Quantum Batteries and Engines

Quantum batteries and engines investigate how quantum states and collective dynamics affect energy storage and conversion. The research question is which advantage survives a complete operational cycle and a fair resource comparison. Faster charging in an ideal model is not the same claim as a practical energy-storage advantage.

Required background. Quantum heat engines and energy, heat, and work establish the thermodynamic baseline.

Stored energy and extractable work are different

For a finite battery with Hamiltonian $H_B$ and state $\rho$, the maximum energy removable by a cyclic unitary operation is its ergotropy,

$$\mathcal W=\operatorname{Tr}(\rho H_B) -\min_U\operatorname{Tr}(U\rho U^\dagger H_B).$$

Here the minimum runs over all unitaries on the battery. Restricting control to a set $\mathcal A$ gives a different quantity, the accessible work $\mathcal W_{\mathcal A}\leq\mathcal W$, obtained by minimizing over $U\in\mathcal A$. A passive state can contain energy while having zero ergotropy; a state with positive ergotropy can still yield no work under a restricted control set. Thus a charging protocol should report both stored energy and the work accessible with its stated controls.

For an engine, efficiency and power require a specified cycle, reservoirs, control work, and reset. If interactions are strong or reservoirs are nonthermal, use a consistent resource ledger rather than importing a weak-coupling thermal bound without its assumptions.

Established: finite quantum models exhibit collective charging and nonclassical energy-conversion mechanisms under specified constraints. Active: scalable implementations, storage and extraction, and advantages after complete resource accounting. Campaioli et al. (2024) review the theory and experiments; Ferraro et al. (2026) discuss opportunities and implementation challenges. These sources do not justify an unrestricted claim of superiority to commercial batteries.

Literature was checked on 2026-10-11; the cited review horizon includes January 2026.

Normalize the charging resource before comparing scaling

Suppose a collective charger becomes faster as the number of cells increases. Ask whether the available interaction strength, field energy, coupling norm, or connectivity also increases. Different constraints produce different scaling comparisons. State exactly which resource is held fixed and why it matches the intended implementation.

Loss during charging, storage, and extraction can change the optimum. Entanglement can accompany a protocol without being the unique cause of its advantage. To establish a resource claim, compare with an appropriate unentangled or classical strategy under the same operations and resource budget.

Starter project: reconstruct the full energy ledger

Compare local and collective charging of a small spin battery. Choose and justify a normalization of the driving Hamiltonian, then compute stored energy, ergotropy, charging time, and peak power. Recover the single-cell limit and verify energy conservation including the charger when it is modeled dynamically.

Add a simple calibrated loss channel and a fixed storage delay. Compare the energy actually extractable after that delay, not only the peak energy during charging. Test more than one legitimate resource constraint to show which part of the conclusion depends on the comparison convention.

A laptop supports the initial exact calculation. Deliver the Hamiltonians, normalization choices, extraction assumption, and energy-versus-time and work-versus-delay curves. The failure test is an apparent scaling advantage that comes entirely from supplying a larger uncounted drive. This is a proposed feasibility exercise; a new device or theorem needs a current literature check and a physically justified constraint.

Starter project: a finite-time Otto engine at its limit cycle

Use a harmonic oscillator whose frequency alternates between $\omega_c$ and $\omega_h$. Two isolated frequency ramps perform compression and expansion; two contacts at fixed frequency exchange heat with thermal baths at $T_c<T_h$. Start with weak-coupling, detailed-balance dissipators and specified contact rates. Rezek and Kosloff’s primary study and their later review supply a tractable baseline; covariance-matrix evolution makes the first calculation feasible on a laptop.

Repeat the four strokes until the covariance matrix returns to itself within a declared tolerance. With heat positive into the working medium, record each stroke’s energy change, heat, and driving work. In this ideal weak-coupling model, a closed cycle gives

$$W_{\mathrm{out}}=Q_h+Q_c,\qquad P=\frac{W_{\mathrm{out}}}{\tau_{\mathrm{cycle}}},\qquad \eta=\frac{W_{\mathrm{out}}}{Q_h}.$$

Report efficiency only in the engine regime $Q_h>0$ and $W_{\mathrm{out}}>0$. Recover the fully thermalized, population-preserving Otto result $\eta_O=1-\omega_c/\omega_h$, with positive work requiring $1<\omega_h/\omega_c<T_h/T_c$. Lengthening all strokes indefinitely sends power to zero; this efficiency benchmark is not an optimum-power claim.

Sweep ramp and contact durations at fixed bath temperatures, frequency endpoints, contact rates, and allowed ramp speed. Compare linear ramps with one bounded smooth ramp, using the same constraints. Deliver a power–efficiency plot, stroke-by-stroke energy ledger, and convergence checks against time-step refinement and a small density-matrix calculation with an increased oscillator cutoff. Flag unstable protocols that never reach a limit cycle. Reject an apparent advantage that relies on residual stored energy, omitted stroke time, or a larger control budget.

The basic model neglects bath-switching and controller overhead. State that limitation explicitly; a full-device extension must include those costs and any imposed reset, without counting driving work twice. A research extension could ask whether a ramp’s gain survives measured frequency noise or a specified switching cost. Establish the gap in current literature before treating that extension as novel, and compare at the same delivered power as well as at each protocol’s optimum.

Research directions and decisive evidence

A theoretical project can bound extractable work with restricted local controls. A computational project can optimize a full charge–store–extract cycle under loss and control limits. An experimental project should measure accepted output energy, repetition rate, stability, and preparation/reset costs with an appropriate reference.

A convincing advance links the collective mechanism to a delivered task and shows robustness to the dominant loss and control errors. An efficient engine at vanishing power and a high-power charger with poor extraction answer different questions. The thermodynamics assessment treats the system-boundary issues that underlie both.

Entry points and references

For batteries, read the ergotropy and cycle concepts in the 2024 review, then the 2026 perspective for implementation questions. For engines, use Kosloff and Rezek’s review to connect stroke dynamics, limit cycles, and power–efficiency tradeoffs. Reassess when a full-cycle comparison or a new resource-normalized bound covers the proposed claim.

References

• F. Campaioli, S. Gherardini, J. Q. Quach, M. Polini, and G. M. Andolina, “Colloquium: Quantum batteries,” Reviews of Modern Physics 96, 031001 (2024). Preprint.
• D. Ferraro, F. Cavaliere, M. G. Genoni, G. Benenti, and M. Sassetti, “Opportunities and challenges of quantum batteries,” Nature Reviews Physics 8, 115–127 (2026). DOI.
• R. Kosloff and Y. Rezek, “The Quantum Harmonic Otto Cycle,” Entropy 19, 136 (2017). DOI — review of finite-time strokes and limit-cycle performance.
• Y. Rezek and R. Kosloff, “Irreversible Performance of a Quantum Harmonic Heat Engine,” New Journal of Physics 8, 83 (2006). DOI — primary dynamical model and finite-time performance analysis.