Dissipative Polaron-Polaritons

• Dissipative polaron-polaritons are hybrid light–matter quasiparticles that combine many-body dressing with engineered dissipation, leading to finite-lifetime, non-Hermitian excitations.
• They emerge in diverse platforms such as semiconductor microcavities, molecular systems, and atomic arrays, where light–matter interactions are tuned by environmental losses.
• Key experimental signatures include branch coexistence, level attraction, and exceptional points, with interactions modulated via bath density and pump–probe techniques.

Searching arXiv for recent and foundational papers on dissipative polaron-polaritons. Dissipative polaron-polaritons are hybrid light–matter quasiparticles in which a polaritonic excitation is additionally dressed by a bosonic environment and evolves under loss, drive, or both. In current usage, the term covers several related nonequilibrium settings: Bose polaron-polaritons formed by impurity exciton-polaritons in a coherently driven polariton bath, dark-state polaron polaritons in superfluid optical media, vibrationally dressed molecular polaritons with Brownian dissipation, and polariton–phonon hybrids in subwavelength arrays of trapped atoms (Tan et al., 2022, Vashisht et al., 2021, Kansanen et al., 2020, Nielsen et al., 28 Jan 2026). The unifying feature is a complex, frequency-dependent dressing self-energy acting on the matter component, together with radiative, vibrational, cavity, or explicitly dissipative light–matter loss channels. This combination produces finite-lifetime quasiparticles with phenomena that are absent in equilibrium polaron physics, including branch coexistence, sign-tunable interactions, anomalous dispersion, negative drag, sideband-enabled transport breakdown, and exceptional points (Vega-Carmona et al., 1 Jul 2026).

1. Conceptual scope and physical realizations

A recurring realization is the cavity-coupled semiconductor platform in which a dilute impurity branch propagates in a dense polariton bath. In monolayer MoSe$_2$ embedded in an open fiber cavity, a resonant $\sigma^+$ pump creates a coherent bath of lower polaritons in one valley, while a weak $\sigma^-$ probe injects impurity lower polaritons in the opposite valley. Because opposite chiralities allow biexciton formation, the impurity–bath interaction is resonantly enhanced through a biexciton Feshbach resonance, leading to attractive and repulsive Bose polarons and, after hybridization with the cavity mode, to polaron-polaritonic branches (Tan et al., 2022).

A second realization is the driven quantum fluid of light formed by lower-branch exciton-polaritons in a coherently pumped semiconductor microcavity. There, an impurity of finite mass is embedded in a lossy polariton fluid whose Bogoliubov excitations have complex eigenfrequencies. The impurity dressed by these collective excitations is a Bose polaron-polariton in a genuinely nonequilibrium environment (Vashisht et al., 2021).

A third realization appears in superfluid optical media based on electromagnetically induced transparency. In a weakly interacting atomic Bose–Einstein condensate, a propagating photon hybridizes with a stable impurity excitation and the impurity becomes dressed by Bogoliubov phonons. The resulting quasiparticle is a dark-state polaron polariton, whose dynamics are determined jointly by Fröhlich dressing and optical control (2002.01435).

Molecular and vibronic settings extend the concept further. In strongly coupled cavity–exciton systems with molecular vibrations, the exciton is dressed by a dissipative vibronic cloud modeled as a Brownian oscillator; the polariton is then a photon hybridized with a dissipative vibronic polaron (Kansanen et al., 2020). A related minimal model is the lossy, periodically driven anharmonic Jaynes–Cummings polariton, in which Kerr anharmonicity and dissipation reshape the stationary Floquet state and the thermodynamics of a polaron-like light–matter excitation (Ochoa, 2024).

Subwavelength atomic arrays provide a distinct implementation. There, optomechanical forces couple collective atomic excitations to lattice phonons, producing polaritons dressed by phonons. The dressing renormalizes dispersion and linewidth off resonance and opens resonant phonon-assisted dissipation channels for dark states (Nielsen et al., 28 Jan 2026). This suggests that “dissipative polaron-polariton” is best understood as a class of non-Hermitian dressed polaritons rather than a single microscopic object.

2. Microscopic structure: self-energy, hybridization, and effective Hamiltonians

The central theoretical object across platforms is a dressed excitonic or impurity propagator whose self-energy is generated by coupling to a bosonic environment. In the monolayer semiconductor case, the probe transmission is determined by the excitonic component of a $2\times2$ exciton–photon Green’s function,

$$\mathbf{G}(E)= \begin{pmatrix} \omega_X(0)-E+\Sigma(E) & \Omega/2\ \Omega/2 & \omega_C(0)-i\gamma_C-E \end{pmatrix}^{-1},$$

with transmission $\mathcal{T}(k,E)=|\mathbf{G}_{22}(k,E)|^2$. The bath-induced self-energy is

$$\Sigma(E)=\cos^2\theta_{\mathbf 0}\, \frac{n_b}{\,g^{-1}+\frac{1}{V}\sum_k \frac{\cos^2\theta_k}{\xi_k}\,},$$

so that the dressing scales linearly with bath density and is resonantly enhanced by the biexciton pole (Tan et al., 2022).

In the non-Hermitian many-body formulation of dissipative coupling polaron-polaritons, the effective exciton–photon problem takes the form

$$H_{\mathrm{eff}}(\mathbf{k},\omega)= \begin{pmatrix} E_x(\mathbf{k})-i\gamma_x+\Sigma_{\uparrow}(\mathbf{k},\omega) & \Omega_{\mathrm{Re}}-i\Omega_{\mathrm{Im}}\ \Omega_{\mathrm{Re}}-i\Omega_{\mathrm{Im}} & E_c(\mathbf{k})-i\gamma_c \end{pmatrix},$$

where $$\Sigma_{\uparrow}(\mathbf{k},\omega)=n_{\downarrow}\,\mathcal{T}(k)$$ is the biexciton-mediated impurity self-energy and $\Omega_{\mathrm{Im}}$ is an anti-Hermitian off-diagonal coupling produced by a common lossy channel and inelastic processes (Vega-Carmona et al., 1 Jul 2026). The matter dressing is therefore not merely a diagonal energy correction; it also renormalizes effective detuning and effective relative decay.

In molecular systems, the same role is played by the Brownian $\sigma^+$0 kernel. The cavity response is

$\sigma^+$1

with

$\sigma^+$2

The corresponding excitonic susceptibility is non-Lorentzian and temperature dependent because the vibrational environment produces a frequency-dependent self-energy rather than a constant linewidth (Kansanen et al., 2020).

In trapped-atom arrays, the same structure emerges in polaron form. The retarded Green’s function is

$\sigma^+$3

with

$\sigma^+$4

Here the self-energy encodes both off-resonant dressing and resonant phonon-assisted decay (Nielsen et al., 28 Jan 2026).

Across these cases, a common description is therefore a light–matter hybrid whose matter constituent acquires a complex, often non-Markovian self-energy from many-body dressing. This shared structure is the basis for the broader category of dissipative polaron-polaritons.

3. Nonequilibrium drive and dissipative channels

The field is defined as much by nonequilibrium conditions as by polaron formation. In the MoSe$\sigma^+$5 cavity experiment, the bath is generated by a narrowband resonant pump with $\sigma^+$6 meV FWHM and $\sigma^+$7 ps duration, while the impurity is injected by a broadband $\sigma^+$8 meV, $\sigma^+$9 fs probe. The empty cavity finesse is $\sigma^-$0, the photon decay rate is $\sigma^-$1 meV, and the corresponding lifetime scale $\sigma^-$2 ps is comparable to the pump duration and observed temporal overlap window. Dissipation is incorporated phenomenologically through imaginary parts in the single-particle energies, and the biexciton channel carries an additional linewidth $\sigma^-$3 meV (Tan et al., 2022).

In the coherently pumped quantum fluid of light, the steady state is fixed by a balance of continuous-wave drive and cavity photon leakage. The Bogoliubov spectrum can be gapless, gapped, or diffusive depending on pump detuning $\sigma^-$4, and for $\sigma^-$5 it can acquire purely imaginary eigenenergies over finite momentum ranges. The impurity therefore moves in a fluid whose linear response is intrinsically non-Hermitian rather than in an equilibrium superfluid (Vashisht et al., 2021).

The molecular Brownian setting illustrates a different dissipative mechanism. Vibrations are described by a damped Brownian oscillator characterized by $\sigma^-$6, $\sigma^-$7, $\sigma^-$8, and $\sigma^-$9, so the matter linewidth is not constant but encoded in a frequency-dependent kernel $2\times2$0. Observable polariton spectra inherit asymmetry, thermal redistribution, and non-Lorentzian tails directly from this dissipation model (Kansanen et al., 2020).

In trapped atomic arrays, motion-induced modulation of photon-mediated dipole–dipole couplings opens two qualitatively distinct loss channels: zero-point-motion corrections of order $2\times2$1 and resonant phonon-assisted scattering between collective modes. The second mechanism is especially important for subradiant states, whose decay can be strongly enhanced when the phonon resonance intersects a large polaritonic density of states (Nielsen et al., 28 Jan 2026).

A common misconception is that dissipation acts only as homogeneous broadening. The cited systems show a broader role: dissipation determines whether dressing is transient or stationary, whether spectra remain Lorentzian, whether collective modes become diffusive, and whether new scattering channels or anti-Hermitian couplings appear.

4. Spectral organization: branches, asymmetry, and exceptional points

Near a biexciton Feshbach resonance, impurity dressing can split the spectral response into multiple branches. In the cavity-coupled monolayer semiconductor, time-resolved spectra show a higher-energy repulsive polaron branch and a lower-energy attractive polaron branch. Their relative weights depend on Feshbach detuning: below the biexciton energy the attractive branch dominates, while above it the repulsive branch dominates. Increasing bath density $2\times2$2 enhances the attractive–repulsive splitting, and the experiment shows an avoided-crossing-like structure during pump–probe overlap (Tan et al., 2022).

When dissipative coupling is added to many-body dressing, the spectral organization becomes explicitly non-Hermitian. Attractive and repulsive matter polarons hybridize with the cavity photon to yield three polaron-polariton branches, denoted lower, middle, and upper. Their dispersions can exhibit level attraction, curvature inversion, and negative effective mass. Exceptional points occur when the dressed discriminant vanishes,

$2\times2$3

so eigenvalues and eigenvectors coalesce. Depending on the sign and magnitude of $2\times2$4, exceptional points can occur between lower–middle or middle–upper branches, and at sufficiently large $2\times2$5 and $2\times2$6 two sets of exceptional points can coexist (Vega-Carmona et al., 1 Jul 2026).

Molecular dissipative polaron-polaritons display a different but related spectral anomaly. Because the excitonic self-energy is frequency dependent, the frequently used coupled oscillator model fails to describe the elastic response. The minimum polariton splitting occurs at a shifted detuning $2\times2$7, the lower- and upper-polariton peak heights are unequal at that detuning, and the peaks remain asymmetric even at nominal zero detuning $2\times2$8. These effects arise from the non-Lorentzian Brownian kernel rather than from a simple constant-width excitonic loss (Kansanen et al., 2020).

This spectral phenomenology shows that “polaron-polariton branch” need not mean a simple avoided crossing of two damped oscillators. In dissipative settings it can mean coexistence of attractive and repulsive polarons, three-branch non-Hermitian hybridization, or line-shape asymmetry driven by a structured environment.

5. Interactions, transport, drag, and directly measured responses

One of the most developed aspects of the subject is the direct extraction of effective interactions. In the MoSe$2\times2$9 cavity platform, the impurity–bath interaction is parameterized through

$$\mathbf{G}(E)= \begin{pmatrix} \omega_X(0)-E+\Sigma(E) & \Omega/2\ \Omega/2 & \omega_C(0)-i\gamma_C-E \end{pmatrix}^{-1},$$0

with measured values $$\mathbf{G}(E)= \begin{pmatrix} \omega_X(0)-E+\Sigma(E) & \Omega/2\ \Omega/2 & \omega_C(0)-i\gamma_C-E \end{pmatrix}^{-1},$$1 to $$\mathbf{G}(E)= \begin{pmatrix} \omega_X(0)-E+\Sigma(E) & \Omega/2\ \Omega/2 & \omega_C(0)-i\gamma_C-E \end{pmatrix}^{-1},$$2 for the repulsive branch and $$\mathbf{G}(E)= \begin{pmatrix} \omega_X(0)-E+\Sigma(E) & \Omega/2\ \Omega/2 & \omega_C(0)-i\gamma_C-E \end{pmatrix}^{-1},$$3 to $$\mathbf{G}(E)= \begin{pmatrix} \omega_X(0)-E+\Sigma(E) & \Omega/2\ \Omega/2 & \omega_C(0)-i\gamma_C-E \end{pmatrix}^{-1},$$4 for the attractive branch. The same-valley baseline without bath is $$\mathbf{G}(E)= \begin{pmatrix} \omega_X(0)-E+\Sigma(E) & \Omega/2\ \Omega/2 & \omega_C(0)-i\gamma_C-E \end{pmatrix}^{-1},$$5. Polaron–polaron interactions are extracted from

$$\mathbf{G}(E)= \begin{pmatrix} \omega_X(0)-E+\Sigma(E) & \Omega/2\ \Omega/2 & \omega_C(0)-i\gamma_C-E \end{pmatrix}^{-1},$$6

and at $$\mathbf{G}(E)= \begin{pmatrix} \omega_X(0)-E+\Sigma(E) & \Omega/2\ \Omega/2 & \omega_C(0)-i\gamma_C-E \end{pmatrix}^{-1},$$7 yield $$\mathbf{G}(E)= \begin{pmatrix} \omega_X(0)-E+\Sigma(E) & \Omega/2\ \Omega/2 & \omega_C(0)-i\gamma_C-E \end{pmatrix}^{-1},$$8 for the attractive branch and $$\mathbf{G}(E)= \begin{pmatrix} \omega_X(0)-E+\Sigma(E) & \Omega/2\ \Omega/2 & \omega_C(0)-i\gamma_C-E \end{pmatrix}^{-1},$$9 for the repulsive branch. The repulsive branch becomes less repulsive and crosses to attractive at the highest bath densities achieved experimentally. To the authors’ knowledge, these were the first direct measurements of Bose polaron–polaron interactions in any physical system (Tan et al., 2022).

In driven quantum fluids of light, transport observables are organized by effective mass and drag rather than by static branch shifts. The self-consistent dressing momentum $\mathcal{T}(k,E)=|\mathbf{G}_{22}(k,E)|^2$0 defines an effective mass $\mathcal{T}(k,E)=|\mathbf{G}_{22}(k,E)|^2$1, and in the diffusive regime $\mathcal{T}(k,E)=|\mathbf{G}_{22}(k,E)|^2$2 the impurity can experience negative drag and accelerate against the flow to a finite terminal momentum. This has no equilibrium counterpart and reflects the complex excitation spectrum of the driven-dissipative polariton fluid (Vashisht et al., 2021).

In superfluid optical media, the most striking transport prediction is optical stabilization above Landau’s critical velocity. For negative one-photon detuning and $\mathcal{T}(k,E)=|\mathbf{G}_{22}(k,E)|^2$3, the dominant phonon-emission channel acquires an extra light-induced energy cost, yielding a modified critical velocity

$\mathcal{T}(k,E)=|\mathbf{G}_{22}(k,E)|^2$4

As a result, the relevant damping rate $\mathcal{T}(k,E)=|\mathbf{G}_{22}(k,E)|^2$5 can remain negligible even when the group velocity $\mathcal{T}(k,E)=|\mathbf{G}_{22}(k,E)|^2$6 and impurity recoil velocity exceed the sound speed by large factors (2002.01435).

In subwavelength atomic arrays, motion controls both transport and optical functionality. The transport length is $\mathcal{T}(k,E)=|\mathbf{G}_{22}(k,E)|^2$7, and transport of dark excitations remains robust over a wide range of trap frequencies except when resonant phonon scattering reaches high-density-of-states regions. Motion also reduces mirror reflectivity, but increasing $\mathcal{T}(k,E)=|\mathbf{G}_{22}(k,E)|^2$8 and reducing $\mathcal{T}(k,E)=|\mathbf{G}_{22}(k,E)|^2$9 can restore on-resonance reflectance above $$\Sigma(E)=\cos^2\theta_{\mathbf 0}\, \frac{n_b}{\,g^{-1}+\frac{1}{V}\sum_k \frac{\cos^2\theta_k}{\xi_k}\,},$$0, with $$\Sigma(E)=\cos^2\theta_{\mathbf 0}\, \frac{n_b}{\,g^{-1}+\frac{1}{V}\sum_k \frac{\cos^2\theta_k}{\xi_k}\,},$$1 reported for $$\Sigma(E)=\cos^2\theta_{\mathbf 0}\, \frac{n_b}{\,g^{-1}+\frac{1}{V}\sum_k \frac{\cos^2\theta_k}{\xi_k}\,},$$2 in the normal-incidence geometry discussed in the paper (Nielsen et al., 28 Jan 2026).

6. Modeling assumptions, limitations, and open questions

The main theoretical descriptions rely on controlled but restrictive approximations. In the semiconductor Bose-polaron setting, the interaction is treated as contact-like and renormalized to the biexciton bound state, finite-range effects are neglected, the pump bath is treated as a Fock state at $$\Sigma(E)=\cos^2\theta_{\mathbf 0}\, \frac{n_b}{\,g^{-1}+\frac{1}{V}\sum_k \frac{\cos^2\theta_k}{\xi_k}\,},$$3, and the two-polaron interaction is extracted from an uncorrelated product ansatz. Quantitative differences between experiment and theory are attributed to density calibration, neglected finite-range effects, neglect of explicit inter-polaron correlations, and a modeling procedure in which impurity density is varied through system volume scaling; the cited work also neglects a finite probe-exciton lifetime to reduce the number of fit parameters (Tan et al., 2022).

In the exceptional-point theory, the dissipative coupling $$\Sigma(E)=\cos^2\theta_{\mathbf 0}\, \frac{n_b}{\,g^{-1}+\frac{1}{V}\sum_k \frac{\cos^2\theta_k}{\xi_k}\,},$$4 is treated phenomenologically as an external common reservoir, and open questions include its microscopic derivation, possible momentum or frequency dependence, the role of finite temperature and higher density, and the fate of the predicted non-Hermitian topology in disordered or strongly driven settings (Vega-Carmona et al., 1 Jul 2026).

The molecular Brownian treatment establishes another important limitation: a constant-linewidth coupled-oscillator model can be qualitatively wrong when molecular dissipation is non-Lorentzian. Symmetric Lorentzian fits at nominal zero detuning, temperature-independent minimum splitting, and equal peak heights at the minimum are therefore not generic signatures of strong coupling in vibronically dressed systems (Kansanen et al., 2020).

The atomic-array and periodically driven anharmonic models add further caveats. In the array problem, the Chevy-type truncation can break down near strong resonances where $$\Sigma(E)=\cos^2\theta_{\mathbf 0}\, \frac{n_b}{\,g^{-1}+\frac{1}{V}\sum_k \frac{\cos^2\theta_k}{\xi_k}\,},$$5 becomes very large or very small, and the present treatment assumes magic trapping and weak heating. In the anharmonic Floquet polariton problem, the master equation is time-local and valid only in the weak-driving, weak-damping regime; there, increasing anharmonicity suppresses bosonic energy storage, lowers maximum power, and enhances interference between exciton and boson drive pathways (Nielsen et al., 28 Jan 2026, Ochoa, 2024).

Taken together, these studies support an objective but expansive view of dissipative polaron-polaritons. They are not merely broadened polarons or lossy polaritons. Rather, they are quasiparticles in which many-body dressing and dissipation are coequal ingredients: the bath can reverse interaction signs, redistribute spectral weight, reshape transport thresholds, destabilize or protect superfluid flow, and generate genuinely non-Hermitian structures such as level attraction and exceptional points. Open problems therefore center not only on stronger correlations and improved quantitative modeling, but also on how to exploit dissipation itself as a tuning parameter for many-body and spectral topology in light–matter quantum fluids.