Power-law spin-coupling gradient turns a time crystal into a quantum battery and sensor ✛ ✎
Context
Discrete time crystals are phases of non-equilibrium matter in which the system spontaneously breaks discrete time-translation symmetry and starts oscillating at a period that is a multiple of the drive period. Usually such phases are stabilized by disorder, which causes many-body localization and prevents the system from heating to a featureless infinite-temperature state. However, disorder complicates control and scaling. Alongside this work, the archive already contains studies of time crystals based on Stark many-body localization, where a linear potential gradient is used instead of disorder, as well as works on quantum batteries and sensors based on time crystals. The present paper generalizes the Stark approach: instead of a linear potential gradient, it introduces a power-law gradient in the interaction strength between neighboring spins, expanding the class of systems with a stable time crystal and opening the path to practical applications.
Methods
The authors consider a one-dimensional chain of L spin-1/2 particles with open boundary conditions. The interaction between neighboring spins has the form j^α $σ_j^z$ $σ_{j+1}^z$, where j is the site index and α is a tunable exponent. The system is subjected to periodic instantaneous kicks (delta pulses) that rotate all spins or only spins on even sites about the x-axis by an angle close to π/2. Evolution over one period is described by the Floquet operator. For analytical analysis, they use the spectral decomposition of the Floquet operator and symmetries of the Hamiltonian: global spin inversion and spin inversion on one sublattice. Numerically, they employ matrix product states with density matrix renormalization group for large systems. To assess sensor sensitivity, they compute the quantum Fisher information, and for the battery, ergotropy, which here coincides with stored energy because the evolution is unitary and the initial state is pure.
Results
The system exhibits period doubling for a wide range of exponents α from 0.5 to 1.5 and above. For a small deviation e from the ideal rotation angle (e.g., e=0.01), the state returns to the initial one after an even number of periods up to a global phase, and after an odd number of periods it transitions to an orthogonal state. This is confirmed by a peak at frequency π in the Fourier spectrum and by averaging over all initial states. The lifetime of the time crystal grows exponentially with system size: t* ~ exp(1.99(2)L) for α=0.5 and t* ~ exp(2.15(9)L) for α=1.5. In the quantum battery application, the stored energy in the time-crystal phase scales as E ~ $L^{2.042(2)}$ for α=1, that is, superlinearly. However, after normalizing the Hamiltonian, the power becomes almost independent of system size, so the true advantage of the battery is not superlinear power but robustness: the charge is preserved under imperfections and does not require precise synchronization with the optimal moment. For the quantum sensor, the quantum Fisher information in the time-crystal phase grows as $F_Q$ ~ $n^2$ $L^{β_s}$, with $β_s$ = 2.73(4) for α=0.5 and $β_s$ = 4.1(2) for α=1.5. This exceeds the standard quantum limit ($F_Q$ ~ L) and approaches or even surpasses the Heisenberg limit ($F_Q$ ~ $L^2$). The exponent $β_s$ depends on α sub-polynomially: $β_s$ ~ $α^{0.37(3)}$. Under Markovian dephasing, the quantum Fisher information still grows with time as $F_Q$ ~ $n^{1.8}$ for α=0.5 and $F_Q$ ~ $n^{2.3}$ for α=1.5, indicating that the quantum advantage persists in an open system.
Implications
This work expands the class of systems that support discrete time crystals without disorder. Unlike previous studies that used a linear potential gradient (Stark localization), here the gradient is introduced into the interaction strength, providing an additional degree of freedom for tuning. This allows combining two functions in one platform: energy storage and quantum sensing. The practical significance is that the time crystal's robustness to preparation and control imperfections reduces the precision requirements for experimental implementation, which is important for scalable quantum devices.
Future development
A natural direction is a detailed characterization of the localization mechanism that stabilizes the time crystal. The authors conjecture that it is Stark many-body localization, but a rigorous proof remains open. Another direction is extending the analysis to open quantum systems with more realistic noise models, including non-Markovian effects and dissipation. It is also interesting to explore two- and three-dimensional lattices with a similar power-law gradient of interactions, where new phases and stronger quantum advantages may emerge.
Impact
The results can be used in the development of quantum batteries and quantum sensors on platforms with trapped ions and Rydberg atom arrays, where tunable power-law interactions and periodic pulses are already available.
Next steps
Test the proposed mechanism on real quantum simulators, for example, in trapped-ion chains, and measure the time crystal lifetime for various α. Also investigate how noise and pulse imperfections affect the quantum Fisher information in larger systems.
Key open problems
For quantum thermodynamics and quantum metrology, the open question remains which non-equilibrium phases provide maximum robustness and sensitivity under realistic noise. This work shows that time crystals based on a power-law interaction gradient can be a candidate, but further studies of their behavior in open systems and at finite temperatures are needed.
The knowledge machine's view
This section is addressed to the author of the paper and their colleagues. Neighbouring works were found by meaning-based search across our archive, not by word overlap. We do not review: we show the field around the work.
What the work does
This work examines a one-dimensional spin chain with interactions that grow according to a power law along the chain. Under periodic driving, the system exhibits a stable discrete time crystal with period doubling, which persists for exponentially long times as the number of spins increases. In the time-crystal phase, the system acts as a quantum battery with superlinear growth of stored energy and as a quantum sensor with a quantum advantage in estimating deviations in the driving period.
Where its strength is
Using a power-law gradient in the interaction strength, instead of disorder or a linear potential, expands the class of systems that host stable time crystals. The tunable exponent allows flexible control over the balance between stability and sensitivity, opening the way to practical applications.
Where it could go next
- It might be interesting to compare the proposed stabilization mechanism with the approach based on Stark many-body localization, where a linear potential gradient is used instead of disorder. In our archive, there is an analysis of a work on Stark-localization-based time crystals, showing how a linear potential suppresses heating and stabilizes periodic dynamics.based on: A Day in the Quantum Beat: How Cesium Atoms Became an Eternal Metronome arXiv:2607.09247
- A possible direction is to study time crystals without external periodic driving, similar to recent works on spontaneous time crystallization in autonomous systems. In the archive, there is a quick analysis of a paper where a time crystal emerges without driving, which may suggest how to adapt the power-law interaction gradient for continuous-time systems.based on: Time Crystallizes Without a Kick arXiv:2605.14250
- Another move is to consider hierarchical time crystals, where symmetry breaking occurs at multiple levels. The archive contains a quick analysis of a work on matryoshka time crystals, showing how multiple oscillation periods can coexist; perhaps the power-law interaction gradient could enable such structures.based on: Matryoshka Time Crystals: A Hierarchy of Rhythms in a System of Five Ions arXiv:2601.09779
Worth trying
- Test whether the time crystal remains stable when weak disorder is added on top of the power-law interaction gradient.
- Investigate how the power-law interaction gradient affects quantum Fisher information in the sensor regime at finite temperatures or with mixed initial states.
- Consider implementing the proposed model on trapped-ion or Rydberg-atom platforms, where interactions can be spatially tuned.
Why this may matter
This work paves the way for creating stable time crystals without disorder, simplifying experimental implementation and scaling. The proposed system simultaneously serves as a quantum battery and a sensor, which could be valuable in quantum technologies where stability and sensitivity are crucial.
Nearby works in our archive
- Time Crystals: Spontaneous Breaking of Time Translation Symmetry in Quantum Systemssimilarity 0.56
- Matryoshka Time Crystals: A Hierarchy of Rhythms in a System of Five Ionssimilarity 0.56
- A Day in the Quantum Beat: How Cesium Atoms Became an Eternal Metronomesimilarity 0.53
- A Time Crystal Trapped in a Cage of Soundsimilarity 0.53
- Time Crystallizes Without a Kicksimilarity 0.52
- A Kiss of Spectra: How a Dance of Energy Survives Frictionsimilarity 0.52
The archive is sparse around this work: the closest review is only 0.56, against a usual 0.64. It looks like few have drilled nearby.
In the ideal case (e=0), the time-crystal battery charges to the maximum possible energy in one odd period and returns to the initial state after an even number of periods, and this cycle repeats indefinitely without losses.
The idea of a perpetual cyclic device that returns to its initial state echoes the concept of a 'perpetual motion machine' from science fiction, but here it does not violate the laws of thermodynamics because the system is not isolated and receives energy from the external periodic drive.
Key numbers
- Battery energy growth exponent: 2.042(2) for α=1
- Quantum Fisher information growth exponent: 2.73(4) for α=0.5 and 4.1(2) for α=1.5
- Exponential lifetime growth: t* ~ exp(1.99(2)L) for α=0.5
- Critical period deviation: ω_c ~ L^{-3.1} for α=0.5 and L^{-5.3} for α=1.5