Dicke-Like Quantum Battery Insights

• Dicke-like quantum batteries are collective energy storage systems where an ensemble of quantum cells is charged via a common bosonic mode, enabling efficient energy transfer.
• They exhibit collective scaling advantages, such as a N^(3/2) enhancement in charging power compared to independent parallel charging protocols.
• Recent studies explore various charging channels—resonant, dispersive, and nonlinear multiphoton—and address trade-offs between stored energy, ergotropy, and control in open-system environments.

Searching arXiv for recent and foundational papers on Dicke-like quantum batteries to ground the article in current literature. arXiv search query: Dicke [quantum battery](https://www.emergentmind.com/topics/quantum-battery) A Dicke-like quantum battery is a cavity-mediated many-body energy-storage architecture in which an ensemble of $N$ quantum cells—most commonly two-level systems (TLSs), but also qutrits, excitons, ions, or Andreev-bound-state doublets—is charged through collective coupling to one or more bosonic modes. Its defining structural feature is inherited from Dicke physics: the battery constituents do not interact with independent chargers, but with a common mode or with an effective interaction descended from a common mode, so charging proceeds through collective light–matter dynamics rather than purely local driving. In the literature this notion spans the full Dicke model, its Tavis–Cummings limit, dispersive descendants such as effective Lipkin–Meshkov–Glick (LMG) batteries, and a range of generalized models with additional matter–matter interactions, nonlinear couplings, inhomogeneity, or solid-state implementations (Ferraro et al., 2017).

1. Defining architecture and canonical models

In its canonical form, a Dicke-like quantum battery consists of $N$ identical TLSs with collective operators $\hat J_\alpha=\sum_{i=1}^N \hat\sigma_i^\alpha/2$, a single bosonic charger mode with annihilation and creation operators $\hat a,\hat a^\dagger$, and a collective light–matter interaction. One standard normalized Hamiltonian is

$$H=\omega_0 J_z+\omega_c a^\dagger a+\frac{2\lambda}{\sqrt N}(a+a^\dagger)J_x,$$

with the battery identified as the qubit ensemble and the charger as the cavity field; the dynamics is typically restricted to the symmetric Dicke sector $J=N/2$ when the initial state is fully symmetric (Jad et al., 11 Mar 2026). Earlier cavity-charging formulations used the time-dependent Dicke Hamiltonian

$$\hat{\mathcal H}^{(N)}_{\lambda_t}=\hbar\omega_{\rm c}\hat a^\dagger\hat a+\omega_{\rm a}\hat J_z+2\omega_{\rm c}\lambda_t\,\hat J_x(\hat a^\dagger+\hat a),$$

and contrasted this collective architecture with a parallel benchmark in which each TLS is charged by its own cavity mode (Ferraro et al., 2017).

A large part of the literature works in the resonant regime $\omega_c=\omega_a$ or $\omega_c=\omega_0$, where real photon absorption by the battery is efficient. Under weak-coupling and rotating-wave assumptions, the Dicke interaction reduces to a Tavis–Cummings form, preserving collective coupling but eliminating counter-rotating terms. In that regime, excitation-number conservation confines the dynamics to a low-dimensional symmetric manifold, which is why the model admits exact or quasi-exact charging analyses for finite $N$ (Zhang et al., 2018).

The label “Dicke-like” is broader than the bare Dicke model. It includes systems in which the common cavity mode is eliminated perturbatively, replaced by effective all-to-all interactions, or generalized by nonlinear, anisotropic, or multilevel couplings. The common structural criterion is collective coupling through a shared bosonic mediator or through an effective Hamiltonian derived from such a mediator, not strict adherence to one microscopic Hamiltonian (Gemme et al., 2023).

2. Charging channels and effective descriptions

The simplest charging channel is resonant cavity-to-battery energy transfer. In the standard resonant Dicke battery, the cavity is prepared with energy—often as photons in a Fock state or coherent state—while the TLS ensemble begins in its ground state. The interaction is switched on for a finite charging window, and the cavity field transfers energy coherently to the battery. In the large-photon Tavis–Cummings regime $N$0, the dynamics reduces to a symmetric tridiagonal problem with equally spaced spectrum, yielding a universal flip duration

$N$1

independent of $N$2 under the stated assumptions (Zhang et al., 2018).

A conceptually distinct mechanism appears in the off-resonant dispersive regime. There the cavity frequency is much larger than the TLS splitting, direct resonant absorption is suppressed, and a Schrieffer–Wolff transformation converts the Dicke Hamiltonian into an effective spin model. In the limit $N$3, the reduced battery Hamiltonian becomes

$N$4

so the charger is no longer a reservoir of real photons but the source of a virtual-photon-mediated infinite-range interaction. In that protocol, both the cavity and the TLS ensemble can start in their ground states; modulating the light–matter coupling switches on the effective interaction and charges the battery through virtual photons emerging from vacuum fluctuations (Gemme et al., 2023).

Generalized charging channels extend this picture. A two-photon Dicke battery supplements the usual linear light–matter term by a collective nonlinear coupling $N$5, allowing charging at the two-photon resonance $N$6 and producing faster dynamics than the single-photon case across weak to ultrastrong coupling (Crescente et al., 2020). In Josephson-junction realizations based on Andreev bound states, the resonator–junction coupling generates not only transverse Dicke-type terms but also longitudinal terms proportional to $N$7, which have no counterpart in the conventional Dicke model and can open additional two-photon resonant channels; the same platform also supports an equivalent charging protocol based on quenching the superconducting phase difference rather than the coupling itself (Varrica et al., 21 May 2026).

These mechanisms show that “Dicke-like” refers less to a unique microscopic interaction than to a family of collective charging architectures. Real-photon charging, virtual-photon charging, and nonlinear multiphoton charging all preserve the same organizing principle: energy transfer is controlled by collective coupling to a shared bosonic structure.

3. Figures of merit and scaling laws

The standard observables are stored energy, charging time, average charging power, and ergotropy. In off-resonant spin-battery formulations with battery Hamiltonian $N$8, the stored energy is

$N$9

its first maximum is $\hat J_\alpha=\sum_{i=1}^N \hat\sigma_i^\alpha/2$0 at time $\hat J_\alpha=\sum_{i=1}^N \hat\sigma_i^\alpha/2$1, and the averaged charging power is $\hat J_\alpha=\sum_{i=1}^N \hat\sigma_i^\alpha/2$2, with maximum $\hat J_\alpha=\sum_{i=1}^N \hat\sigma_i^\alpha/2$3 at $\hat J_\alpha=\sum_{i=1}^N \hat\sigma_i^\alpha/2$4 (Gemme et al., 2023). Open-system studies additionally track the energy cost of switching the interaction on and off, because a fast protocol is not operationally meaningful if the quench itself dominates the energetic balance (Canzio et al., 2024).

The best-known finite-$\hat J_\alpha=\sum_{i=1}^N \hat\sigma_i^\alpha/2$5 collective scaling law is the Dicke result $\hat J_\alpha=\sum_{i=1}^N \hat\sigma_i^\alpha/2$6, contrasted with $\hat J_\alpha=\sum_{i=1}^N \hat\sigma_i^\alpha/2$7 for parallel charging, hence a $\hat J_\alpha=\sum_{i=1}^N \hat\sigma_i^\alpha/2$8 enhancement in charging power for $\hat J_\alpha=\sum_{i=1}^N \hat\sigma_i^\alpha/2$9 (Ferraro et al., 2017). In the large-photon Tavis–Cummings limit, the same conclusion can be derived by a direct fair-resource comparison: if the collective protocol uses one cavity with $\hat a,\hat a^\dagger$0 photons and the parallel protocol uses $\hat a,\hat a^\dagger$1 independent cavities with $\hat a,\hat a^\dagger$2 photons each, then

$\hat a,\hat a^\dagger$3

for the same transferred energy (Zhang et al., 2018).

The dispersive Dicke-like battery shows two distinct regimes. In weak effective coupling $\hat a,\hat a^\dagger$4, perturbation theory yields superextensive formulas $\hat a,\hat a^\dagger$5 and $\hat a,\hat a^\dagger$6, but the absolute energy is very small and the single-cell ergotropy is essentially zero. In the strong effective regime $\hat a,\hat a^\dagger$7, the numerics instead give

$\hat a,\hat a^\dagger$8

so the off-resonant battery reproduces the familiar Dicke collective advantage while charging from an empty cavity through virtual photons (Gemme et al., 2023).

Beyond the single-photon Dicke architecture, two-photon collective batteries can sharpen the scaling. For the pure two-photon model, the reported finite-$\hat a,\hat a^\dagger$9 laws are

$$H=\omega_0 J_z+\omega_c a^\dagger a+\frac{2\lambda}{\sqrt N}(a+a^\dagger)J_x,$$0

which exceeds the conventional Dicke $$H=\omega_0 J_z+\omega_c a^\dagger a+\frac{2\lambda}{\sqrt N}(a+a^\dagger)J_x,$$1 power law in the regime studied (Crescente et al., 2020).

At the same time, the literature makes clear that scaling statements depend strongly on normalization conventions. Under the normalized Dicke Hamiltonian with coupling $$H=\omega_0 J_z+\omega_c a^\dagger a+\frac{2\lambda}{\sqrt N}(a+a^\dagger)J_x,$$2, a rigorous ergotropy-first quantum speed limit gives

$$H=\omega_0 J_z+\omega_c a^\dagger a+\frac{2\lambda}{\sqrt N}(a+a^\dagger)J_x,$$3

equivalently $$H=\omega_0 J_z+\omega_c a^\dagger a+\frac{2\lambda}{\sqrt N}(a+a^\dagger)J_x,$$4 with $$H=\omega_0 J_z+\omega_c a^\dagger a+\frac{2\lambda}{\sqrt N}(a+a^\dagger)J_x,$$5. This is a fixed-resource statement: at fixed $$H=\omega_0 J_z+\omega_c a^\dagger a+\frac{2\lambda}{\sqrt N}(a+a^\dagger)J_x,$$6, the time to reach a given normalized ergotropy fraction grows as $$H=\omega_0 J_z+\omega_c a^\dagger a+\frac{2\lambda}{\sqrt N}(a+a^\dagger)J_x,$$7 (Jad et al., 11 Mar 2026). An analogous normalization issue appears in the organic excitonic Dicke-like battery: under a density-preserving normalization $$H=\omega_0 J_z+\omega_c a^\dagger a+\frac{2\lambda}{\sqrt N}(a+a^\dagger)J_x,$$8, no quantum advantage is observed in the scaling of maximum stored energy density or power density, whereas under fixed cavity length $$H=\omega_0 J_z+\omega_c a^\dagger a+\frac{2\lambda}{\sqrt N}(a+a^\dagger)J_x,$$9, both quantities exhibit quantum advantages relative to the Dicke benchmark (Li et al., 1 Apr 2025).

4. Correlations, entanglement, and extractable work

A central distinction in Dicke-like batteries is that stored energy need not coincide with extractable work. Under closed pure-state evolution, the total ergotropy of the full battery can equal the stored energy, but local or reduced ergotropy can be much smaller because collective charging generates correlations. In the off-resonant LMG-type battery, the single-TLS ergotropy can vanish in weak coupling even when the global stored energy scales superextensively, because the energy is trapped in collective correlations rather than in locally usable excitations (Gemme et al., 2023).

This separation is especially explicit in multilevel generalizations. In the three-level Dicke battery, the reduced battery ergotropy $J=N/2$0 is systematically smaller than the stored energy $J=N/2$1 once charger–battery entanglement develops, and the locked energy tracks the battery von Neumann entropy closely. Among equal-energy Fock, coherent, and squeezed charger states, the coherent state yields the highest stored energy, ergotropy, and charging power because it generates the least entanglement and therefore the least locked energy; the same study reports that the system becomes asymptotically free as $J=N/2$2, with the stored energy becoming fully extractable around $J=N/2$3 (Yang et al., 2023).

Open Dicke and Tavis–Cummings batteries exhibit a related asymptotic-freedom phenomenon. Under permutation-invariant initial states and permutation-invariant generators, the passive part of the battery energy is bounded by

$J=N/2$4

so if the stored energy is extensive, the fraction of nonextractable energy vanishes as $J=N/2$5 grows. The paper derives this bound analytically and shows numerically that both Dicke and Tavis–Cummings batteries remain asymptotically free even with single-atom dissipation and dephasing (Canzio et al., 2024).

The role of entanglement in the charging advantage itself remains contested. The original Dicke battery analysis argued that the common photonic mode creates genuine many-body entanglement and a $J=N/2$6 enhancement in charging power (Ferraro et al., 2017). A later Tavis–Cummings study in the large-photon regime found a $J=N/2$7 enhancement without entanglement among the cells and attributed the effect instead to coherent cooperative interactions mediated by the common cavity mode (Zhang et al., 2018). A subsequent Comment went further, arguing that in the particular setup of Ferraro et al. the apparent $J=N/2$8 enhancement results from a stronger cavity field—hence a classical resource mismatch—rather than a genuine many-body quantum advantage (Xu et al., 2024). Taken together, these works indicate that “collective advantage” in Dicke-like batteries is not a single mechanism but a family of effects whose interpretation depends on the model, the resource normalization, and whether the figure of merit is stored energy, ergotropy, power, or speed limit.

5. Generalizations and physical realizations

The most active branch of the field extends the canonical Dicke battery by adding internal interactions or additional bosonic structures. In extended Dicke models with all-to-all interatomic interactions $J=N/2$9 and transverse driving $$\hat{\mathcal H}^{(N)}_{\lambda_t}=\hbar\omega_{\rm c}\hat a^\dagger\hat a+\omega_{\rm a}\hat J_z+2\omega_{\rm c}\lambda_t\,\hat J_x(\hat a^\dagger+\hat a),$$0, the maximum stored energy can exhibit a critical phenomenon tied to the phase structure of the underlying model, and the maximum charging power can scale as $$\hat{\mathcal H}^{(N)}_{\lambda_t}=\hbar\omega_{\rm c}\hat a^\dagger\hat a+\omega_{\rm a}\hat J_z+2\omega_{\rm c}\lambda_t\,\hat J_x(\hat a^\dagger+\hat a),$$1 with $$\hat{\mathcal H}^{(N)}_{\lambda_t}=\hbar\omega_{\rm c}\hat a^\dagger\hat a+\omega_{\rm a}\hat J_z+2\omega_{\rm c}\lambda_t\,\hat J_x(\hat a^\dagger+\hat a),$$2 up to $$\hat{\mathcal H}^{(N)}_{\lambda_t}=\hbar\omega_{\rm c}\hat a^\dagger\hat a+\omega_{\rm a}\hat J_z+2\omega_{\rm c}\lambda_t\,\hat J_x(\hat a^\dagger+\hat a),$$3 in the ultrastrong-coupling regime, while returning to the Dicke value $$\hat{\mathcal H}^{(N)}_{\lambda_t}=\hbar\omega_{\rm c}\hat a^\dagger\hat a+\omega_{\rm a}\hat J_z+2\omega_{\rm c}\lambda_t\,\hat J_x(\hat a^\dagger+\hat a),$$4 in deep-strong coupling (Dou et al., 2021). A related cavity-QED extension with dipole–dipole interactions and external cavity driving reports more stable and faster charging in weak ultrastrong coupling and a fitted exponent up to $$\hat{\mathcal H}^{(N)}_{\lambda_t}=\hbar\omega_{\rm c}\hat a^\dagger\hat a+\omega_{\rm a}\hat J_z+2\omega_{\rm c}\lambda_t\,\hat J_x(\hat a^\dagger+\hat a),$$5 (Zhang et al., 2023). More recently, collective one-axis-twisting interactions $$\hat{\mathcal H}^{(N)}_{\lambda_t}=\hbar\omega_{\rm c}\hat a^\dagger\hat a+\omega_{\rm a}\hat J_z+2\omega_{\rm c}\lambda_t\,\hat J_x(\hat a^\dagger+\hat a),$$6 have been repurposed as a charging resource: in low excitations they induce spin squeezing and soft-mode enhancement, while in high excitations they act as nonlinear torque that lowers dynamical barriers; the reported cooperative enhancement remains robust under realistic dissipation and can outperform an ideal dissipationless noninteracting Dicke battery in some regimes (Yan et al., 30 Jun 2026).

Several works transplant Dicke-like charging into explicitly engineered platforms. A trapped-ion Dicke-Ising battery replaces the photonic charger by a mechanical oscillator and augments the battery with long-range $$\hat{\mathcal H}^{(N)}_{\lambda_t}=\hbar\omega_{\rm c}\hat a^\dagger\hat a+\omega_{\rm a}\hat J_z+2\omega_{\rm c}\lambda_t\,\hat J_x(\hat a^\dagger+\hat a),$$7-$$\hat{\mathcal H}^{(N)}_{\lambda_t}=\hbar\omega_{\rm c}\hat a^\dagger\hat a+\omega_{\rm a}\hat J_z+2\omega_{\rm c}\lambda_t\,\hat J_x(\hat a^\dagger+\hat a),$$8 couplings, showing that counter-rotating-wave terms can dramatically affect both charging energy and ergotropy and can suppress a quantum phase transition by destroying quantum coherence (Wen et al., 12 Feb 2025). A two-mode extended Dicke battery introduces an auxiliary dispersive anisotropic cavity that generates effective $$\hat{\mathcal H}^{(N)}_{\lambda_t}=\hbar\omega_{\rm c}\hat a^\dagger\hat a+\omega_{\rm a}\hat J_z+2\omega_{\rm c}\lambda_t\,\hat J_x(\hat a^\dagger+\hat a),$$9 and $\omega_c=\omega_a$0 terms; in that setup both the battery energy variance and the speed of evolution scale as $\omega_c=\omega_a$1, and the maximum charging power can approach a quadratic law $\omega_c=\omega_a$2 in an optimal parameter window (Sharma et al., 17 Dec 2025).

Material-specific realizations have also become prominent. In organic microcavity batteries, the Dicke-like charger is a collectively coupled molecular layer forming exciton-polaritons, while the main limitation is self-discharge by radiative loss. A multilayer microcavity design that transfers energy from a bright charging layer to molecular triplet states in a storage layer extends the self-discharge time from the nanosecond regime of earlier bright-state devices to $\omega_c=\omega_a$3 in the demonstrated triplet-polariton regime (Tibben et al., 2024). A complementary organic excitonic battery replaces independent spins by a one-dimensional molecular aggregate with hopping $\omega_c=\omega_a$4 and exciton–exciton interaction $\omega_c=\omega_a$5, thereby realizing an interacting Frenkel–Dicke model in which optimal $\omega_c=\omega_a$6 values maximize both stored-energy density and power density (Li et al., 1 Apr 2025). In superconducting hardware, a two-dimensional-material Josephson junction inductively coupled to an LC resonator realizes a Dicke-like battery whose cells are the Andreev-bound-state doublets of the junction; here the longitudinal coupling $\omega_c=\omega_a$7 and quadratic $\omega_c=\omega_a$8 terms are intrinsic and can enhance energy storage in the appropriate resonance windows (Varrica et al., 21 May 2026).

6. Noise, control, and unresolved issues

Open-system analyses show that Dicke-like batteries are intrinsically transient devices unless the battery–charger interaction or the environmental coupling is cut off. With cavity loss, single-atom relaxation, and single-atom dephasing, the asymptotic battery state is determined by the environment rather than by the initial energy placed in the cavity, so a genuine charging process occurs only in the transient regime. In that setting, subextensive charging times can survive depending on how the coupling scales with $\omega_c=\omega_a$9, but in the Dicke model this benefit comes with a larger switch-off cost than in the Tavis–Cummings case; the same study reports that the optimal Dicke regime is off resonance, whereas Tavis–Cummings charging is optimized at resonance (Canzio et al., 2024).

Control-theoretic work has sharpened this point by showing that poor Dicke-battery performance is often a dynamical rather than purely structural problem. In the chaotic regime of the Dicke model, standard on–off charging stores energy but yields low local ergotropy because the reduced battery state becomes nearly passive. Reinforcement learning can counter this by modulating either the coupling or the detuning, greatly improving ergotropy and reducing quantum energy fluctuations; notably, in the detuning-control scheme the collective $\omega_c=\omega_0$0 speedup of charging time can be preserved even when the battery is nearly fully charged (Yanagi, 2022). A later study generalized this to an inhomogeneous Dicke battery under partial observability and found that second-order correlation observables recover most of the performance gap to the full-state baseline, reaching $\omega_c=\omega_0$1 of the fully observed policy performance (Song et al., 15 Nov 2025).

The principal unresolved issues are therefore not only experimental. They include how to compare collective and parallel protocols under strictly fair resource accounting, how to define advantage when stored energy, ergotropy, power, and switching cost do not optimize simultaneously, and how robust collective enhancements remain once dissipation, disorder, and control constraints are treated on equal footing. The existing literature supports no single universal verdict. Instead, it establishes Dicke-like quantum batteries as a broad class of collective energy-storage models whose behavior depends decisively on detuning, normalization convention, interaction architecture, extraction protocol, and the operational definition of useful work.