Spin-Chain Quantum Battery: Insights and Protocols

• Spin-chain quantum batteries are many-body energy storage systems with interacting spins charged via noncommuting fields, quench protocols, or engineered environments.
• Their performance is quantified by stored energy relative to reference states, charging power, and ergotropy, with enhancements arising from intrinsic interactions and anisotropy.
• Diverse Hamiltonian architectures and charging protocols—ranging from cavity-coupled to quenched setups—demonstrate trade-offs between interaction-induced benefits and challenges like decoherence and disorder.

Spin-chain quantum batteries are many-body temporary energy-storage systems in which the working medium is an interacting spin chain and the charging stage is implemented by a noncommuting field, a cavity, a quench, a Floquet kick sequence, or an engineered environment. Across XX, XY, XXZ, XYZ, Heisenberg, dimerized XY, Ising, and cavity-coupled realizations, the central observables are the stored energy relative to a ground or thermal reference, the charging power, and the ergotropy, namely the maximum extractable work under unitary operations. A recurrent theme is that intrinsic spin-spin interactions, anisotropy, boundary conditions, and critical spectral rearrangements can strongly modify both capacity and charging rate, while decoherence, feedback, and disorder may either degrade or stabilize performance depending on the protocol (Le et al., 2017, Zhao et al., 2020).

1. Hamiltonian architectures and model classes

The literature uses several distinct spin-chain architectures, but most are organized around a battery Hamiltonian $H_B$ that defines the stored-energy observable and a noncommuting charging sector that injects energy. Nearest-neighbor XX/XY/XXZ/XYZ chains charged by uniform transverse fields form one major class. A second class couples the chain to a cavity mode or an auxiliary charger. A third uses internal-parameter quenches or Floquet kicks rather than a separate charger. Dimerization, long-range couplings, anisotropy, and graph topology are then used to tune both the spectrum and the charging dynamics (Zhao et al., 2020, Le et al., 2017, Grazi et al., 2024, Romero et al., 21 Nov 2025).

Archetype	Core Hamiltonian structure	Characteristic outcome
XX/XXZ/XYZ chain with local field	Nearest-neighbor exchange plus a uniform charging field	Interaction-enhanced power; anisotropy-dependent advantage
Cavity-coupled spin chain	Spin chain + single bosonic mode + collective spin-cavity coupling	Open-system charging, thermal ergotropy, cavity-mediated control
Dimerized or quenched integrable chain	Internal parameter quench of $\delta$, $h$, or field	Robust storage across quantum phase transitions
Floquet kicked Ising chain	Static battery Hamiltonian plus kicked Ising charger	Exact stroboscopic charging, parity/boundary effects, disorder robustness
Heisenberg chain with DM/KSEA terms	Exchange anisotropy plus antisymmetric/symmetric spin-orbit terms	Capacity and ergotropy enhancement with interaction thresholds

A canonical cavity realization embeds an $N$-spin chain in a single-mode microcavity and decomposes the total Hamiltonian as

$H_S=H_A+H_B+H_I,$

with

$$H_A=\omega_c c^\dag c,\qquad H_B=\omega_a \sum_{i=1}^{N}\sigma_+^i\sigma_-^i+J\sum_{i=1}^{N-1}\bigl(\sigma_+^i\sigma_-^{i+1}+h.c.\bigr),$$

$H_I=\sum_{i=1}^{N}g(\sigma_+^ic+h.c.).$

This open-boundary XY/XX-type hopping chain is a standard reference for cavity-mediated spin-chain batteries (Zhao et al., 2020).

Other representative battery Hamiltonians are structurally different but conceptually analogous. The cavity-Heisenberg spin-$j$ model uses

$$H_B=\omega_a \sum_{n=1}^{N} S^z_n + \omega_a J \sum_{n=1}^{N-1}\Big[(1+\gamma)S^x_nS^x_{n+1}+(1-\gamma)S^y_nS^y_{n+1}+\Delta S^z_nS^z_{n+1}\Big],$$

with local Hilbert-space dimension $2j+1$ and $\delta$0 (Sun et al., 2024). The dimerized XY battery instead uses alternating bonds,

$\delta$1

which introduces a two-site unit cell and a phase diagram controlled by $\delta$2 and $\delta$3 (Grazi et al., 2024). In Floquet formulations, the battery Hamiltonian is often taken as

$\delta$4

while the charger is a kicked Ising sector $\delta$5, with the choice of $\delta$6 depending on the charger variant (Romero et al., 21 Nov 2025).

2. Charging protocols and performance metrics

Charging protocols span coherent local fields, cavity pumping, thermal baths, sudden and double quenches, Landau-Zener ramps, periodic drives, feedback-controlled dissipation, reinforcement-learned couplings, and Floquet kick trains. In cavity setups, charging can be produced either by a coherent cavity drive

$\delta$7

or by coupling the cavity to a bosonic thermal bath with occupation

$\delta$8

In quench-based batteries, the charger is not a separate subsystem: a Hamiltonian parameter such as dimerization $\delta$9 or transverse field $h$0 is switched for a finite interval and then restored (Zhao et al., 2020, Grazi et al., 2024, Siddique et al., 2024).

Two stored-energy conventions dominate. One measures energy relative to the instantaneous ground-state energy,

$h$1

while another measures work relative to the initial battery state,

$h$2

The average charging power is commonly

$h$3

and some works also use the instantaneous derivative $h$4. A frequently used charging time is the maximizer of average power,

$h$5

Ergotropy is defined through the passive-state construction,

$h$6

or equivalently as $h$7 after ordering state eigenvalues $h$8 in decreasing order and Hamiltonian eigenvalues $h$9 in increasing order (Zhao et al., 2020, Ali et al., 2024).

Several works also introduce explicit efficiency-type ratios. One is the useful-work fraction

$N$0

and another, used for noninteracting spin-$N$1 chains, normalizes the stored energy by the spectral bound

$N$2

For $N$3 and $N$4, this gives $N$5 for $N$6, respectively (Sun et al., 2024). In several closed, unitary charging protocols starting from passive states, the stored energy coincides with ergotropy, so the full stored energy is in principle extractable (Le et al., 2017, Ghosh et al., 2021, Verma et al., 2024).

3. Interaction-induced enhancement, anisotropy, and scaling laws

Intrinsic interactions can increase charging power even when the charger is only a uniform local field. For the XXZ chain charged by

$N$7

the isotropic case $N$8 satisfies $N$9, so interactions do not affect charging and the dynamics reduce to the noninteracting benchmark. By contrast, for $H_S=H_A+H_B+H_I,$0 the weak-coupling stored energy contains an explicit interaction contribution,

$H_S=H_A+H_B+H_I,$1

with $H_S=H_A+H_B+H_I,$2. This yields $H_S=H_A+H_B+H_I,$3 for finite-range interactions, $H_S=H_A+H_B+H_I,$4 for $H_S=H_A+H_B+H_I,$5, and $H_S=H_A+H_B+H_I,$6 for infinite-range $H_S=H_A+H_B+H_I,$7 interactions without Kac normalization (Le et al., 2017).

Nearest-neighbor XYZ batteries charged by a local $H_S=H_A+H_B+H_I,$8-field also show a strong interaction-assisted regime. In an ordered XY chain with $H_S=H_A+H_B+H_I,$9 and $$H_A=\omega_c c^\dag c,\qquad H_B=\omega_a \sum_{i=1}^{N}\sigma_+^i\sigma_-^i+J\sum_{i=1}^{N-1}\bigl(\sigma_+^i\sigma_-^{i+1}+h.c.\bigr),$$0, the interaction-induced enhancement in maximal average power can reach $$H_A=\omega_c c^\dag c,\qquad H_B=\omega_a \sum_{i=1}^{N}\sigma_+^i\sigma_-^i+J\sum_{i=1}^{N-1}\bigl(\sigma_+^i\sigma_-^{i+1}+h.c.\bigr),$$1 relative to the $$H_A=\omega_c c^\dag c,\qquad H_B=\omega_a \sum_{i=1}^{N}\sigma_+^i\sigma_-^i+J\sum_{i=1}^{N-1}\bigl(\sigma_+^i\sigma_-^{i+1}+h.c.\bigr),$$2 reference, and the advantage is reported to be almost independent of system size. In the same framework, finite temperature and quenched disorder can produce even larger maximal average power in specific parameter windows, leading to a disorder-induced order effect in the charging dynamics (Ghosh et al., 2019).

When long-range collective structures are added, stronger scaling exponents appear. In the cavity-Heisenberg spin chain with all-to-all spin-spin interactions, the maximum stored energy scales as $$H_A=\omega_c c^\dag c,\qquad H_B=\omega_a \sum_{i=1}^{N}\sigma_+^i\sigma_-^i+J\sum_{i=1}^{N-1}\bigl(\sigma_+^i\sigma_-^{i+1}+h.c.\bigr),$$3, whereas the maximum average charging power obeys

$$H_A=\omega_c c^\dag c,\qquad H_B=\omega_a \sum_{i=1}^{N}\sigma_+^i\sigma_-^i+J\sum_{i=1}^{N-1}\bigl(\sigma_+^i\sigma_-^{i+1}+h.c.\bigr),$$4

with $$H_A=\omega_c c^\dag c,\qquad H_B=\omega_a \sum_{i=1}^{N}\sigma_+^i\sigma_-^i+J\sum_{i=1}^{N-1}\bigl(\sigma_+^i\sigma_-^{i+1}+h.c.\bigr),$$5 and able to exceed $$H_A=\omega_c c^\dag c,\qquad H_B=\omega_a \sum_{i=1}^{N}\sigma_+^i\sigma_-^i+J\sum_{i=1}^{N-1}\bigl(\sigma_+^i\sigma_-^{i+1}+h.c.\bigr),$$6 under parameter optimization, while the corresponding long-range Heisenberg spin-chain battery without cavity reaches only about $$H_A=\omega_c c^\dag c,\qquad H_B=\omega_a \sum_{i=1}^{N}\sigma_+^i\sigma_-^i+J\sum_{i=1}^{N-1}\bigl(\sigma_+^i\sigma_-^{i+1}+h.c.\bigr),$$7 (Dou et al., 2022). In the open Kitaev $$H_A=\omega_c c^\dag c,\qquad H_B=\omega_a \sum_{i=1}^{N}\sigma_+^i\sigma_-^i+J\sum_{i=1}^{N-1}\bigl(\sigma_+^i\sigma_-^{i+1}+h.c.\bigr),$$8-$$H_A=\omega_c c^\dag c,\qquad H_B=\omega_a \sum_{i=1}^{N}\sigma_+^i\sigma_-^i+J\sum_{i=1}^{N-1}\bigl(\sigma_+^i\sigma_-^{i+1}+h.c.\bigr),$$9 chain, parallel charging remains extensive with $H_I=\sum_{i=1}^{N}g(\sigma_+^ic+h.c.).$0, but collective charging with XY anisotropy and nonzero $H_I=\sum_{i=1}^{N}g(\sigma_+^ic+h.c.).$1 reaches super-extensive scaling up to $H_I=\sum_{i=1}^{N}g(\sigma_+^ic+h.c.).$2 for $H_I=\sum_{i=1}^{N}g(\sigma_+^ic+h.c.).$3, with the best reported set

$H_I=\sum_{i=1}^{N}g(\sigma_+^ic+h.c.).$4

(Ali et al., 2024).

Interaction topology and drive protocol can also be paired deliberately. In the anisotropic XY battery driven by a Landau-Zener field, long-range couplings with $H_I=\sum_{i=1}^{N}g(\sigma_+^ic+h.c.).$5 and $H_I=\sum_{i=1}^{N}g(\sigma_+^ic+h.c.).$6 give smoother dynamics and larger deposited work than nearest-neighbor couplings at the same $H_I=\sum_{i=1}^{N}g(\sigma_+^ic+h.c.).$7. The reported $H_I=\sum_{i=1}^{N}g(\sigma_+^ic+h.c.).$8 grows almost linearly with $H_I=\sum_{i=1}^{N}g(\sigma_+^ic+h.c.).$9 in the long-range Landau-Zener regime, while nearest-neighbor chains may instead favor periodic driving at maximal sweep rate $j$0 (Siddique et al., 2024).

A separate enhancement mechanism is enlargement of the local spin dimension. In the cavity-Heisenberg spin-$j$1 chain with $j$2, larger $j$3 systematically increases both stored energy and average charging power, with $j$4 in both closed and open cases (Sun et al., 2024).

4. Spectral rearrangements, frustration, and quantum criticality

Critical spectral structure is one of the most distinctive features of spin-chain batteries. In the cavity-coupled nearest-neighbor hopping chain, the hopping term splits the many-body bands and causes ground-state crossings. For $j$5, the steady stored energy $j$6 has a non-differentiable point at $j$7, while the steady ergotropy shows non-differentiable points at $j$8 and $j$9. More generally, discontinuities in

$$H_B=\omega_a \sum_{n=1}^{N} S^z_n + \omega_a J \sum_{n=1}^{N-1}\Big[(1+\gamma)S^x_nS^x_{n+1}+(1-\gamma)S^y_nS^y_{n+1}+\Delta S^z_nS^z_{n+1}\Big],$$0

occur at $$H_B=\omega_a \sum_{n=1}^{N} S^z_n + \omega_a J \sum_{n=1}^{N-1}\Big[(1+\gamma)S^x_nS^x_{n+1}+(1-\gamma)S^y_nS^y_{n+1}+\Delta S^z_nS^z_{n+1}\Big],$$1 and $$H_B=\omega_a \sum_{n=1}^{N} S^z_n + \omega_a J \sum_{n=1}^{N-1}\Big[(1+\gamma)S^x_nS^x_{n+1}+(1-\gamma)S^y_nS^y_{n+1}+\Delta S^z_nS^z_{n+1}\Big],$$2, identifying first-order ground-state quantum phase transitions. Before the first crossing the ground-state energy is zero; afterward it decreases, so $$H_B=\omega_a \sum_{n=1}^{N} S^z_n + \omega_a J \sum_{n=1}^{N-1}\Big[(1+\gamma)S^x_nS^x_{n+1}+(1-\gamma)S^y_nS^y_{n+1}+\Delta S^z_nS^z_{n+1}\Big],$$3 increases sharply because the reference energy $$H_B=\omega_a \sum_{n=1}^{N} S^z_n + \omega_a J \sum_{n=1}^{N-1}\Big[(1+\gamma)S^x_nS^x_{n+1}+(1-\gamma)S^y_nS^y_{n+1}+\Delta S^z_nS^z_{n+1}\Big],$$4 is lowered (Zhao et al., 2020).

In cyclic battery-charger Ising devices, criticality can directly improve thermodynamic performance. For a battery defined as a subsystem of a transverse-field Ising chain, operation near the critical point $$H_B=\omega_a \sum_{n=1}^{N} S^z_n + \omega_a J \sum_{n=1}^{N-1}\Big[(1+\gamma)S^x_nS^x_{n+1}+(1-\gamma)S^y_nS^y_{n+1}+\Delta S^z_nS^z_{n+1}\Big],$$5 enhances both extracted work and efficiency. In the single-spin battery limit, the extracted ergotropy scales as

$$H_B=\omega_a \sum_{n=1}^{N} S^z_n + \omega_a J \sum_{n=1}^{N-1}\Big[(1+\gamma)S^x_nS^x_{n+1}+(1-\gamma)S^y_nS^y_{n+1}+\Delta S^z_nS^z_{n+1}\Big],$$6

on the ordered side of the transition. The efficiency also acquires critical scaling; for generic phase choice it follows the same $$H_B=\omega_a \sum_{n=1}^{N} S^z_n + \omega_a J \sum_{n=1}^{N-1}\Big[(1+\gamma)S^x_nS^x_{n+1}+(1-\gamma)S^y_nS^y_{n+1}+\Delta S^z_nS^z_{n+1}\Big],$$7 exponent, while for the phase choice $$H_B=\omega_a \sum_{n=1}^{N} S^z_n + \omega_a J \sum_{n=1}^{N-1}\Big[(1+\gamma)S^x_nS^x_{n+1}+(1-\gamma)S^y_nS^y_{n+1}+\Delta S^z_nS^z_{n+1}\Big],$$8 the reported scaling is approximately $$H_B=\omega_a \sum_{n=1}^{N} S^z_n + \omega_a J \sum_{n=1}^{N-1}\Big[(1+\gamma)S^x_nS^x_{n+1}+(1-\gamma)S^y_nS^y_{n+1}+\Delta S^z_nS^z_{n+1}\Big],$$9 (Barra et al., 2021).

Integrable quench batteries display another critical mechanism. In the dimerized XY chain, the phase boundaries are

$2j+1$0

A double quench of the dimerization parameter produces three regimes: a short-time dimer regime, a finite-size recurrence regime, and an intermediate thermodynamic-limit regime. In that intermediate regime, if the quench crosses a phase boundary, the stored energy becomes almost unaffected by the charging time and by the precise charging parameters, yielding a robust plateau $2j+1$1 (Grazi et al., 2024).

Topological frustration modifies the picture further. In an odd-site Ising ring with periodic boundary conditions and antiferromagnetic couplings, the frustrated chain shows better resilience, charging time, and stored energy than the non-frustrated ferromagnetic chain. Under a simple discharging protocol to an external spin, only the frustrated battery transfers work rather than just heat, reaching

$2j+1$2

while the non-frustrated chain gives $2j+1$3 (Catalano et al., 2023).

Criticality is not always beneficial. In a charger-battery setup where the charger is coupled to an anisotropic XY spin-chain bath, the bath critical point at $2j+1$4 strongly suppresses long-time ergotropy, stored energy, and charging power; $2j+1$5 decays rapidly and the power becomes predominantly negative at long times (Yadav et al., 4 Nov 2025). This suggests that criticality enhances performance when it restructures the battery spectrum or the charging overlaps, but can suppress performance when it primarily amplifies dissipative environmental channels.

5. Open-system charging, noise engineering, and control protocols

Open-system spin-chain batteries are often formulated with Lindblad or non-Markovian master equations. In the cavity-hopping chain, coherent driving and thermal charging are described by

$2j+1$6

and

$2j+1$7

Without hopping, coherent charging produces transient oscillations, whereas thermal charging approaches its maximum monotonically at long times. The key nontrivial result is that thermal charging yields nonzero ergotropy for finite chains: for $2j+1$8, $2j+1$9, but for $\delta$00, $\delta$01 at steady state because the thermal bath acts on the cavity rather than directly thermalizing the spin chain, so the reduced battery state is generically nonpassive (Zhao et al., 2020).

Feedback control can convert dissipation from an obstacle into the charging resource itself. In an open XXX spin chain with local spin decay, homodyne-based quantum feedback uses a photocurrent-driven local field

$\delta$02

At zero temperature and perfect measurement efficiency $\delta$03, the optimal parameters are $\delta$04 with $\delta$05, or equivalently $\delta$06 with $\delta$07. Under these conditions the battery reaches full capacity $\delta$08 and full extractability $\delta$09, first analytically for $\delta$10 and then numerically for $\delta$11 (Yao et al., 2022).

Adaptive control appears in a different form in the cavity-Heisenberg spin-$\delta$12 chain. There the cavity-spin coupling $\delta$13 is optimized by the Soft Actor-Critic algorithm, with state $\delta$14 and action $\delta$15. In closed systems, the learned policy raises cavity-spin entanglement during charging and can push the stored energy close to $\delta$16. In open systems, the policy often switches $\delta$17 off near the pre-optimization peak, uses the environment to stabilize the energy, and then resumes charging. The reported gain is “several times” higher final stored energy than the baseline constant-coupling protocol in multiple scenarios (Sun et al., 2024).

Beyond Lindblad dynamics, strong-coupling memory effects can be regulated by energy-invariant catalysis. In the non-Markovian cavity-coupled XY chain, a low-energy catalyst qubit coupled to all spins through

$\delta$18

and constrained to approximately constant average catalyst energy suppresses ergotropy oscillations and generates a quasi-stationary regime. Increasing $\delta$19, $\delta$20, or $\delta$21 can stabilize $\delta$22, while increasing $\delta$23 toward the battery scale or simultaneously increasing the system-environment coupling $\delta$24 destabilizes work extraction (Zhao et al., 4 Aug 2025).

Noise itself can also become a control knob. In the $\delta$25 XYZ chain with local reservoirs and local Pauli noise, phase-flip noise slows charging but stabilizes both stored energy and ergotropy; bit-flip noise speeds the initial charging but saturates at low capacities; bit-phase-flip noise combines both tendencies and can outperform the pure channels in intermediate windows. In that model, the ergotropy-to-energy ratio for $\delta$26 approaches $\delta$27 at its peak and remains close to unity thereafter, so almost all stored energy is extractable (Fajar et al., 1 Sep 2025).

6. Correlations, experimental realizations, and unresolved issues

The role of genuinely quantum correlations is subtle and model-dependent. In the cavity-Heisenberg long-range battery, the Wigner function, von Neumann entropy, and logarithmic negativity show that entanglement can be necessary to store more energy, but not sufficient: $\delta$28 can continue to grow after entanglement saturates (Dou et al., 2022). In the long-range XXZ chain charged by local transverse fields, by contrast, the charging advantage is explicitly identified as a mean-field interaction effect that does not involve correlations and is reproduced by a correlation-free classical mean-field model (Le et al., 2017). In two-spin Heisenberg batteries with Dzyaloshinskii-Moriya interaction, first-order coherence increases with $\delta$29 and correlates positively with ergotropy and power, whereas quantum steering decreases and is reported not to be conducive to energy storage (Zhang et al., 2024). The finite-chain thermal-ergotropy result provides another cautionary example: nonzero ergotropy under an incoherent thermal source does not contradict passivity of genuine thermal states, because the reduced battery state is not itself thermal (Zhao et al., 2020).

Experimental proposals span cavity and circuit QED, solid-state spin chains embedded in microcavities, trapped ions with engineered spin-spin couplings, cold atoms in cavities, superconducting qudits, Rydberg arrays, NV centers, and quantum-dot or nanographene spin systems (Zhao et al., 2020, Sun et al., 2024, Siddique et al., 2024, Bhattacharya et al., 28 Aug 2025). The kicked-Ising battery goes further and has been verified on IBM hardware: on ibm_torino, a $\delta$30 periodic chain with a $\delta$31-charger was implemented for up to $\delta$32 kicks, and the measured energy-per-site curves agreed with the exact analytical predictions within error bars (Romero et al., 21 Nov 2025). Geometric generalizations beyond a strict 1D chain have also been explored: XXZ batteries with Dzyaloshinskii-Moriya interaction on supercube, cuboctahedral, and icosahedral graphs show near-ideal sinusoidal charge-discharge cycles at optimized couplings, with $\delta$33 for the $\delta$34-qubit supercube, $\delta$35 for the $\delta$36-qubit cuboctahedron, and $\delta$37 for the $\delta$38-qubit icosahedron (Bhattacharya et al., 28 Aug 2025).

Several limitations recur across the literature. Many analytically transparent studies are restricted to $\delta$39 or $\delta$40, so finite-size effects remain significant. Open-system models often include cavity loss but omit direct spin relaxation and dephasing, or use phenomenological memory kernels rather than microscopically derived non-Markovian equations (Zhao et al., 2020, Zhao et al., 4 Aug 2025). Control-theoretic works assume high measurement efficiency or idealized actuator bandwidths (Yao et al., 2022, Sun et al., 2024). Even when super-extensive power or near-perfect ergotropy fractions are reported, the robustness of those gains under realistic hardware noise, larger $\delta$41, and constrained controls remains an open problem. The data collectively indicate that the decisive design variables are not only the interaction strength and chain length, but also anisotropy, boundary conditions, spectral topology, reservoir placement, and the distinction between energy storage and extractable work.