over sliding windows to monitor spectral peak evolution from transient to steady periodic behavior. The autocorrelation function G(τ)=∫dt I(t) I(t+τ) (with I(t) zero-mean transmission) quantifies temporal correlations and long-range order (LRO). Multiple (≈100–200) independent realizations assess frequency stabilization and shot-to-shot phase randomness.

Noise robustness: Temporal intensity noise is added to the probe via RF amplitude modulation of an AOM. A noise strength N is defined from the single-sided amplitude spectra of the probe with/without noise. A crystalline fraction Q=Σω∈[ω0,ω0+δω] |A(ω)| / Σω A(ω) (around the fundamental frequency ω0 with DFT resolution δω) quantifies order under noise. DFT spectra and ACFs are evaluated across noise levels to characterize melting.

Theory: A mean-field V-type three-level model with two driven Rydberg states captures the essential mechanism. Rydberg–Rydberg interactions produce a nonlinear energy shift εNL=χ(nr+ns), effectively coupling the dynamics of levels r and s. With decay γ, the nonlinear Bloch equations exhibit stationary, bistable, and, via a Hopf bifurcation when level competition is strong (detuning separation comparable to interaction shift), a limit-cycle phase with persistent oscillations. The model reproduces key qualitative features and trends (e.g., frequency decreasing with stronger drive in the oscillatory regime), though high-frequency branches can overestimate experimental values. The role of ions can be incorporated as an additional nonlinear Stark shift proportional to total Rydberg density, effectively modifying χ.

• Observation of persistent oscillations (limit cycles): Transmission through a continuously driven room-temperature 85Rb Rydberg vapor exhibits stable, non-decaying oscillations sustained for at least 125 ms after a brief (~200 μs) transient; typical fundamental frequencies are in the few-kHz range (e.g., ~5–5.5 kHz).
• Multi-level competition is essential: Oscillations appear only when at least two nearby Rydberg Zeeman sublevels are simultaneously driven. With elliptical coupling polarization and small magnetic fields (levels close in energy), oscillations are present; with purely circular polarization (driving only one Rydberg state), oscillations disappear. Increasing magnetic field to separate levels also suppresses oscillations.
• Parameter regime and phase diagram: A limit-cycle (OSC) phase emerges above threshold Rabi frequencies; as Ωp increases, oscillations first appear weak with long periods, then grow in amplitude and span a broader detuning range. An experimental phase diagram using oscillation contrast C identifies wide regions of OSC robust to variations of detuning and drive strengths.
• Quench dynamics establishes long-range temporal order: DFT spectra of the transmission evolve from broad, drifting peaks (2–32 ms) to equidistant, narrow peaks (50–80 ms and later), evidencing stable periodicity. Across 100 realizations, the fundamental frequency drifts upward during transients and converges to a steady value. The autocorrelation function G(τ) shows nearly constant-amplitude oscillations over ~80 periods in the steady regime, demonstrating true long-range temporal order.
• Robustness to temporal perturbations (rigidity): Adding probe-intensity noise leads to a gradual (approximately linear at first) decrease of the normalized crystalline fraction Q with noise strength N. In the intermediate noise regime, DFT spectra retain equally spaced peaks atop a noisy background and ACF oscillations persist (LRO maintained with slight late-time decay). At strong noise, higher harmonics are suppressed and ACF decays for τ>0 (melting of temporal order), yet the fundamental peak can remain visible.
• Mechanism and theory support: Mean-field analysis with two interacting Rydberg states predicts a Hopf bifurcation between bistable regions, producing stable limit cycles. The model captures the observed dependence of oscillation frequency on drive strength (frequency decreases with increasing effective drive in the OSC regime) and attributes oscillations to competition between Rydberg components coupled by nonlinear interaction shifts.
• Additional observations: Higher principal quantum number n enhances oscillatory behavior due to stronger interactions and closer manifold spacings; oscillations were particularly strong for n≈75–85. The oscillation frequency shows long-term drifts over many realizations (environmental fluctuations), and single-shot phases are random across runs, consistent with spontaneous symmetry breaking and phase diffusion.

The results address whether a dissipative, continuously driven Rydberg gas can realize a continuous time crystal. The observed stable, self-sustained oscillations, non-decaying autocorrelation over many periods, and robustness to temporal noise establish the spontaneous breaking of continuous time-translation symmetry with long-range temporal order—key hallmarks of a CTC. The dependence on multi-level competition, controlled via magnetic field and polarization, identifies the microscopic mechanism: strong interactions couple distinct Rydberg components, generating nonlinear shifts that drive a Hopf bifurcation and stabilize limit cycles. Compared to previous CTC observations in cavity BECs limited by atom loss, the room-temperature vapor platform enables long-time dynamics and direct transmission readout. The mean-field model explains qualitative features and parameter trends, supporting the interpretation that interaction-mediated competition among Rydberg levels underlies the time-crystalline phase.

The study demonstrates a dissipative continuous time crystal in a room-temperature Rydberg gas, observed via persistent oscillations in optical transmission arising from interaction-enabled competition among multiple Rydberg states. The oscillations exhibit true long-range temporal order and are robust to temporal intensity noise, satisfying the defining criteria of a CTC. Future work could map the full phase diagram, including bistable regions characterized by hysteresis; explore discrete time crystals under periodic driving; implement in Rydberg tweezer arrays to investigate dimensionality, quantum fluctuations, and system-size scaling; and leverage time-crystalline rigidity for quantum synchronization and sensing applications.

• Theoretical modeling is mean-field and qualitative: it neglects intermediate-state dynamics, Doppler effects from thermal motion, velocity classes, and spatial inhomogeneities; high-frequency branches overestimate observed frequencies. A more complete model is needed for quantitative predictions.
• Possible ionic effects: While spectra are narrow and fluorescence measurements show no clear plasma signatures, limited sensitivity prevents ruling out ion-induced interactions; such effects could modify the effective nonlinearity.
• Environmental drifts and phase randomness: Long-term frequency drifts across realizations and random oscillation phase indicate sensitivity to slow experimental fluctuations; distinguishing quantum from classical sources of phase diffusion is challenging.
• Parameter constraints: Observation of limit cycles requires sufficiently large Rabi frequencies and higher principal quantum numbers n; lower abundance of 87Rb demands stronger driving or heating to reach similar regimes.