Unruh Prethermalization in Accelerated Systems

• Unruh prethermalization is a nonequilibrium regime where accelerated many-atom systems temporarily adopt a generalized Gibbs state protected by emergent conservation laws.
• This phenomenon is marked by a clear separation of timescales, with a rapid prethermal plateau preceding the slower approach to the standard Unruh Gibbs state.
• It bridges rigorous detector thermalization theories with experimental analogues in ultracold atoms and condensates, offering actionable insights into open-system dynamics.

Searching arXiv for the cited papers on Unruh prethermalization and related detector thermalization. Unruh prethermalization denotes, in the most specific current usage, a nonequilibrium regime in which a uniformly accelerating many-atom system relaxes first to a long-lived prethermal generalized Gibbs state and only later to the ordinary Gibbs state associated with Unruh thermalization (Saha et al., 6 Sep 2025). In a broader and looser usage across adjacent literature, the phrase also refers to acceleration-induced thermal-like response before exact equilibration, to asymptotic detector thermal plateaus that emerge only in the long-time limit, and to analog settings in which KMS-like or thermal-looking correlations arise without global equilibrium (Perche, 2021). The topic therefore lies at the intersection of the Unruh effect, open-system thermalization, metastable nonequilibrium states, and analogue gravity.

1. Definition and relation to the standard Unruh effect

The standard Unruh effect states that a uniformly accelerated detector probing the Minkowski vacuum responds as if immersed in a thermal environment with inverse temperature

$\beta_U=\frac{2\pi}{a},$

equivalently

$T_U=\frac{\hbar a}{2\pi c k_B},$

or, in natural units, $T_U=a/(2\pi)$ (Perche, 2021). In the detector language, this statement is usually formulated through KMS analyticity, detailed balance, or asymptotic relaxation of the detector’s reduced state.

This reference problem matters because “prethermalization” is not part of the minimal Unruh statement. For general smeared particle detectors coupled to arbitrary QFT operators, the asymptotic result is thermalization to the KMS temperature, provided the detector is rigid, sufficiently localized, and switched on adiabatically for a long time (Perche, 2021). In a uniformly accelerated Minkowski setting, the detector size must satisfy

$a^2 D_a^* D^a \ll 1,$

so that the detector is much smaller than the acceleration length scale (Perche, 2021). Nonperturbative cavity analyses similarly show that, with sufficiently smooth switching, an accelerated detector thermalizes to an Unruh-like temperature regardless of periodic, Dirichlet, or Neumann boundary conditions, and regardless of whether the coupling is to the field amplitude or field momentum (Brenna et al., 2013).

A first conceptual boundary therefore follows. Ordinary Unruh thermalization concerns the long-time Gibbs or detailed-balance behavior of accelerated probes. Unruh prethermalization, by contrast, becomes relevant when there is an intermediate quasi-stationary regime, a sector-selective thermal response, or a thermal-like plateau protected by approximate conservation laws rather than by full equilibration.

2. Emergence of a prethermal generalized Gibbs state in accelerating many-atom systems

The sharpest formulation of Unruh prethermalization appears in the uniformly accelerating many-atom problem of non-interacting two-level atoms weakly coupled to a free massless scalar field in the vacuum (Saha et al., 6 Sep 2025). For two atoms, the trajectories are hyperbolic,

$$t_a=\frac{1}{\alpha}\sinh(\alpha \tau),\qquad X_a=\frac{1}{\alpha}\cosh(\alpha \tau),$$

with proper separation $L$. Under the Born-Markov and weak-coupling approximations, the reduced density matrix satisfies a Lindblad master equation,

$$\frac{d\rho_s(\tau)}{d\tau} = -i[\mathcal{H}_{\rm lamb},\rho_s(\tau)] + \mathcal{L}(\rho_s(\tau)).$$

The control parameter for collective behavior is the bath-induced correlation factor $f_{ab}(\alpha)$,

$$f_{ab}(\alpha)= \frac{\sin\!\left(2\frac{\omega_\circ}{\alpha}\sinh^{-1}\!\left(\frac{\alpha L}{2}\right)\right)} {L\,\omega_\circ\sqrt{1+\alpha^2L^2/4}} \quad (a\neq b), \qquad f_{aa}=1.$$

Its limiting behavior organizes the dynamics: $L\ll 1/\alpha$ implies $T_U=\frac{\hbar a}{2\pi c k_B},$0, while $T_U=\frac{\hbar a}{2\pi c k_B},$1 implies $T_U=\frac{\hbar a}{2\pi c k_B},$2 (Saha et al., 6 Sep 2025). When $T_U=\frac{\hbar a}{2\pi c k_B},$3, the Liouvillian has only one zero eigenvalue, and the atoms relax to the ordinary Gibbs state at the Unruh temperature. When $T_U=\frac{\hbar a}{2\pi c k_B},$4, the Liouvillian acquires additional zero modes, the dynamics becomes nearly integrable, and the system first approaches a prethermal plateau.

In that regime, the relevant steady state is not the universal thermal Gibbs state but a generalized Gibbs state. For the two-atom problem, it can be written as

$T_U=\frac{\hbar a}{2\pi c k_B},$5

This state depends on initial data through additional conserved structure, which is precisely what distinguishes it from standard Unruh thermalization (Saha et al., 6 Sep 2025).

The emergent conservation law is explicit. At $T_U=\frac{\hbar a}{2\pi c k_B},$6, the dynamics preserves

$T_U=\frac{\hbar a}{2\pi c k_B},$7

This protected quantity, together with the extra zero modes of the Liouvillian, is the mechanism behind the long-lived prethermal generalized Gibbs state (Saha et al., 6 Sep 2025). The phenomenon is therefore not a small perturbation of single-detector thermality but a qualitatively different many-body open-system regime.

3. Dynamical signatures, timescales, and many-body structure

The defining signature of Unruh prethermalization is a separation of timescales. In the nearly integrable regime $T_U=\frac{\hbar a}{2\pi c k_B},$8, the system first reaches the prethermal plateau on the scale

$T_U=\frac{\hbar a}{2\pi c k_B},$9

and only later thermalizes on the longer scale

$T_U=a/(2\pi)$0

A simple estimate of the prethermal lifetime is therefore $T_U=a/(2\pi)$1, which becomes parametrically large as $T_U=a/(2\pi)$2 (Saha et al., 6 Sep 2025).

Entanglement provides a diagnostic of this lifetime. For two atoms, the concurrence is

$T_U=a/(2\pi)$3

and the measure of entanglement is a good estimation of the lifetime of the prethermal state (Saha et al., 6 Sep 2025). In the same regime, entanglement is generated and persists much longer than in the ordinary thermalizing case.

A particularly transparent protected state is the singlet Bell state

$T_U=a/(2\pi)$4

which is a dark state for $T_U=a/(2\pi)$5 and does not evolve (Saha et al., 6 Sep 2025). This identifies a decoherence-free subspace associated with the emergent symmetry.

For many atoms, the collective structure becomes more pronounced. In the extreme collective limit $T_U=a/(2\pi)$6, the dynamics is organized in total-spin sectors with collective operator $T_U=a/(2\pi)$7, and there are

$T_U=a/(2\pi)$8

conserved pairwise quantities such as $T_U=a/(2\pi)$9 (Saha et al., 6 Sep 2025). The system then behaves like a nearly integrable model, and the prethermal generalized Gibbs regime has entropy essentially independent of $a^2 D_a^* D^a \ll 1,$0, in sharp contrast with the extensive entropy of independent thermalization.

The transient approach to the plateau is also nontrivial. In the $a^2 D_a^* D^a \ll 1,$1 regime, the collective radiation intensity

$a^2 D_a^* D^a \ll 1,$2

shows a Dicke superradiance-type radiation burst before prethermalization, with peak intensity scaling as $a^2 D_a^* D^a \ll 1,$3 and burst decay time roughly $a^2 D_a^* D^a \ll 1,$4 (Saha et al., 6 Sep 2025). Ordinary Unruh thermalization, by contrast, exhibits only mono-exponential decay.

4. Detector-level approximate, non-KMS, and partial thermality

Outside the strict many-atom usage, the phrase “Unruh prethermalization” has been used more loosely for situations in which detector response is thermal-like without constituting full finite-time thermal equilibrium. A central result is that KMS implies asymptotic thermalization, but the asymptotic excitation-to-deexcitation ratio by itself does not prove that the detector’s reduced density matrix is exactly Gibbs-like at finite time (Perche, 2021). The long-time thermal plateau is therefore universal, while finite-time behavior retains dependence on switching, smearing, and other non-asymptotic details.

Finite-time effective temperatures make this explicit. For an accelerated Unruh-DeWitt detector in a stationary/KMS field state, the excitation-to-deexcitation ratio defines an effective inverse EDR temperature

$a^2 D_a^* D^a \ll 1,$5

which can be nearly gap-independent before full equilibration (Garay et al., 2016). In this regime, the detector looks approximately thermal even though the interaction time is finite. The same analysis also reveals the Anti-Unruh effect: for accelerated trajectories, the detector can click less often as the KMS temperature increases, and this may persist even in the adiabatic infinite-time limit (Garay et al., 2016). This excludes a naive identification of “more acceleration” with “more detector excitation” in all regimes.

Thermal detector response can also survive without exact KMS analyticity. Uniformly accelerated detectors can satisfy thermal detailed balance and even have a Planckian response identical to the standard case when the Wightman function fails to satisfy KMS; the exact criterion is instead encoded in analytic conditions on the pole structure of the deformed Wightman function (Carballo-Rubio et al., 2018). This separates observable thermality from strict equilibrium in the algebraic sense.

At the same time, thermality need not extend to all detector sectors. When a uniformly accelerated two-level atom is given a dynamical center-of-mass degree of freedom, the internal energy thermalizes with the Unruh bath, but the center-of-mass does not: recoil breaks the fluctuation-dissipation structure for the motional sector, even though it remains valid for the internal sector (Stargen et al., 2 Sep 2025). Conversely, genuinely extended systems can thermalize. A pair of uniformly accelerated, weakly coupled spins relaxes to

$a^2 D_a^* D^a \ll 1,$6

with $a^2 D_a^* D^a \ll 1,$7, showing that the Minkowski vacuum acts as a legitimate thermal reservoir for an accelerated extended system even though its nonlocal correlators are not identical to those of an inertial thermal bath (Lima et al., 2018).

Taken together, these results suggest that detector thermality, KMS equilibrium, Gibbs reduction, and prethermal plateaus are distinct notions. They coincide in some regimes but separate in others.

5. Analog, simulator, and spacelike realizations

Unruh prethermalization has strong connections to analogue platforms, especially where acceleration-induced thermal signatures appear inside explicitly nonequilibrium states. A notable example is the supersonic splitting of a one-dimensional condensate. There, the low-energy Luttinger-liquid dynamics is Lorentz invariant, and a supersonic moving splitting front maps to a sudden quench in a boosted frame. The resulting state is not an ordinary thermal state but a chiral prethermal state with right- and left-moving sectors at different Doppler-shifted temperatures,

$a^2 D_a^* D^a \ll 1,$8

with

$a^2 D_a^* D^a \ll 1,$9

and $$t_a=\frac{1}{\alpha}\sinh(\alpha \tau),\qquad X_a=\frac{1}{\alpha}\cosh(\alpha \tau),$$0 (Agarwal et al., 2014). By accelerating the release point instead of moving it at constant velocity, the same setup realizes a spacelike analogue of the Unruh effect, producing thermal-looking correlations with

$$t_a=\frac{1}{\alpha}\sinh(\alpha \tau),\qquad X_a=\frac{1}{\alpha}\cosh(\alpha \tau),$$1

along an accelerated spacelike trajectory (Agarwal et al., 2014). This is not Unruh prethermalization in the many-atom detector sense, but it directly connects acceleration-induced thermality to prethermal many-body dynamics.

A second analogue direction is the quench from a Minkowski to a Rindler Hamiltonian in interacting ultracold fermions on an optical lattice. The post-quench Wightman response is thermal-Unruh-like in the low-energy relativistic regime, and BCS pairing leads to an inversion of statistics for low-lying excitations (Kosior et al., 2018). That work uses a sudden quench and a thermal-like response function, but it does not identify a long-lived prethermal plateau in the many-body sense.

A third development is the relativistic-fluid counterpart of the Unruh effect. In first-order hydrodynamics and its modern extensions, frame changes can be interpreted as hyperbolic rotations in a thermodynamic state space, and the proper-time variation of the associated rapidity defines a thermodynamic acceleration

$$t_a=\frac{1}{\alpha}\sinh(\alpha \tau),\qquad X_a=\frac{1}{\alpha}\cosh(\alpha \tau),$$2

A comoving probe then sees a thermal spectrum with

$$t_a=\frac{1}{\alpha}\sinh(\alpha \tau),\qquad X_a=\frac{1}{\alpha}\cosh(\alpha \tau),$$3

which is explicitly described as a transient KMS/thermal response during out-of-equilibrium evolution rather than as the material temperature of the fluid (Erkul, 15 Nov 2025). This is one of the clearest analogues of Unruh prethermalization outside detector QFT.

6. Conceptual boundaries and common misconceptions

A recurring misconception is that any thermal detector response under acceleration already constitutes Unruh prethermalization. The literature does not support that identification. The strongest current use of the term is the many-atom regime in which emergent conserved quantities protect a prethermal generalized Gibbs state before eventual Gibbs thermalization (Saha et al., 6 Sep 2025). By contrast, many single-detector studies establish asymptotic thermalization, not a distinct prethermal plateau [(Perche, 2021); (Brenna et al., 2013)].

A second misconception is that strict KMS equilibrium is necessary for Unruh thermality. It is sufficient but not necessary: detector detailed balance and Planckian response can persist even when KMS fails (Carballo-Rubio et al., 2018). Conversely, a KMS state does not imply exact finite-time Gibbs reduction of every observable sector (Perche, 2021, Stargen et al., 2 Sep 2025).

A third misconception is that acceleration-induced temperature is always a unique scalar attached to the state. The split-condensate example shows that probe trajectory can interpolate between chiral Doppler-shifted temperatures, making it impossible to assign a unique naive Lorentz-transformed temperature to the moving prethermal state (Agarwal et al., 2014). This point is conceptually close to Unruh prethermalization because it emphasizes observer- and probe-dependent thermality.

The boundary between dynamical prethermalization and structural precursor is also important. In a Mellin-transform representation of the $$t_a=\frac{1}{\alpha}\sinh(\alpha \tau),\qquad X_a=\frac{1}{\alpha}\cosh(\alpha \tau),$$4-dimensional massless field, the Unruh field emerges only in the singular limit $$t_a=\frac{1}{\alpha}\sinh(\alpha \tau),\qquad X_a=\frac{1}{\alpha}\cosh(\alpha \tau),$$5; the resulting thermal spectrum is obtained by quantization of that singular representation, not by real-time relaxation (Aref'eva et al., 2013). Likewise, a source-driven interpretation of the Unruh response treats bilocal entangled excitations as localized microstates that populate the thermal ensemble after wedge restriction and coarse-graining (Player, 20 Sep 2025). These are suggestive of prethermal structure, but they are not metastable dynamical plateaus.

Finally, non-uniform acceleration further complicates the picture. For a uniformly rotating detector in $$t_a=\frac{1}{\alpha}\sinh(\alpha \tau),\qquad X_a=\frac{1}{\alpha}\cosh(\alpha \tau),$$6 dimensions, the long-time effective temperature tends to zero linearly in the small-gap limit for ordinary nonnegative switchings, but a positive limiting temperature can be recovered in a controlled long-time-small-gap double limit if the switching changes sign and satisfies small frequency suppression (Parry et al., 27 Aug 2025). This result reinforces a general lesson: approximate thermality can depend on interaction-time scaling and probe design even when exact thermality is absent.

7. Significance and current understanding

The present state of the subject suggests that “Unruh prethermalization” is best understood as a family resemblance term rather than a single theorem. In its strict sense, it names a many-body acceleration-induced prethermal plateau protected by emergent conservation laws, with Liouvillian near-zero modes, generalized Gibbs structure, long lifetime scaling like $$t_a=\frac{1}{\alpha}\sinh(\alpha \tau),\qquad X_a=\frac{1}{\alpha}\cosh(\alpha \tau),$$7, and a superradiant burst preceding equilibration (Saha et al., 6 Sep 2025). In a broader sense, it encompasses acceleration-induced thermal-like regimes that are asymptotic, approximate, sector-dependent, or analogue rather than globally thermal.

This broader understanding is technically useful because it aligns several strands of research that would otherwise appear disconnected: rigorous detector thermalization in the long-time limit (Perche, 2021), nonperturbative universality of Unruh equilibration (Brenna et al., 2013), thermal response without exact KMS (Carballo-Rubio et al., 2018), sector-selective athermality due to recoil (Stargen et al., 2 Sep 2025), and experimentally accessible analogue realizations in condensates, optical lattices, and relativistic fluids [(Agarwal et al., 2014); (Kosior et al., 2018); (Erkul, 15 Nov 2025)].

A plausible implication is that future work will continue to sharpen the distinction between four notions that are often conflated: KMS structure of the field state, detailed balance of a detector response, Gibbs form of the reduced state, and metastable prethermalization of a many-body accelerated system. Unruh prethermalization is precisely the regime in which those notions cease to be equivalent.