Dissipative chaos and steady state of open Tavis-Cummings dimer

Debabrata Mondal1, Andrey Kolovsky2,3, S. Sinha1 1Indian Institute of Science Education and Research-Kolkata, Mohanpur, Nadia-741246, India
2Kirensky Institute of Physics, Federal Research Centre KSC SB RAS, 660036 Krasnoyarsk, Russia
3School of Engineering Physics and Radio Electronics, Siberian Federal University, 660041 Krasnoyarsk, Russia
Abstract

We consider a coupled atom-photon system described by the Tavis-Cummings dimer (two coupled cavities) in the presence of photon loss and atomic pum**, to investigate the quantum signature of dissipative chaos. The appropriate classical limit of the model allows us to obtain a phase diagram identifying different dynamical phases, especially the onset of chaos. Both classically and quantum mechanically, we demonstrate the emergence of a steady state in the chaotic regime and analyze its properties. The interplay between quantum fluctuation and chaos leads to enhanced mixing dynamics and dephasing, resulting in the formation of an incoherent photonic fluid. The steady state exhibits an intriguing phenomenon of subsystem thermalization even outside the chaotic regime; however, its effective temperature increases with the degree of chaos. Moreover, the statistical properties of the steady state show a close connection with the random matrix theory. Finally, we discuss the experimental relevance of our findings, which can be tested in cavity and circuit quantum electrodynamics setups.

Introduction: Understanding the signature of chaos in quantum systems [1, 2, 3] still remains a vibrant area of research over the past decade. In spite of the lack of phase-space trajectories in quantum systems, the Bohigas-Giannoni-Schmit (BGS) conjecture plays a pivotal role in diagnosing chaos from spectral statistics of the Hamiltonian system [2]. While the connection between chaos, thermalization, and ergodicity in isolated quantum systems has been investigated extensively [4, 5, 6, 7, 8, 9], it is less explored in open quantum systems [17, 18, 19, 10, 11, 12, 13, 14, 15, 16], particularly regarding the fate of thermal steady states in dissipative environments [17, 18, 19, 20]. In recent years there has been an impetus to study dissipative quantum chaos from spectral properties of the Liouvillian [10, 21, 22, 23, 24, 25], however, such correspondence remains unclear for certain systems [25]. On the other hand, the classical-quantum correspondence in collective quantum systems with an appropriate semiclassical limit can facilitate the detection of chaos in the quantum counterpart [26, 27, 28, 10, 11, 12, 13, 14, 29, 30, 31, 32].

Ultracold atomic systems coupled to cavity modes offer a pathway to study open quantum systems, where dissipation naturally arises from various loss processes [33, 34, 35, 36, 37, 39, 40, 41, 42, 43, 44, 45, 46, 48, 47, 38, 49, 50, 51, 52]. Such atom-photon systems have recently been realized experimentally in cavity and circuit quantum electrodynamics (QED) setups, exhibiting intriguing nonequilibrium phenomena [53, 54, 55, 56, 41, 57, 58, 59, 60, 61]. Within a certain regime, a collection of two-level atoms interacting with a single cavity mode can be described by the Tavis-Cummings model [62, 63], which also facilitates the study of classical-quantum correspondence due to the appropriate classical limit for a large number of atoms.

In this work, we explore different dynamical phases, especially the onset of chaos and its quantum signature, in a dimer of atom-photon system under dissipation, which offers classical-quantum correspondence. Our study focuses on the properties of the emergent steady state in the chaotic regime, particularly the combined effects of quantum fluctuations and chaos on mixing dynamics and dephasing. Furthermore, we delve into the issue of thermalization in this open quantum system and examine the statistical properties of the steady-state density matrix.

Model and semiclassical analysis: We consider two coupled cavities, each containing N𝑁Nitalic_N two-level atoms interacting with a single cavity mode, which can be described by the Tavis-Cummings dimer (TCD) model with the following Hamiltonian

^^\displaystyle\hat{\mathcal{H}}\!\!over^ start_ARG caligraphic_H end_ARG =\displaystyle== J(a^La^R+a^Ra^L)+i=L,R[ωa^ia^i+ω0S^zi\displaystyle\!\!\!-J\left(\hat{a}_{\mathrm{L}}^{\dagger}\hat{a}_{\mathrm{R}}+% \hat{a}_{\mathrm{R}}^{\dagger}\hat{a}_{\mathrm{L}}\right)\!+\!\!\sum_{i=\rm L,% R}\left[\omega\hat{a}_{i}^{\dagger}\hat{a}_{i}+\omega_{0}\hat{S}_{zi}\right.- italic_J ( over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT roman_L end_POSTSUBSCRIPT start_POSTSUPERSCRIPT † end_POSTSUPERSCRIPT over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT roman_R end_POSTSUBSCRIPT + over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT roman_R end_POSTSUBSCRIPT start_POSTSUPERSCRIPT † end_POSTSUPERSCRIPT over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT roman_L end_POSTSUBSCRIPT ) + ∑ start_POSTSUBSCRIPT italic_i = roman_L , roman_R end_POSTSUBSCRIPT [ italic_ω over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT † end_POSTSUPERSCRIPT over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT (1)
+λ2S(a^iS^i+a^iS^i+)],\displaystyle\!\!\!+\left.\frac{\lambda}{\sqrt{2S}}\left(\hat{a}_{i}^{\dagger}% \hat{S}_{i}^{-}+\hat{a}_{i}\hat{S}_{i}^{+}\right)\!\right],+ divide start_ARG italic_λ end_ARG start_ARG square-root start_ARG 2 italic_S end_ARG end_ARG ( over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT † end_POSTSUPERSCRIPT over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT + over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) ] ,

where the site index i=L(R)𝑖LRi=\rm L(R)italic_i = roman_L ( roman_R ) represents left(right) cavity, a^isubscript^𝑎𝑖\hat{a}_{i}over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT annihilates photon mode with frequency ω𝜔\omegaitalic_ω, and J𝐽Jitalic_J is the hop** amplitude of the photons between the cavities. Collectively, N𝑁Nitalic_N two-level atoms with energy gap ω0subscript𝜔0\omega_{0}italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT can be represented by large spins Si^^subscript𝑆𝑖\hat{\vec{S_{i}}}over^ start_ARG over→ start_ARG italic_S start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG end_ARG with magnitude S=N/2𝑆𝑁2S=N/2italic_S = italic_N / 2, and λ𝜆\lambdaitalic_λ is the atom-photon coupling strength. Note that the total excitation 𝒩^=i(a^ia^i+S+S^zi)^𝒩subscript𝑖superscriptsubscript^𝑎𝑖subscript^𝑎𝑖𝑆subscript^𝑆𝑧𝑖\hat{\mathcal{N}}=\sum_{i}(\hat{a}_{i}^{\dagger}\hat{a}_{i}+S+\hat{S}_{zi})over^ start_ARG caligraphic_N end_ARG = ∑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT † end_POSTSUPERSCRIPT over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_S + over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT ) is conserved as a result of U(1)U1{\rm U}(1)roman_U ( 1 ) symmetry of ^^\hat{{\mathcal{H}}}over^ start_ARG caligraphic_H end_ARG [63]. The integrability of the single-cavity Tavis-Cummings model is broken by coupling two such cavities.

Photon loss from cavities can be compensated by atomic pum**, leading to the non-unitary evolution of the density matrix (DM) ρ^^𝜌\hat{\rho}over^ start_ARG italic_ρ end_ARG described by the Lindblad master equation [64, 65, 66],

ρ^˙=i[^,ρ^]+κi𝒟[a^i]+1Si(γ𝒟[S^i+]+γ𝒟[S^i]),˙^𝜌𝑖^^𝜌𝜅subscript𝑖𝒟delimited-[]subscript^𝑎𝑖1𝑆subscript𝑖subscript𝛾𝒟delimited-[]superscriptsubscript^𝑆𝑖subscript𝛾𝒟delimited-[]superscriptsubscript^𝑆𝑖\displaystyle\dot{\hat{\rho}}=-i[\mathcal{\hat{H}},\hat{\rho}]+\kappa\sum_{i}% \mathcal{D}[\hat{a}_{i}]+\frac{1}{S}\sum_{i}(\gamma_{\uparrow}\mathcal{D}[\hat% {S}_{i}^{+}]+\gamma_{\downarrow}\mathcal{D}[\hat{S}_{i}^{-}]),over˙ start_ARG over^ start_ARG italic_ρ end_ARG end_ARG = - italic_i [ over^ start_ARG caligraphic_H end_ARG , over^ start_ARG italic_ρ end_ARG ] + italic_κ ∑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT caligraphic_D [ over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ] + divide start_ARG 1 end_ARG start_ARG italic_S end_ARG ∑ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ( italic_γ start_POSTSUBSCRIPT ↑ end_POSTSUBSCRIPT caligraphic_D [ over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ] + italic_γ start_POSTSUBSCRIPT ↓ end_POSTSUBSCRIPT caligraphic_D [ over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT ] ) , (2)

with 𝒟[L^]=12(2L^ρ^L^L^L^ρ^ρ^L^L^)𝒟delimited-[]^𝐿122^𝐿^𝜌superscript^𝐿superscript^𝐿^𝐿^𝜌^𝜌superscript^𝐿^𝐿\mathcal{D}[\hat{L}]=\frac{1}{2}\left(2\hat{L}\hat{\rho}\hat{L}^{\dagger}-\hat% {L}^{\dagger}\hat{L}\hat{\rho}-\hat{\rho}\hat{L}^{\dagger}\hat{L}\right)caligraphic_D [ over^ start_ARG italic_L end_ARG ] = divide start_ARG 1 end_ARG start_ARG 2 end_ARG ( 2 over^ start_ARG italic_L end_ARG over^ start_ARG italic_ρ end_ARG over^ start_ARG italic_L end_ARG start_POSTSUPERSCRIPT † end_POSTSUPERSCRIPT - over^ start_ARG italic_L end_ARG start_POSTSUPERSCRIPT † end_POSTSUPERSCRIPT over^ start_ARG italic_L end_ARG over^ start_ARG italic_ρ end_ARG - over^ start_ARG italic_ρ end_ARG over^ start_ARG italic_L end_ARG start_POSTSUPERSCRIPT † end_POSTSUPERSCRIPT over^ start_ARG italic_L end_ARG ) describing the dissipative process corresponding to the Lindblad operator L^^𝐿\hat{L}over^ start_ARG italic_L end_ARG. The dissipators L^=a^i^𝐿subscript^𝑎𝑖\hat{L}=\hat{a}_{i}over^ start_ARG italic_L end_ARG = over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT and S^isuperscriptsubscript^𝑆𝑖\hat{S}_{i}^{-}over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT account for the decay processes of photon and spins with amplitudes κ𝜅\kappaitalic_κ and γsubscript𝛾\gamma_{\downarrow}italic_γ start_POSTSUBSCRIPT ↓ end_POSTSUBSCRIPT respectively. The incoherent pum** of the atoms is represented by 𝒟[S^i+]𝒟delimited-[]superscriptsubscript^𝑆𝑖\mathcal{D}[\hat{S}_{i}^{+}]caligraphic_D [ over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ] with rate γ>γsubscript𝛾subscript𝛾\gamma_{\uparrow}>\gamma_{\downarrow}italic_γ start_POSTSUBSCRIPT ↑ end_POSTSUBSCRIPT > italic_γ start_POSTSUBSCRIPT ↓ end_POSTSUBSCRIPT. From the time evolved DM, we obtain an average of any operator O^^𝑂\hat{O}over^ start_ARG italic_O end_ARG using O^=Tr(ρ^O^)delimited-⟨⟩^𝑂Tr^𝜌^𝑂\langle\hat{O}\rangle={\rm Tr}(\hat{\rho}\hat{O})⟨ over^ start_ARG italic_O end_ARG ⟩ = roman_Tr ( over^ start_ARG italic_ρ end_ARG over^ start_ARG italic_O end_ARG ). Note that, in the presence of dissipation, total excitation number 𝒩^^𝒩\hat{\mathcal{N}}over^ start_ARG caligraphic_N end_ARG is no longer conserved. Throughout the paper, we set ,kB=1Planck-constant-over-2-pisubscript𝑘𝐵1\hbar,k_{B}=1roman_ℏ , italic_k start_POSTSUBSCRIPT italic_B end_POSTSUBSCRIPT = 1 and scale energy (time) by J𝐽Jitalic_J (1/J)1𝐽(1/J)( 1 / italic_J ).

For S1much-greater-than𝑆1S\gg 1italic_S ≫ 1, the scaled operators α^i=(x^i+ıp^i)/2=a^i/Ssubscript^𝛼𝑖subscript^𝑥𝑖italic-ısubscript^𝑝𝑖2subscript^𝑎𝑖𝑆\hat{\alpha}_{i}=(\hat{x}_{i}+\imath\hat{p}_{i})/\sqrt{2}=\hat{a}_{i}/\sqrt{S}over^ start_ARG italic_α end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = ( over^ start_ARG italic_x end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_ı over^ start_ARG italic_p end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ) / square-root start_ARG 2 end_ARG = over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT / square-root start_ARG italic_S end_ARG and s^i=S^i/Ssubscript^𝑠𝑖subscript^𝑆𝑖𝑆\hat{\vec{s}}_{i}=\hat{\vec{S}}_{i}/Sover^ start_ARG over→ start_ARG italic_s end_ARG end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = over^ start_ARG over→ start_ARG italic_S end_ARG end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT / italic_S attain classical limit, as they satisfy, [α^i,α^i]=1/Ssubscript^𝛼𝑖superscriptsubscript^𝛼𝑖1𝑆[\hat{\alpha}_{i},\hat{\alpha}_{i}^{\dagger}]=1/S[ over^ start_ARG italic_α end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , over^ start_ARG italic_α end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT † end_POSTSUPERSCRIPT ] = 1 / italic_S and [s^ai,s^bi]=ıϵabcs^ci/Ssubscript^𝑠𝑎𝑖subscript^𝑠𝑏𝑖italic-ısubscriptitalic-ϵ𝑎𝑏𝑐subscript^𝑠𝑐𝑖𝑆[\hat{s}_{ai},\hat{s}_{bi}]=\imath\epsilon_{abc}\hat{s}_{ci}/S[ over^ start_ARG italic_s end_ARG start_POSTSUBSCRIPT italic_a italic_i end_POSTSUBSCRIPT , over^ start_ARG italic_s end_ARG start_POSTSUBSCRIPT italic_b italic_i end_POSTSUBSCRIPT ] = italic_ı italic_ϵ start_POSTSUBSCRIPT italic_a italic_b italic_c end_POSTSUBSCRIPT over^ start_ARG italic_s end_ARG start_POSTSUBSCRIPT italic_c italic_i end_POSTSUBSCRIPT / italic_S, where 1/S1𝑆1/S1 / italic_S plays the role of reduced Planck constant. Within the mean-field approximation, the expectation of the product of operators can be decomposed as A^B^=A^B^delimited-⟨⟩^𝐴^𝐵delimited-⟨⟩^𝐴delimited-⟨⟩^𝐵\langle\hat{A}\hat{B}\rangle=\langle\hat{A}\rangle\langle\hat{B}\rangle⟨ over^ start_ARG italic_A end_ARG over^ start_ARG italic_B end_ARG ⟩ = ⟨ over^ start_ARG italic_A end_ARG ⟩ ⟨ over^ start_ARG italic_B end_ARG ⟩, which is valid for S𝑆S\rightarrow\inftyitalic_S → ∞ [32, 52, 67]. The semiclassical dynamics of the scaled observables is obtained from Eq.(2), which is described by the following equations of motion (EOM),

αi˙˙subscript𝛼𝑖\displaystyle\dot{\alpha_{i}}over˙ start_ARG italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG =\displaystyle== (κ/2+ıω)αiı2λsi+ıαj𝜅2italic-ı𝜔subscript𝛼𝑖italic-ı2𝜆superscriptsubscript𝑠𝑖italic-ısubscript𝛼𝑗\displaystyle-\left(\kappa/2+\imath\omega\right)\alpha_{i}-\frac{\imath}{% \scalebox{0.8}{$\sqrt{2}$}}\lambda s_{i}^{-}+\imath\alpha_{j}- ( italic_κ / 2 + italic_ı italic_ω ) italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT - divide start_ARG italic_ı end_ARG start_ARG square-root start_ARG 2 end_ARG end_ARG italic_λ italic_s start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT + italic_ı italic_α start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT (3a)
s˙i+superscriptsubscript˙𝑠𝑖\displaystyle\dot{s}_{i}^{+}over˙ start_ARG italic_s end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT =\displaystyle== ıω0si+ıλ2sziαifcszisi+italic-ısubscript𝜔0superscriptsubscript𝑠𝑖italic-ı𝜆2subscript𝑠𝑧𝑖superscriptsubscript𝛼𝑖subscript𝑓𝑐subscript𝑠𝑧𝑖superscriptsubscript𝑠𝑖\displaystyle\imath\omega_{0}s_{i}^{+}-\imath\lambda\scalebox{0.8}{$\sqrt{2}$}% s_{zi}\alpha_{i}^{*}-f_{c}s_{zi}s_{i}^{+}italic_ı italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT italic_s start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT - italic_ı italic_λ square-root start_ARG 2 end_ARG italic_s start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT - italic_f start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT italic_s start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT italic_s start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT (3b)
s˙zisubscript˙𝑠𝑧𝑖\displaystyle\dot{s}_{zi}over˙ start_ARG italic_s end_ARG start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT =\displaystyle== ı2λ(αisi+αisi)+fc(1szi2)italic-ı2𝜆subscript𝛼𝑖superscriptsubscript𝑠𝑖superscriptsubscript𝛼𝑖superscriptsubscript𝑠𝑖subscript𝑓𝑐1superscriptsubscript𝑠𝑧𝑖2\displaystyle-\frac{\imath}{\scalebox{0.8}{$\sqrt{2}$}}\lambda(\alpha_{i}s_{i}% ^{+}-\alpha_{i}^{*}s_{i}^{-})+f_{c}(1-s_{zi}^{2})- divide start_ARG italic_ı end_ARG start_ARG square-root start_ARG 2 end_ARG end_ARG italic_λ ( italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT italic_s start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT - italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT italic_s start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT ) + italic_f start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 1 - italic_s start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) (3c)

where ji𝑗𝑖j\neq iitalic_j ≠ italic_i, fc=γγsubscript𝑓𝑐subscript𝛾subscript𝛾f_{c}=\gamma_{\uparrow}-\gamma_{\downarrow}italic_f start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT = italic_γ start_POSTSUBSCRIPT ↑ end_POSTSUBSCRIPT - italic_γ start_POSTSUBSCRIPT ↓ end_POSTSUBSCRIPT, αi=niexp(iψi)=(xi+ιpi)/2subscript𝛼𝑖subscript𝑛𝑖𝑖subscript𝜓𝑖subscript𝑥𝑖𝜄subscript𝑝𝑖2\alpha_{i}=\sqrt{n_{i}}\exp(-i\psi_{i})=(x_{i}+\iota p_{i})/\sqrt{2}italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = square-root start_ARG italic_n start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_ARG roman_exp ( - italic_i italic_ψ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ) = ( italic_x start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT + italic_ι italic_p start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ) / square-root start_ARG 2 end_ARG and si=(sinθicosϕi,sinθisinϕi,cosθi)subscript𝑠𝑖subscript𝜃𝑖subscriptitalic-ϕ𝑖subscript𝜃𝑖subscriptitalic-ϕ𝑖subscript𝜃𝑖\vec{s}_{i}=(\sin\theta_{i}\cos\phi_{i},\sin\theta_{i}\sin\phi_{i},\cos\theta_% {i})over→ start_ARG italic_s end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT = ( roman_sin italic_θ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT roman_cos italic_ϕ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , roman_sin italic_θ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT roman_sin italic_ϕ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , roman_cos italic_θ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ).

Next, we investigate the stable fixed points (FP) and other attractors of the EOM (see Eq.(3), describing various nonequilibrium phases of this open TCD, listed below and summarized in the phase diagram in Fig.1(a).

Refer to caption
Figure 1: Dynamical phases and steady states: (a) The classical phase-diagram in λκ𝜆𝜅\lambda-\kappaitalic_λ - italic_κ plane with average Lyapunov exponent Λ¯¯Λ\bar{\Lambda}over¯ start_ARG roman_Λ end_ARG as color scale. The NP to SR transition at λcsubscript𝜆𝑐\lambda_{c}italic_λ start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT (solid line) and instability of SR phase at λIsubscript𝜆I\lambda_{\rm I}italic_λ start_POSTSUBSCRIPT roman_I end_POSTSUBSCRIPT (dashed line) are shown. (b) The chaotic trajectory is represented by scattered points on the Bloch sphere at regular time intervals (λ=4.0,κ=0.05formulae-sequence𝜆4.0𝜅0.05\lambda=4.0,\kappa=0.05italic_λ = 4.0 , italic_κ = 0.05 (blue circle in (a))). Steady state: Variation of average (c) photon number nLdelimited-⟨⟩subscript𝑛L\langle n_{\rm L}\rangle⟨ italic_n start_POSTSUBSCRIPT roman_L end_POSTSUBSCRIPT ⟩ and (d) spin polarization szLdelimited-⟨⟩subscript𝑠𝑧L\langle s_{z{\rm L}}\rangle⟨ italic_s start_POSTSUBSCRIPT italic_z roman_L end_POSTSUBSCRIPT ⟩ of one cavitity with λ𝜆\lambdaitalic_λ for κ=0.05𝜅0.05\kappa=0.05italic_κ = 0.05. TWA results (blue dashed line), quantum steady-state value (solid green line) and fixed point (n,szsuperscript𝑛superscriptsubscript𝑠𝑧n^{*},s_{z}^{*}italic_n start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT , italic_s start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT) of SR phase are compared. The stable (unstable) SR phase is shown by black solid (dashed dotted) line. The color scales in (c,d) represent Λ¯¯Λ\bar{\Lambda}over¯ start_ARG roman_Λ end_ARG. All energies (time) are measured by J(1/J)𝐽1𝐽J(1/J)italic_J ( 1 / italic_J ). We set ,kB=1Planck-constant-over-2-pisubscript𝑘𝐵1\hbar,k_{B}=1roman_ℏ , italic_k start_POSTSUBSCRIPT italic_B end_POSTSUBSCRIPT = 1 and ω=2.0,ω0=0.5,γ=0.2,γ=0.1,S=2formulae-sequence𝜔2.0formulae-sequencesubscript𝜔00.5formulae-sequencesubscript𝛾0.2formulae-sequencesubscript𝛾0.1𝑆2\omega=2.0,\omega_{0}=0.5,\gamma_{\uparrow}=0.2,\gamma_{\downarrow}=0.1,S=2italic_ω = 2.0 , italic_ω start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT = 0.5 , italic_γ start_POSTSUBSCRIPT ↑ end_POSTSUBSCRIPT = 0.2 , italic_γ start_POSTSUBSCRIPT ↓ end_POSTSUBSCRIPT = 0.1 , italic_S = 2 for all figures.

Normal phase (NP): This phase is characterized by vanishing photon number ni=0superscriptsubscript𝑛𝑖0n_{i}^{*}=0italic_n start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT = 0 and spin polarization szi=1superscriptsubscript𝑠𝑧𝑖1s_{zi}^{*}=1italic_s start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT = 1.

Superradiant phase (SR): At a critical coupling λcsubscript𝜆𝑐\lambda_{c}italic_λ start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT the normal phase undergoes a continuous transition to the superradiant phase (see Fig.1(a)) with a non-vanishing photon number n0superscript𝑛0n^{*}\neq 0italic_n start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT ≠ 0 and spin polarization |sz|<1superscriptsubscript𝑠𝑧1|s_{z}^{*}|<1| italic_s start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT | < 1, same for each cavity. Due to the U(1) symmetry, FPs lie on a circle of radius 2n2superscript𝑛\sqrt{2n^{*}}square-root start_ARG 2 italic_n start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT end_ARG (1(sz)21superscriptsuperscriptsubscript𝑠𝑧2\sqrt{1-(s_{z}^{*})^{2}}square-root start_ARG 1 - ( italic_s start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT end_ARG) in xp𝑥𝑝x-pitalic_x - italic_p (sxsysubscript𝑠𝑥subscript𝑠𝑦s_{x}-s_{y}italic_s start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT - italic_s start_POSTSUBSCRIPT italic_y end_POSTSUBSCRIPT) plane, where the dynamics always converges regardless of the initial condition. The details of this phase are given in the supplementary material [68].

Oscillatory phase (OP): Once the SR phase becomes unstable for λ>λI𝜆subscript𝜆I\lambda>\lambda_{\rm I}italic_λ > italic_λ start_POSTSUBSCRIPT roman_I end_POSTSUBSCRIPT (see Fig.1(a,c,d)), the oscillatory phase emerges, where the periodic motion is identified from a single peak in the Fourier transform of trajectories [68]. Although the photon number oscillates, its phase in both cavities remains the same ψL=ψRsubscript𝜓Lsubscript𝜓R\psi_{\rm L}=\psi_{\rm R}italic_ψ start_POSTSUBSCRIPT roman_L end_POSTSUBSCRIPT = italic_ψ start_POSTSUBSCRIPT roman_R end_POSTSUBSCRIPT, similar to the SR phase.

Coexistance of chaos and oscillatory dynamics: Further increasing λ𝜆\lambdaitalic_λ gives rise to the mixed type of dynamics, where oscillatory motion coexists with chaotic dynamics, depending on the initial conditions.

Chaotic dynamics: We also identify a regime in the phase diagram where the trajectories exhibit chaotic behavior, as depicted in Fig.1(b). To quantify the degree of chaos, we compute the mean Lyapunov exponent Λ¯¯Λ\bar{\Lambda}over¯ start_ARG roman_Λ end_ARG [69, 70, 71] averaged over random initial phase space points, as shown in Fig.1(a).

The stable fixed point attractors, such as NP and SR phases, uniquely describe the system’s asymptotic steady state. To understand the state of the system after the instability of the SR phase, we study the semiclassical dynamics on an ensemble of phase space points using truncated Wigner approximation (TWA) [72, 73, 74]. Quantum fluctuations are incorporated by sampling initial conditions from the Husimi distribution [75] of a product of bosonic and spin coherent states, |Ψc=i|αi|szi,ϕiketsubscriptΨ𝑐subscriptproduct𝑖tensor-productketsubscript𝛼𝑖ketsubscript𝑠𝑧𝑖subscriptitalic-ϕ𝑖|{\Psi_{c}}\rangle=\prod_{i}|\alpha_{i}\rangle\otimes|s_{zi},\phi_{i}\rangle| roman_Ψ start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ⟩ = ∏ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT | italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ⟩ ⊗ | italic_s start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT , italic_ϕ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ⟩ [76] corresponding to large spin S1much-greater-than𝑆1S\gg 1italic_S ≫ 1, which semiclassically represents an arbitrary phase space point {αi,szi=cosθi,ϕi}formulae-sequencesubscript𝛼𝑖subscript𝑠𝑧𝑖subscript𝜃𝑖subscriptitalic-ϕ𝑖\{\alpha_{i},s_{zi}=\cos\theta_{i},\phi_{i}\}{ italic_α start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_s start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT = roman_cos italic_θ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , italic_ϕ start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT }. Importantly, we observe the emergence of a unique steady state in both oscillatory and chaotic phases, characterized by the stationary value of physical quantities such as photon number nTWAsubscriptdelimited-⟨⟩𝑛TWA\langle n\rangle_{\scriptscriptstyle\text{TWA}}⟨ italic_n ⟩ start_POSTSUBSCRIPT TWA end_POSTSUBSCRIPT and spin polarization szTWAsubscriptdelimited-⟨⟩subscript𝑠𝑧TWA\langle s_{z}\rangle_{\scriptscriptstyle\text{TWA}}⟨ italic_s start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT ⟩ start_POSTSUBSCRIPT TWA end_POSTSUBSCRIPT after sufficient time, which smoothly connects to the stable SR phase, as shown in Fig.1(c,d). Moreover, these dynamical variables follow stationary distribution P(ni)subscript𝑛𝑖(n_{i})( italic_n start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ), P(szi)subscript𝑠𝑧𝑖(s_{zi})( italic_s start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT ) irrespective of initial condition (see Fig.2(a,b)). Notably, Our analysis reveals the ergodic nature of the steady state, which is based on (i) the equivalence between time and ensemble averages Ot=OTWAsubscriptdelimited-⟨⟩𝑂𝑡subscriptdelimited-⟨⟩𝑂TWA\langle O\rangle_{t}=\langle O\rangle_{\scriptscriptstyle\text{TWA}}⟨ italic_O ⟩ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT = ⟨ italic_O ⟩ start_POSTSUBSCRIPT TWA end_POSTSUBSCRIPT of dynamical quantities and (ii) their independence from initial condition [77, 78, 80, 79].

The different dynamical regimes of the steady state particularly, the onset of chaos can be revealed from the autocorrelation function,

CTWA(t)=O(t+τ)O(τ)O(t+τ)O(τ),superscript𝐶TWA𝑡delimited-⟨⟩𝑂𝑡𝜏𝑂𝜏delimited-⟨⟩𝑂𝑡𝜏delimited-⟨⟩𝑂𝜏\displaystyle C^{\scriptscriptstyle\text{TWA}}(t)=\langle O(t+\tau)O(\tau)% \rangle-\langle O(t+\tau)\rangle\langle O(\tau)\rangle,italic_C start_POSTSUPERSCRIPT TWA end_POSTSUPERSCRIPT ( italic_t ) = ⟨ italic_O ( italic_t + italic_τ ) italic_O ( italic_τ ) ⟩ - ⟨ italic_O ( italic_t + italic_τ ) ⟩ ⟨ italic_O ( italic_τ ) ⟩ , (4)

computed within TWA. Here, the observable O𝑂Oitalic_O is first evolved up to a long transient time τ𝜏\tauitalic_τ until O(τ)TWAsubscriptdelimited-⟨⟩𝑂𝜏TWA\langle O(\tau)\rangle_{\scriptscriptstyle\text{TWA}}⟨ italic_O ( italic_τ ) ⟩ start_POSTSUBSCRIPT TWA end_POSTSUBSCRIPT reaches the steady state. Although the TWA analysis indicates the formation of a steady state in the oscillatory regime, the autocorrelation function exhibits persistent oscillations, revealing its signature (see Fig.2(c)). On the contrary, in the chaotic regime, the autocorrelation function decays rapidly, as seen from Fig.2(d), confirming chaotic mixing [79, 81, 80]. Interestingly, the steady state in the oscillatory regime displays ergodicity in the absence of mixing. This contrasts with the typical closed system, where chaotic mixing leads to the ergodic steady state.

Refer to caption
Figure 2: Comparison between oscillatory and chaotic phases: Steady-state distribution and autocorrelation function of the spin polarization szsubscript𝑠𝑧s_{z}italic_s start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT of one of the cavities in the (a,c) oscillatory (OP) (λ=1.4𝜆1.4\lambda=1.4italic_λ = 1.4, violet square in Fig.1(a)), (b,d) chaotic regime (λ=4.0𝜆4.0\lambda=4.0italic_λ = 4.0, blue circle in Fig.1(a)). Squares and circles in (a,b) represent distributions starting from different initial conditions. We choose κ=0.05𝜅0.05\kappa=0.05italic_κ = 0.05.

Quantum steady state: To this end, we investigate the properties of the quantum steady state and the signature of chaos in the presence of quantum fluctuation. We obtain the steady state density matrix (DM) ρ^sssubscript^𝜌ss\hat{\rho}_{\rm ss}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT by solving the Master equation (see Eq.(2)) using the stochastic wavefunction approach [82, 83, 75], considering spin magnitude S=2𝑆2S=2italic_S = 2.

Refer to caption
Figure 3: Quantum dynamics: (a) Evolution of total entropy 𝒮totsubscript𝒮tot\mathcal{S}_{\rm tot}caligraphic_S start_POSTSUBSCRIPT roman_tot end_POSTSUBSCRIPT and (b) survival probability f(t)¯¯𝑓𝑡\overline{f(t)}over¯ start_ARG italic_f ( italic_t ) end_ARG at different coupling λ𝜆\lambdaitalic_λ. Inset in (b) shows variation of decay time τdsubscript𝜏𝑑\tau_{d}italic_τ start_POSTSUBSCRIPT italic_d end_POSTSUBSCRIPT of f(t)¯¯𝑓𝑡\overline{f(t)}over¯ start_ARG italic_f ( italic_t ) end_ARG with λ𝜆\lambdaitalic_λ. (c) Dynamics of autocorrelation of spin Cz(t)subscript𝐶𝑧𝑡C_{z}(t)italic_C start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT ( italic_t ) for different λ𝜆\lambdaitalic_λ. (d) Steady-state value of the coherence function 𝒞LRsubscript𝒞LR\mathcal{C}_{\rm LR}caligraphic_C start_POSTSUBSCRIPT roman_LR end_POSTSUBSCRIPT of the photon field between the cavities (solid red), corresponding TWA result (dashed greenish blue) and the purity 𝒫𝒫\mathcal{P}caligraphic_P (solid blue, right axis), as a function of λ𝜆\lambdaitalic_λ. For all figures, κ=0.05𝜅0.05\kappa=0.05italic_κ = 0.05.

Similar to the classical analysis, a unique quantum steady state is formed across different dynamical regimes with increasing coupling strength. This state is characterized by the average quantities n^isssubscriptdelimited-⟨⟩subscript^𝑛𝑖ss\langle\hat{n}_{i}\rangle_{\rm ss}⟨ over^ start_ARG italic_n end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ⟩ start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT and S^zisssubscriptdelimited-⟨⟩subscript^𝑆𝑧𝑖ss\langle\hat{S}_{zi}\rangle_{\rm ss}⟨ over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT ⟩ start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT obtained from ρ^sssubscript^𝜌ss\hat{\rho}_{\rm ss}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT, as shown in Fig.1(c,d). For comparison with classical results, we scale these physical quantities by S𝑆Sitalic_S. The steady state also exhibits ergodicity analogous to its classical counterpart, as the time average of the observables n^itsubscriptdelimited-⟨⟩subscript^𝑛𝑖𝑡\langle\hat{n}_{i}\rangle_{t}⟨ over^ start_ARG italic_n end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ⟩ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT and S^zitsubscriptdelimited-⟨⟩subscript^𝑆𝑧𝑖𝑡\langle\hat{S}_{zi}\rangle_{t}⟨ over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT ⟩ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT over a typical quantum trajectory approach to their steady-state values n^isssubscriptdelimited-⟨⟩subscript^𝑛𝑖ss\langle\hat{n}_{i}\rangle_{\rm ss}⟨ over^ start_ARG italic_n end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ⟩ start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT and S^zisssubscriptdelimited-⟨⟩subscript^𝑆𝑧𝑖ss\langle\hat{S}_{zi}\rangle_{\rm ss}⟨ over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT ⟩ start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT respectively [68].

Next, we study the nonequilibrium dynamics to investigate the signature of quantum mixing, as the chaotic regime is approached with increasing λ𝜆\lambdaitalic_λ. Starting from an arbitrary initial coherent state |ΨcketsubscriptΨ𝑐|{\Psi_{c}}\rangle| roman_Ψ start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ⟩, representing a phase space point, we study the time evolution of survival probability f(t)=Tr(ρ^(t)ρ^(0))𝑓𝑡Tr^𝜌𝑡^𝜌0f(t)={\rm Tr}(\hat{\rho}(t)\hat{\rho}(0))italic_f ( italic_t ) = roman_Tr ( over^ start_ARG italic_ρ end_ARG ( italic_t ) over^ start_ARG italic_ρ end_ARG ( 0 ) ) [84, 85, 86, 87] between the initial and time-evolved density matrices. The survival probability f(t)¯¯𝑓𝑡\overline{f(t)}over¯ start_ARG italic_f ( italic_t ) end_ARG, averaged over an ensemble of initial states, decays exponentially with time and the decay rate increases as we approach the chaotic regime (see Fig.3(b)).

Refer to caption
Figure 4: Subsystem thermalization: (a) Overlap between subsystem reduced DM obtained from the steady state ρ^sssubscript^𝜌ss\hat{\rho}_{\rm ss}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT and the corresponding DM derived from the thermal state ρ^thsuperscript^𝜌th\hat{\rho}^{\rm th}over^ start_ARG italic_ρ end_ARG start_POSTSUPERSCRIPT roman_th end_POSTSUPERSCRIPT. The overlap of the left cavity (LsubscriptL\mathcal{F}_{\rm L}caligraphic_F start_POSTSUBSCRIPT roman_L end_POSTSUBSCRIPT), photon (PhsubscriptPh\mathcal{F}_{\rm Ph}caligraphic_F start_POSTSUBSCRIPT roman_Ph end_POSTSUBSCRIPT), spin (SsubscriptS\mathcal{F}_{\rm S}caligraphic_F start_POSTSUBSCRIPT roman_S end_POSTSUBSCRIPT) of the same cavity (left axis) and effective temperature T𝑇Titalic_T (right axis) as a function of λ𝜆\lambdaitalic_λ. Variation of (b) total entropy 𝒮totsubscript𝒮tot\mathcal{S}_{\rm tot}caligraphic_S start_POSTSUBSCRIPT roman_tot end_POSTSUBSCRIPT, entanglement entropy of the left cavity 𝒮Lsubscript𝒮L\mathcal{S}_{\rm L}caligraphic_S start_POSTSUBSCRIPT roman_L end_POSTSUBSCRIPT and photon field of that cavity 𝒮Phsubscript𝒮Ph\mathcal{S}_{\rm Ph}caligraphic_S start_POSTSUBSCRIPT roman_Ph end_POSTSUBSCRIPT in the steady state ρ^sssubscript^𝜌ss\hat{\rho}_{\rm ss}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT with λ𝜆\lambdaitalic_λ. The solid lines (markers) represent the entropies obtained from ρ^sssubscript^𝜌ss\hat{\rho}_{\rm ss}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT (ρ^thsuperscript^𝜌th\hat{\rho}^{\rm th}over^ start_ARG italic_ρ end_ARG start_POSTSUPERSCRIPT roman_th end_POSTSUPERSCRIPT). Classical phases are marked by blue colored regimes. Statistics of ρ^sssubscript^𝜌ss\hat{\rho}_{\rm ss}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT: (c) distribution of structural entropy 𝒮νstrsuperscriptsubscript𝒮𝜈str\mathcal{S}_{\nu}^{\rm str}caligraphic_S start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_str end_POSTSUPERSCRIPT of eigenstates of ρ^sssubscript^𝜌ss\hat{\rho}_{\rm ss}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT in the chaotic regime. Black dashed line represents the corresponding GUE value. (d) Distribution of elements η=|Ψνj|2Ndim𝜂superscriptsuperscriptsubscriptΨ𝜈𝑗2subscript𝑁dim\eta=|\Psi_{\nu}^{j}|^{2}N_{\rm dim}italic_η = | roman_Ψ start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_j end_POSTSUPERSCRIPT | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_N start_POSTSUBSCRIPT roman_dim end_POSTSUBSCRIPT of typical eigenstate of steady state at different λ𝜆\lambdaitalic_λ. Black dashed line represents the corresponding GUE distribution. Parameter chosen: κ=0.05𝜅0.05\kappa=0.05italic_κ = 0.05.

From the time evolved density matrix, we also obtain the total entropy 𝒮tot=Tr(ρ^(t)lnρ^(t))subscript𝒮totTr^𝜌𝑡^𝜌𝑡\mathcal{S}_{\rm tot}=-{\rm Tr}(\hat{\rho}(t)\ln\hat{\rho}(t))caligraphic_S start_POSTSUBSCRIPT roman_tot end_POSTSUBSCRIPT = - roman_Tr ( over^ start_ARG italic_ρ end_ARG ( italic_t ) roman_ln over^ start_ARG italic_ρ end_ARG ( italic_t ) ), which grows linearly and attains a saturation value corresponding to the steady state. As evident from Fig.3(a), both the growth rate and the saturation value of 𝒮totsubscript𝒮tot\mathcal{S}_{\rm tot}caligraphic_S start_POSTSUBSCRIPT roman_tot end_POSTSUBSCRIPT increases with λ𝜆\lambdaitalic_λ. Also, the reduced density matrix ρ^A=TrA¯(ρ^)subscript^𝜌𝐴subscriptTr¯𝐴^𝜌\hat{\rho}_{A}={\rm Tr}_{\bar{A}}(\hat{\rho})over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT = roman_Tr start_POSTSUBSCRIPT over¯ start_ARG italic_A end_ARG end_POSTSUBSCRIPT ( over^ start_ARG italic_ρ end_ARG ) of subsystem A𝐴Aitalic_A can be obtained by tracing out the remaining degrees of freedom A¯¯𝐴\bar{A}over¯ start_ARG italic_A end_ARG, which in turn yields the corresponding entanglement entropy (EE) 𝒮A=Tr(ρ^Alnρ^A)subscript𝒮𝐴Trsubscript^𝜌𝐴subscript^𝜌𝐴\mathcal{S}_{A}=-{\rm Tr}(\hat{\rho}_{A}\ln\hat{\rho}_{A})caligraphic_S start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT = - roman_Tr ( over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT roman_ln over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT ). Additionally, the EE corresponding to one of the cavities 𝒮Lsubscript𝒮L\mathcal{S}_{\rm L}caligraphic_S start_POSTSUBSCRIPT roman_L end_POSTSUBSCRIPT exhibits similar behavior [68]. Such linear growth in EE is typically observed as a signature of chaos in isolated quantum systems [4, 6, 88, 89, 90, 91]. The underlying chaos can also be unveiled from the individual quantum trajectories, which show a spreading of the power spectrum over a wide range of frequencies [68, 92].

We also analyze the autocorrelation function corresponding to the steady state C(t)=O(t)O(0)O(t)O(0)𝐶𝑡delimited-⟨⟩𝑂𝑡𝑂0delimited-⟨⟩𝑂𝑡delimited-⟨⟩𝑂0C(t)=\langle O(t)O(0)\rangle-\langle O(t)\rangle\langle O(0)\rangleitalic_C ( italic_t ) = ⟨ italic_O ( italic_t ) italic_O ( 0 ) ⟩ - ⟨ italic_O ( italic_t ) ⟩ ⟨ italic_O ( 0 ) ⟩, following the prescription of stochastic wave-function [93, 83]. As seen from Fig.3(c), the autocorrelation function Cz(t)subscript𝐶𝑧𝑡C_{z}(t)italic_C start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT ( italic_t ) of the spin in one cavity vanishes rapidly and the decay time decreases with increasing coupling strength λ𝜆\lambdaitalic_λ. Such fast decay of Cz(t)subscript𝐶𝑧𝑡C_{z}(t)italic_C start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT ( italic_t ) signifies enhanced mixing dynamics due to the onset of chaos.

In addition, the quantum fluctuations enhance the decay rate of Cz(t)subscript𝐶𝑧𝑡C_{z}(t)italic_C start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT ( italic_t ) as compared to the classical counterpart. Importantly, the oscillatory phase is washed out due to the strong quantum fluctuation, which is evident from the decaying autocorrelation function (see Fig.3(c)). Furthermore, the combined effect of quantum fluctuations and chaotic mixing suppresses the purity of the steady state 𝒫=Tr(ρ^ss2)𝒫Trsuperscriptsubscript^𝜌ss2\mathcal{P}={\rm Tr}(\hat{\rho}_{\rm ss}^{2})caligraphic_P = roman_Tr ( over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) significantly, even in the superradiant phase, as depicted in Fig.3(d). The coherence of the photon fields between the two cavities in the steady state can be quantified from,

𝒞LR=a^La^R+a^Ra^L2n^Ln^R,subscript𝒞LRdelimited-⟨⟩superscriptsubscript^𝑎Lsubscript^𝑎Rsuperscriptsubscript^𝑎Rsubscript^𝑎𝐿2delimited-⟨⟩subscript^𝑛Ldelimited-⟨⟩subscript^𝑛R\displaystyle\mathcal{C}_{\rm LR}=\frac{\langle\hat{a}_{\rm L}^{\dagger}\hat{a% }_{\rm R}+\hat{a}_{\rm R}^{\dagger}\hat{a}_{L}\rangle}{2\sqrt{\langle\hat{n}_{% \rm L}\rangle\langle\hat{n}_{\rm R}\rangle}},caligraphic_C start_POSTSUBSCRIPT roman_LR end_POSTSUBSCRIPT = divide start_ARG ⟨ over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT roman_L end_POSTSUBSCRIPT start_POSTSUPERSCRIPT † end_POSTSUPERSCRIPT over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT roman_R end_POSTSUBSCRIPT + over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT roman_R end_POSTSUBSCRIPT start_POSTSUPERSCRIPT † end_POSTSUPERSCRIPT over^ start_ARG italic_a end_ARG start_POSTSUBSCRIPT italic_L end_POSTSUBSCRIPT ⟩ end_ARG start_ARG 2 square-root start_ARG ⟨ over^ start_ARG italic_n end_ARG start_POSTSUBSCRIPT roman_L end_POSTSUBSCRIPT ⟩ ⟨ over^ start_ARG italic_n end_ARG start_POSTSUBSCRIPT roman_R end_POSTSUBSCRIPT ⟩ end_ARG end_ARG , (5)

which classically reduces to 𝒞LRTWA=cos(ψLψR)TWAsuperscriptsubscript𝒞LRTWAsubscriptdelimited-⟨⟩subscript𝜓Lsubscript𝜓RTWA\mathcal{C}_{\rm LR}^{\rm TWA}=\langle\cos(\psi_{\rm L}-\psi_{\rm R})\rangle_{% \scriptscriptstyle\text{TWA}}caligraphic_C start_POSTSUBSCRIPT roman_LR end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_TWA end_POSTSUPERSCRIPT = ⟨ roman_cos ( italic_ψ start_POSTSUBSCRIPT roman_L end_POSTSUBSCRIPT - italic_ψ start_POSTSUBSCRIPT roman_R end_POSTSUBSCRIPT ) ⟩ start_POSTSUBSCRIPT TWA end_POSTSUBSCRIPT. In the superradiant and oscillatory phase, 𝒞LRTWAsuperscriptsubscript𝒞LRTWA\mathcal{C}_{\rm LR}^{\rm TWA}caligraphic_C start_POSTSUBSCRIPT roman_LR end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_TWA end_POSTSUPERSCRIPT attains the value unity and decays in the chaotic regime, as observed from the TWA analysis. Similar behavior can also be observed quantum mechanically, however, the decay starts even before the instability of SR phase, as a result of quantum fluctuations (see Fig.3(d)). It is clear from this analysis that the mixing dynamics due to the underlying chaos as well as quantum fluctuation can lead to the destruction of the coherence of the system, which can be probed from the state of the photon field of the respective cavities. The semiclassical distribution of the photon field for the sufficiently small value of λ𝜆\lambdaitalic_λ in the stable SR phase is peaked around the circle of the classical fixed points of radius 2n2superscript𝑛\sqrt{2n^{*}}square-root start_ARG 2 italic_n start_POSTSUPERSCRIPT ∗ end_POSTSUPERSCRIPT end_ARG. As we approach the chaotic regime, this distribution spreads and peaks around the center, resembling a thermal distribution [68]. Such an observation suggests chaos-induced thermalization in this dissipative system, which we discuss next.

Subsystem thermalization: To investigate thermalization of this open system, we consider the thermal density matrix of TCD ρ^th=exp(β(^μ𝒩^))/Zsuperscript^𝜌th𝛽^𝜇^𝒩𝑍\hat{\rho}^{\rm th}=\exp(-\beta(\hat{\mathcal{H}}-\mu\hat{\mathcal{N}}))/Zover^ start_ARG italic_ρ end_ARG start_POSTSUPERSCRIPT roman_th end_POSTSUPERSCRIPT = roman_exp ( - italic_β ( over^ start_ARG caligraphic_H end_ARG - italic_μ over^ start_ARG caligraphic_N end_ARG ) ) / italic_Z with Z𝑍Zitalic_Z being the partition function. The inverse temperature β𝛽\betaitalic_β and the chemical potential μ𝜇\muitalic_μ are uniquely determined from the mean energy ^delimited-⟨⟩^\langle\hat{\mathcal{H}}\rangle⟨ over^ start_ARG caligraphic_H end_ARG ⟩ and the average number of excitations 𝒩^delimited-⟨⟩^𝒩\langle\hat{\mathcal{N}}\rangle⟨ over^ start_ARG caligraphic_N end_ARG ⟩ obtained from the steady-state density matrix ρ^sssubscript^𝜌ss\hat{\rho}_{\rm ss}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT. Clearly, ρ^thsuperscript^𝜌th\hat{\rho}^{\rm th}over^ start_ARG italic_ρ end_ARG start_POSTSUPERSCRIPT roman_th end_POSTSUPERSCRIPT does not correspond to the steady state of the Master equation (Eq.(2)) in the presence of dissipation. However, the state of the subsystem A𝐴Aitalic_A can be well described by the reduced DM ρ^Athsuperscriptsubscript^𝜌𝐴th\hat{\rho}_{A}^{\rm th}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_th end_POSTSUPERSCRIPT derived from the thermal state ρ^thsuperscript^𝜌th\hat{\rho}^{\rm th}over^ start_ARG italic_ρ end_ARG start_POSTSUPERSCRIPT roman_th end_POSTSUPERSCRIPT of the full system. We compare the reduced DMs ρ^Asssuperscriptsubscript^𝜌𝐴ss\hat{\rho}_{A}^{\rm ss}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_ss end_POSTSUPERSCRIPT and ρ^Athsuperscriptsubscript^𝜌𝐴th\hat{\rho}_{A}^{\rm th}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_th end_POSTSUPERSCRIPT obtained from the steady state and thermal DM respectively, using their overlap A=Trρ^Assρ^Athρ^Asssubscript𝐴Trsuperscriptsubscript^𝜌𝐴sssuperscriptsubscript^𝜌𝐴thsuperscriptsubscript^𝜌𝐴ss\mathcal{F}_{A}={\rm Tr}\sqrt{\sqrt{\hat{\rho}_{A}^{\rm ss}}\hat{\rho}_{A}^{% \rm th}\sqrt{\hat{\rho}_{A}^{\rm ss}}}caligraphic_F start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT = roman_Tr square-root start_ARG square-root start_ARG over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_ss end_POSTSUPERSCRIPT end_ARG over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_th end_POSTSUPERSCRIPT square-root start_ARG over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_ss end_POSTSUPERSCRIPT end_ARG end_ARG [84, 85]. Considering one of the cavities as well as its photonic and spin sector as subsystems, we compute Asubscript𝐴\mathcal{F}_{A}caligraphic_F start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT, which remains close to unity, as evident from Fig.4(a). Moreover, trace-distances between ρ^Asssuperscriptsubscript^𝜌𝐴ss\hat{\rho}_{A}^{\rm ss}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_ss end_POSTSUPERSCRIPT and ρ^Athsuperscriptsubscript^𝜌𝐴th\hat{\rho}_{A}^{\rm th}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT italic_A end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_th end_POSTSUPERSCRIPT remain small [68], indicating the similarity between them. Consequently, the entanglement entropies of the corresponding subsystems (see Fig.4(b)) as well as the average values of the observables such as n^idelimited-⟨⟩subscript^𝑛𝑖\langle\hat{n}_{i}\rangle⟨ over^ start_ARG italic_n end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT ⟩ and S^zidelimited-⟨⟩subscript^𝑆𝑧𝑖\langle\hat{S}_{zi}\rangle⟨ over^ start_ARG italic_S end_ARG start_POSTSUBSCRIPT italic_z italic_i end_POSTSUBSCRIPT ⟩ exhibit agreement with those of the thermal state [68], confirming the validity of subsystem thermalization [94, 95, 96] in open TCD over a range of λ𝜆\lambdaitalic_λ. Surprisingly, the steady state follows the subsystem thermalization over a large range of coupling strength even outside the chaotic regime, however, the effective temperature T𝑇Titalic_T increases with the degree of chaos, as shown in Fig.4(a).

For a deeper understanding of this scenario, we also study the statistics of eigenstates |ΨνketsubscriptΨ𝜈|{\Psi_{\nu}}\rangle| roman_Ψ start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ⟩ of the steady-state DM ρ^sssubscript^𝜌ss\hat{\rho}_{\rm ss}over^ start_ARG italic_ρ end_ARG start_POSTSUBSCRIPT roman_ss end_POSTSUBSCRIPT. We compute the structural entropy [97, 98] of |ΨνketsubscriptΨ𝜈|{\Psi_{\nu}}\rangle| roman_Ψ start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ⟩,

𝒮νstr=j|Ψνj|2ln(|Ψνj|2)+ln(j|Ψνj|4),superscriptsubscript𝒮𝜈strsubscript𝑗superscriptsuperscriptsubscriptΨ𝜈𝑗2superscriptsuperscriptsubscriptΨ𝜈𝑗2subscript𝑗superscriptsuperscriptsubscriptΨ𝜈𝑗4\displaystyle\mathcal{S}_{\nu}^{\rm str}=-\sum_{j}|\Psi_{\nu}^{j}|^{2}\ln(|% \Psi_{\nu}^{j}|^{2})+\ln\left(\sum_{j}|\Psi_{\nu}^{j}|^{4}\right),caligraphic_S start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_str end_POSTSUPERSCRIPT = - ∑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT | roman_Ψ start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_j end_POSTSUPERSCRIPT | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT roman_ln ( | roman_Ψ start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_j end_POSTSUPERSCRIPT | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) + roman_ln ( ∑ start_POSTSUBSCRIPT italic_j end_POSTSUBSCRIPT | roman_Ψ start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_j end_POSTSUPERSCRIPT | start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT ) , (6)

and study its distribution, which is sharply peaked around a value 𝒮str0.27superscript𝒮str0.27\mathcal{S}^{\rm str}\approx 0.27caligraphic_S start_POSTSUPERSCRIPT roman_str end_POSTSUPERSCRIPT ≈ 0.27 (see Fig.4(c)) corresponding to the Gaussian unitary ensemble (GUE) class of the random matrix theory (RMT) [3, 99]. Moreover, the elements η=|Ψνj|2Ndim𝜂superscriptsuperscriptsubscriptΨ𝜈𝑗2subscript𝑁dim\eta=|\Psi_{\nu}^{j}|^{2}N_{\rm dim}italic_η = | roman_Ψ start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_j end_POSTSUPERSCRIPT | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_N start_POSTSUBSCRIPT roman_dim end_POSTSUBSCRIPT of the eigenstate |ΨνketsubscriptΨ𝜈|{\Psi_{\nu}}\rangle| roman_Ψ start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ⟩ (with dimension Ndimsubscript𝑁dimN_{\rm dim}italic_N start_POSTSUBSCRIPT roman_dim end_POSTSUBSCRIPT), follows the GUE distribution P(η)=exp(η)P𝜂𝜂{\rm P}(\eta)=\exp(-\eta)roman_P ( italic_η ) = roman_exp ( - italic_η ) [1], as seen from Fig.4(d), indicating the chaotic nature of such states. The steady state of a random Liouvillian also follows the GUE distribution [22, 100], however in the present case, we do not find any signature of level repulsion in the Liouvillian spectrum, which on the contrary resembles to 2d-Poisson distribution, even in the chaotic regime [68].

Discussion: Our study unveils the emergence of a steady state with intriguing properties, especially the onset of chaos in an open atom-photon dimer system. The appropriate classical limit of this system enables us to explore the classical-quantum correspondence of dissipative chaos, which is absent in a generic quantum system. The rapid decay of the correlation function and survival probability, along with the growth of entropy, manifest the chaotic mixing of the emergent steady state, which serves as a more tangible signature of dissipative chaos compared to spectral statistics. Both chaos and quantum fluctuations result in the loss of coherence, leading to the formation of an incoherent photonic fluid, which can be probed experimentally. Remarkably, this steady state follows thermalization of the subsystems, revealing its connection with the random matrix.

In conclusion, this atom-photon dimer system exhibits intriguing nonequilibrium phenomena and sheds light on dissipative quantum chaos, with results readily testable in current cavity and circuit QED setups.

Acknowledgments: We thank Krishnendu Sengupta and Sudip Sinha for comments and fruitful discussions. D.M. acknowledges support from Prime Minister Research Fellowship (PMRF).

References

  • [1] F. Haake, Quantum Signatures of Chaos, Springer Science and Business Media (Springer, Berlin, Heidelberg, 2013), Vol. 54.
  • [2] O. Bohigas, M. J. Giannoni, and C. Schmit, Phys. Rev. Lett. 52, 1 (1984).
  • [3] F. M. Izrailev, Phys. Rep. 196, 299 (1990);
  • [4] L. D’Alessio, Y. Kafri, A. Polkovnikov, and M. Rigol, Adv. Phys. 65, 239 (2016).
  • [5] M. Ueda, Nat. Rev. Phys. 2, 669 (2020).
  • [6] F. Borgonovi, F. M. Izrailev, L. F. Santos, and V. G. Zelevinsky, Phys. Rep. 626, 1 (2016).
  • [7] M. Rigol, V. Dunjko, M. Olshanii, Nature 452, 854 (2008).
  • [8] J. M. Deutsch, Phys. Rev. A 43, 2046 (1991).
  • [9] M. Srednicki, Phys. Rev. E 50, 888 (1994).
  • [10] R. Grobe, F. Haake, and Hans-Jürgen Sommers, Phys. Rev. Lett. 61, 1899 (1988); R. Grobe and F. Haake, Phys. Rev. Lett. 62, 2893–2896 (1989).
  • [11] G. G. Carlo, G. Benenti, and D. L. Shepelyansky, Phys. Rev. Lett. 95, 164101 (2005).
  • [12] G. G. Carlo, G. Benenti, G. Casati, and D. L. Shepelyansky, Phys. Rev. Lett. 94, 164101 (2005).
  • [13] D. Dahan, G. Arwas, and E. Grosfeld, npj Quantum Inf 8, 14 (2022).
  • [14] D. Mondal, K. Sengupta, S. Sinha, arXiv:2310.12779 (2023).
  • [15] G. Vivek, D. Mondal, S. Chakraborty, S. Sinha, arXiv:2405.13809 (2024).
  • [16] F. Ferrari, L. Gravina, D. Eeltink, P. Scarlino, V. Savona, F. Minganti, arXiv.2305.15479 (2023).
  • [17] M. Žnidarič, T. Prosen, G. Benenti, G. Casati, and D. Rossini, Phys. Rev. E 81, 051135 (2010).
  • [18] I. Reichental, A. Klempner, Y. Kafri, and D. Podolsky, Phys. Rev. B 97, 134301 (2018).
  • [19] T. Shirai and T. Mori, Phys. Rev. E 101, 042116 (2020).
  • [20] F. Vicentini, A. Biella, N. Regnault, and C. Ciuti, Phys. Rev. Lett. 122, 250503 (2019).
  • [21] L. Sá, P. Ribeiro, and T. Prosen, Phys. Rev. X 10, 021019 (2020).
  • [22] L. Sá, P. Ribeiro, and T. Prosen, J. Phys. A: Math. Theor. 53 305303 (2020).
  • [23] S. Denisov, T. Laptyeva, W. Tarnowski, D. Chruściński, and K. Życzkowski, Phys. Rev. Lett. 123, 140403 (2019).
  • [24] M. Prasad, H. K. Yadalam, C. Aron, and M. Kulkarni, Phys. Rev. A 105, L050201 (2022).
  • [25] D. Villaseǹor, L. F. Santos, P. Barberis-Blostein, arXiv:2406.07616 (2024).
  • [26] C. Emary and T. Brandes, Phys. Rev. Lett. 90, 044101 (2003);
  • [27] A. Altland and F. Haake, Phys. Rev. Lett. 108, 073601 (2012); New J. Phys. 14, 073011 (2012).
  • [28] D. Villaseñor, S. Pilatowsky-Cameo, M. A. Bastarrachea-Magnani, S. Lerma-Hernández, L. F. Santos, and J. G. Hirsch, Entropy 25, 8 (2022).
  • [29] T. Dittrich, R. Graham, Europhys. Lett. 7, 287 (1988).
  • [30] P.S.Muraev, D.N.Maksimov, and A.R.Kolovsky, Entropy 2023, 25, 117.
  • [31] A. R. Kolovsky, Phys. Rev. E 106, 014209 (2022).
  • [32] S. Sinha, S. Ray and S. Sinha, J. Phys.: Condens. Matter 36 163001 (2024).
  • [33] F. Mivehvar, F. Piazza, T. Donner, H. Ritsch, Adv. Phys. 70, 1–153 (2021).
  • [34] H. Ritsch, P. Domokos, F. Brennecke, and T. Esslinger, Rev. Mod. Phys. 85, 553 (2013).
  • [35] M. Müller, S. Diehl, G. Pupillo, and P. Zoller, Adv. At. Mol. Opt. Phys. 61, 1 (2012).
  • [36] F. Damanet, E. Mascarenhas, D. Pekker, and A. J. Daley, Phys. Rev. Lett. 123, 180402 (2019).
  • [37] P.M. Harrington, E.J. Mueller, and K.W. Murch, Nat Rev Phys 4, 660–671 (2022).
  • [38] R. Lin, R. Rosa-Medina, F. Ferri, F. Finger, K. Kroeger, T. Donner, T. Esslinger, and R. Chitra, Phys. Rev. Lett. 128, 153601 (2022).
  • [39] H. Weimer, A. Kshetrimayum, and R. Orús, Rev. Mod. Phys. 93, 015008 (2021).
  • [40] F. Damanet, A. J. Daley, and J. Keeling, Phys. Rev. A 99, 033845 (2019).
  • [41] J. Klinder, H. Keßler, M. Wolke, L. Mathey, and A. Hemmerich, Proc. Natl. Acad. Sci. U.S.A. 112, 3290 (2015).
  • [42] S. Diehl, A. Tomadin, A. Micheli, R. Fazio, and P. Zoller, Phys. Rev. Lett. 105, 015702 (2010).
  • [43] H. J. Carmichael, Phys. Rev. X 5, 031028 (2015).
  • [44] K. C. Stitely, A. Giraldo, B. Krauskopf, and S. Parkins, Phys. Rev. Research 2, 033131 (2020).
  • [45] K. C. Stitely, A. Giraldo, B Krauskopf, and S. Parkins, Phys. Rev. Research 4, 023101 (2022).
  • [46] K. C. Stitely, S. J. Masson, A. Giraldo, B. Krauskopf, and S. Parkins, Phys. Rev. A 102, 063702 (2020).
  • [47] C. J. Zhu, L. L. **, Y. P. Yang, and G. S. Agarwal, Phys. Rev. Lett. 124, 073602 (2020).
  • [48] S. Ray, A. Vardi, and D. Cohen, Phys. Rev. Lett. 128, 130604 (2022).
  • [49] K. C. Stitely, F. Finger, R. Rosa-Medina, F. Ferri, T. Donner, T. Esslinger, S. Parkins, and B. Krauskopf, Phys. Rev. Lett. 131, 143604 (2023).
  • [50] J. Li, R. Fazio, and S. Chesi, New J. Phys. 24, 083039 (2022).
  • [51] W. Kopylov, M. Radonjić, T. Brandes, A. Balaž, and A. Pelster, Phys. Rev. A 92, 063832 (2015).
  • [52] F. Carollo and I. Lesanovsky, Phys. Rev. Lett. 126, 230601 (2021).
  • [53] J. M. Raimond, M. Brune, and S. Haroche, Rev. Mod. Phys. 73, 565 (2001).
  • [54] A. Blais, A. L. Grimsmo, S. M. Girvin, and A. Wallraff, Rev. Mod. Phys. 93, 025005 (2021).
  • [55] J. Léonard, A. Morales, P. Zupancic, T. Esslinger, and T. Donner, Nature (London) 543, 87 (2017).
  • [56] J. Léonard, A. Morales, P. Zupancic, T. Donner, and T. Esslinger, Science 358, 1415 (2017).
  • [57] J.A. Muniz, D. Barberena, R.J. Lewis-Swan, D. J. Young, J. R. K. Cline, A. M. Rey, Nature 580, 602–607 (2020).
  • [58] D.J. Young, A. Chu, E.Y.Song, D. Barberena, D. Wellnitz, Z. Niu, V. M. Schäfer, R. J. Lewis-Swan, A. M. Rey and J. K. Thompson, Nature 625, 679–684 (2024).
  • [59] F. Letscher, O. Thomas, T. Niederprüm, M. Fleischhauer, and H. Ott, Phys. Rev. X 7, 021020 (2017).
  • [60] R. M. Kroeze, Y. Guo, V. D. Vaidya, J. Keeling, and B. L. Lev, Phys. Rev. Lett. 121, 163601 (2018).
  • [61] M. Fitzpatrick, N. M. Sundaresan, A. C. Y. Li, J. Koch and A. A. Houck, Phys. Rev. X 7, 011016 (2017).
  • [62] M. Tavis and F. W. Cummings, Phys. Rev. 170, 379 (1968).
  • [63] H-P Eckle, Models of Quantum Matter: A First Course on Integrability and the Bethe Ansatz, Chap 12, 474 (Oxford University Press, (2021)).
  • [64] V. Gorini, A. Kossakowski, and E. C. G. Sudarshan, J. Math. Phys. (N.Y.) 17, 821–825 (1976).
  • [65] G. Lindblad, Commun. Math. Phys. 48, 119–130 (1976).
  • [66] H. P. Breuer and F. Petruccione, The Theory of Open Quantum Systems (Oxford University Press, Oxford, 2007).
  • [67] L. da Silva Souza, L. F. dos Prazeres, and F. Iemini, Phys. Rev. Lett. 130, 180401 (2023).
  • [68] See the Supplementary material for the details of the classical dynamical phases, quantum ergodicity, subsystem thermalization and statistics of the Liouvillian spectrum in the present model.
  • [69] S. H. Strogatz, Nonlinear Dynamics and Chaos (Westview Press, Boulder, CO, 2007).
  • [70] A. J. Lichtenberg and M. A. Lieberman, Regular and chaotic dynamics, (Springer-Verlag,1992).
  • [71] J. Chávez-Carlos, M. A. Bastarrachea-Magnani, S. Lerma-Hernández, and J. G. Hirsch, Phys. Rev. E 94, 022209 (2016).
  • [72] P. B. Blakie, A. S. Bradley, M. J. Davis, R. J. Ballagh, and C. W. Gardiner, Adv. Phys. 57, 363 (2008).
  • [73] A. Polkovnikov, Ann. Phys. 325, 1790 (2010).
  • [74] J. Schachenmayer, A. Pikovsky, and A. M. Rey, Phys. Rev. X 5, 011022 (2015).
  • [75] C. Gardiner and P. Zoller, Quantum Noise: A Handbook of Markovian and Non-Markovian Quantum Stochastic Methods with Applications to Quantum Optics (Springer Science, New York, 2004).
  • [76] J. M. Radcliffe, J. Phys. A: Gen.Phys., 4, 313 (1971).
  • [77] I. P. Cornfield, S. V. Fomin, and Y. G. Sinai, 1982 Ergodic Theory (Springer) .
  • [78] P. R. Halmos 2017 Lectures on Ergodic Theory (Dover).
  • [79] R. Frigg, J. Berkovitz, and F. Kronz, The Ergodic Hierarchy, in The Stanford Encyclopedia of Philosophy, (Fall 2020 Edition), edited by E. N. Zalta, https://plato.stanford.edu/archives/ fall2020/entries/ergodic-hierarchy/.
  • [80] E. Ott, Chaos in dynamical systems (1993, Cambridge University Press).
  • [81] D. Ruelle, Chaotic evolution and strange attractors (1989, Cambridge University Press).
  • [82] H. Carmichael, An Open Systems Approach to Quantum Optics. Springer-Verlag (1993).
  • [83] K. Mølmer, Y. Castin and J. Dalibard, J. Opt. Soc. Am. B 10, 524 (1993).
  • [84] M. A. Nielsen and I. L. Chuang, Quantum Computation and Quantum Information, (Cambridge University Press, Cambridge, England, 2000).
  • [85] H. Nha and H. J. Carmichael, Phys. Rev. A 71, 032336 (2005).
  • [86] F. Tonielli, R. Fazio, S. Diehl, and J. Marino, Phys. Rev. Lett. 122, 040604 (2019).
  • [87] For unitary dynamics of initial pure state ρ^(0)=|Ψ(0)Ψ(0)|^𝜌0ketΨ0braΨ0\hat{\rho}(0)=|{\Psi(0)}\rangle\langle{\Psi(0)}|over^ start_ARG italic_ρ end_ARG ( 0 ) = | roman_Ψ ( 0 ) ⟩ ⟨ roman_Ψ ( 0 ) |, the quantity f(t)=Tr(ρ^(t)ρ^(0))𝑓𝑡Tr^𝜌𝑡^𝜌0f(t)={\rm Tr}(\hat{\rho}(t)\hat{\rho}(0))italic_f ( italic_t ) = roman_Tr ( over^ start_ARG italic_ρ end_ARG ( italic_t ) over^ start_ARG italic_ρ end_ARG ( 0 ) ) reduces to the usual definition of survival probability f(t)=|Ψ(t)|Ψ(0)|2𝑓𝑡superscriptinner-productΨ𝑡Ψ02f(t)=|\langle\Psi(t)|\Psi(0)\rangle|^{2}italic_f ( italic_t ) = | ⟨ roman_Ψ ( italic_t ) | roman_Ψ ( 0 ) ⟩ | start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT. Additionally, it is related with the more general definition of overlap between two mixed density matrices (see Ref. [84, 85]).
  • [88] A. Piga, M. Lewenstein, and J. Q. Quach, Phys. Rev. E 99 032213 (2019).
  • [89] X. Wang, S. Ghose, B. C. Sanders, and B. Hu, Phys. Rev. E 70 016217 (2004).
  • [90] H. Fujisaki, T. Miyadera, and A. Tanaka, Phys. Rev. E 67, 066201 (2003).
  • [91] S. Ray, A. Ghosh, and S. Sinha, Phys. Rev. E 94, 032103 (2016).
  • [92] T. Geisel, Phys. Rev. A 41, 2989 (1990); Erratum Phys. Rev. A 42, 7491 (1990).
  • [93] S. Wolff, A. Sheikhan, C. Kollath, SciPost Phys. Core 3, 010 (2020)
  • [94] A. Dymarsky, N. Lashkari, and H. Liu, Phys. Rev. E 97, 012140 (2018);
  • [95] Z. Huang and Xiao-Kan Guo, Phys. Rev. E 109, 054120 (2024).
  • [96] J. R. Garrison and T. Grover, Phys. Rev. X 8, 021026 (2018).
  • [97] J. Pipek and I. Varga, Phys. Rev. A 46, 3148 (1992).
  • [98] P. Jacquod and I. Varga, Phys. Rev. Lett. 89, 134101 (2002).
  • [99] S. Ray, B. Mukherjee, S. Sinha, and K. Sengupta, Phys. Rev. A 96, 023607 (2017).
  • [100] T. Prosen and M. Žnidarič, Phys. Rev. Lett. 111, 124101 (2013).