Floquet Phases and Discrete Time Crystals

A discrete time crystal is not merely a periodically driven system with an oscillating observable. It is a many-body regime whose response has a period $nT$ larger than the drive period $T$, is locked rather than fine-tuned, and remains ordered for a lifetime whose dependence on size, drive frequency, perturbations, and openness matches an identified stabilization mechanism.

Required background. Floquet engineering defines quasienergy and micromotion. Helpful background. Prethermalization and Floquet ensembles supplies the mechanisms that delay heating.

Quasienergy order and period multiplication

Let $U_F=U(T,0)$ be the one-period evolution. A period-doubled signal obeys, over its ordered window,

$$\langle O(mT)\rangle\simeq(-1)^m A$$

after transients. In a disorder-stabilized eigenstate description, the Floquet spectrum contains paired many-body states separated by quasienergy $\pi/T$ modulo $2\pi/T$. An order parameter odd under the emergent $\mathbb Z_2$ operation connects the partners, producing the factor $(-1)^m$.

Three features distinguish order from a driven two-level oscillation:

• subharmonicity: the response occurs at $\Omega/n$ with $n>1$;
• rigidity: its frequency remains locked over a finite interval of pulse errors or other symmetry-preserving perturbations; and
• many-body stability: correlations, lifetime, and system-size trends follow a collective stabilization mechanism.

A narrow Fourier peak alone supplies only the first item. Any finite record of a nearly periodic signal has a narrow peak, and pulse imperfections, beat notes, or detector aliasing can make one at a subharmonic.

The quasienergy-pairing definition and its many-body rigidity criterion were formulated by Else, Bauer, and Nayak 2016.

Two isolated-system stabilization mechanisms

In an ideal one-dimensional many-body-localized Floquet model, disorder can prevent indefinite energy absorption and allow eigenstate order throughout the spectrum. A schematic period-doubling unitary is

$$U_F\simeq X e^{-iDT}, \qquad X^2=1, \qquad [D,X]=0,$$

where $D$ supports spontaneous breaking of the emergent $X$ symmetry. Because $X$ flips the order parameter every period, the signal returns after $2T$. This picture requires localization stable on the relevant length and time scales; finite-chain slow heating does not by itself establish an asymptotic localized phase.

At high drive frequency, a clean interacting system can instead realize a prethermal time crystal. An approximately conserved effective Hamiltonian $D$ supports ordinary symmetry breaking, while the drive applies the discrete rotation $X$. The response persists for

$$t_{\mathrm{obs}}\ll t_*\sim g^{-1}e^{c\Omega/g}$$

under the bounded-local assumptions of high-frequency prethermalization. It ultimately heats in an isolated generic system. This mechanism also requires an effective temperature or energy density inside the symmetry-broken regime of $D$; high frequency alone is insufficient.

The two mechanisms have different negative tests. Disorder-stabilized order should be tested against increasing size, rare thermal inclusions, and changes in disorder. Prethermal order should show a lifetime that grows rapidly with frequency before extrinsic decoherence dominates, plus thermal behavior when the effective state lies outside the ordered region.

Zaletel et al. 2023 review the distinction among localized, prethermal, dissipative, and classical subharmonic mechanisms.

Floquet topology and open temporal order

Periodic drives also support anomalous topological phases whose edge evolution cannot be generated by a static bulk Hamiltonian with the same symmetries. Their invariants involve the full loop $U(t)$, not only a chosen logarithm $H_F$. Edge motion or a quasienergy gap at $0$ or $\pi/T$ therefore requires a bulk invariant, boundary-condition checks, and micromotion-aware measurement.

An open driven system can sustain a stable limit cycle because drive and dissipation balance. Such dissipative time-crystalline behavior is meaningful, but its classification is not the same as isolated Floquet eigenstate order. The claim must say whether it concerns the unconditional density matrix, individual monitored trajectories, or a postselected record. Conditional oscillations can disappear after averaging over measurement outcomes.

Evidence ladder for a time-crystal claim

The strongest practical test combines several observables:

1. map the subharmonic amplitude and frequency across a perturbation interval rather than at one tuned point;
2. measure unequal-time correlations, not only the mean magnetization;
3. vary $L$ and extract a lifetime with a fixed operational threshold;
4. compare that lifetime with heating, decoherence, and recurrence times measured independently;
5. prepare generic and symmetry-related initial states to exclude a special-state revival; and
6. show the predicted breakdown when the stabilizing mechanism is removed.

The pioneering 2017 trapped-ion and diamond-spin experiments established robust finite-duration subharmonic responses Zhang et al. 2017, Choi et al. 2017. Later programmable experiments resolved prethermal stabilization and many-body correlations more directly Kyprianidis et al. 2021. These are substantial realizations of time-crystalline dynamics in finite, imperfect systems. They do not convert a finite number of cycles into proof of infinite-time order before the thermodynamic limit.

Evidence status

This assessment was checked through 10 August 2026. Discrete time-crystal phenomena are experimentally established as robust finite-window many-body responses in several platforms; which thermodynamic classification applies remains mechanism and dimension dependent. Current work continues to separate prethermal, localized, dissipative, and constrained-sector stabilization. Superseding results and negative evidence belong in the Quantum Matter and Emergence Research dossier.

Exercises

1. Quasienergy pairing. Suppose $U_F\lvert\phi_+\rangle=e^{-i\varepsilon T}\lvert\phi_+\rangle$ and $U_F\lvert\phi_-\rangle=-e^{-i\varepsilon T}\lvert\phi_-\rangle$. For $\lvert\psi_0\rangle=(\lvert\phi_+\rangle+\lvert\phi_-\rangle)/\sqrt2$ and an operator with $\langle\phi_+\rvert O\lvert\phi_-\rangle=A$, find the stroboscopic oscillating term.

Solution

After $m$ periods the relative phase is $(-1)^m$. The off-diagonal contribution is $(-1)^m(A+A^*)/2=(-1)^m\operatorname{Re}A$. Thus the quasienergy separation $\pi/T$ produces a $2T$ response. Stability of that pairing, not the algebra at one parameter point, is the phase question.

2. Frequency and bath ceilings. A prethermal lifetime follows $t_*=t_0e^{c\Omega/g}$, while decoherence limits coherence to $t_{\mathrm{dec}}$. Sketch the measured lifetime as $\Omega$ increases.

Solution

The observed lifetime is approximately $\min(t_*,t_{\mathrm{dec}})$, with additional preparation and finite-size limits if present. It initially grows exponentially with $\Omega/g$ and then saturates near $t_{\mathrm{dec}}$. That saturation does not refute prethermal scaling unless the intrinsic regime was never resolved below the bath ceiling.

References

• Choi, Soonwon, Joonhee Choi, Renate Landig, Georg Kucsko, Hengyun Zhou, Junichi Isoya, Fedor Jelezko, Shinobu Onoda, Hitoshi Sumiya, Vedika Khemani, Curt von Keyserlingk, Norman Y. Yao, Eugene Demler, and Mikhail D. Lukin. “Observation of Discrete Time-Crystalline Order in a Disordered Dipolar Many-Body System.” Nature 543, 221–225 (2017). DOI.
• Else, Dominic V., Bela Bauer, and Chetan Nayak. “Floquet Time Crystals.” Physical Review Letters 117, 090402 (2016). DOI.
• Kyprianidis, Antonis, Francisco Machado, William Morong, Patrick Becker, Kate S. Collins, Dominic V. Else, Lei Feng, Paul W. Hess, Chetan Nayak, Guido Pagano, Norman Y. Yao, and Christopher Monroe. “Observation of a Prethermal Discrete Time Crystal.” Science 372, 1192–1196 (2021). DOI.
• Zaletel, Michael P., Mikhail Lukin, Christopher Monroe, Chetan Nayak, Frank Wilczek, and Norman Y. Yao. “Colloquium: Quantum and Classical Discrete Time Crystals.” Reviews of Modern Physics 95, 031001 (2023). DOI.
• Zhang, Jiehang, Paul W. Hess, Antonis Kyprianidis, Patrick Becker, A. Lee, J. Smith, Guido Pagano, I.-D. Potirniche, Andrew C. Potter, Ashvin Vishwanath, Norman Y. Yao, and Christopher Monroe. “Observation of a Discrete Time Crystal.” Nature 543, 217–220 (2017). DOI.