Trion Polaritons: Charged Exciton Hybrids

• Trion polaritons are hybrid light–matter quasiparticles that emerge when an optical mode couples strongly to charged-exciton resonances in doped semiconductors.
• They exhibit a multibranch spectrum with clear avoided crossings, where features are tunable via carrier density, detuning, and the photonic environment.
• Their strong nonlinearity arises from phase-space filling, oscillator strength saturation, and Coulomb-mediated trion–trion interactions, providing opportunities for advanced quantum photonics.

Trion polaritons are hybrid light–matter quasiparticles formed when an optical mode strongly couples to a charged-exciton resonance in a doped semiconductor. In the minimal low-density picture, the matter constituent is a trion, i.e. an exciton bound to an extra carrier; in more explicitly many-body formulations, the same resonance can be described as an attractive exciton Fermi polaron or as an exciton coupled to a manifold of four-body trion states. Across these descriptions, the defining experimental signature is a trion-related avoided crossing within a multibranch polaritonic spectrum whose composition, oscillator strength, and nonlinear response are controlled by carrier density, detuning, and the photonic environment (Emmanuele et al., 2019, Hu et al., 2022, Koksal et al., 2021).

1. Definitions and microscopic interpretations

An exciton is a neutral bound electron–hole pair, a trion is a charged excitonic complex, and a polariton is a mixed eigenstate produced by strong coupling between a material resonance and a confined optical mode. In electron-doped TMDC monolayers the relevant trion is usually the negative trion, and the resulting hybrid modes often appear as lower, middle, and upper branches because both neutral-exciton and trion-related resonances participate in the coupling (Emmanuele et al., 2019).

A central conceptual issue is that a trion is not a simple boson. In a strict few-body language it is a composite fermion, which makes a naive photon–trion strong-coupling picture implausible: the direct photon–trion matrix element is parametrically small, scaling as $\left(a_T/L\right)^D$, and trion recombination naturally connects to many photon momentum channels. A moderately dense Fermi sea changes this conclusion because Pauli blocking pins the emitted photon to the initial momentum and because the occupied electron states provide a volume-linear manifold of trion–conduction-hole states that collectively amplifies the effective coupling (Shiau et al., 2016).

This few-body trion picture is not universal. In charge-tunable monolayers, one widely used interpretation is that the “trion” optical line is better viewed as an attractive exciton Fermi polaron with a finite excitonic residue $Z_X$, so that the effective light–matter coupling is $\Omega_{\mathrm{eff}} \simeq \Omega \sqrt{Z_X}$. In that formulation, trion-polaritons are cavity modes coupled to the attractive polaron branch rather than to a bare isolated three-body trion (Hu et al., 2022).

A different many-body viewpoint treats the optically relevant charged sector as a set of bosonic four-body states that are dark with respect to the material ground state and become visible only through Coulomb mixing with bright excitons. In this picture, photons couple directly only to the exciton sector, while excitons couple by Coulomb interaction to bound and unbound trion states. A recurrent misconception is therefore that every charged resonance in a doped cavity can be modeled as an independent bright trion oscillator; several microscopic theories instead regard trion-related polaritons as indirect, exciton-mediated hybrids (Rana et al., 2020, Koksal et al., 2021).

2. Experimental platforms and strong-coupling signatures

The experimental literature spans dielectric microcavities, waveguides, Tamm-plasmon structures, and plasmonic nanoantennas. In all cases, the basic spectroscopic criterion is an avoided crossing or a resolved multibranch structure associated with a trion-related optical resonance.

Platform	Representative signature	Reference
Open-access DBR microcavity with hBN/MoSe$_2$ at $4$ K	LPB/MPB/UPB and trion Rabi splitting $\hbar\Omega_T \approx 5.8~\text{meV}$	(Emmanuele et al., 2019)
Tamm-plasmon MoSe$_2$ structure	Normal-mode splitting $5.2~\text{meV}$ at positive detuning $4.2~\text{meV}$	(Lundt et al., 2017)
Single Ag nanoprism on WS$_2$	Three bright plasmon–exciton–trion branches at $Z_X$0 K; $Z_X$1	(Cuadra et al., 2017)
Photonic-crystal waveguide with doped MoSe$Z_X$2	Three branches with splittings $Z_X$3 and $Z_X$4	(Koksal et al., 2021)
Dielectric-screened WS$Z_X$5 reflectance structure	Three-branch exciton–trion–photon response with screening-tunable damping	(Cao et al., 8 Jun 2025)

In the monolayer MoSe$Z_X$6 DBR microcavity, the relevant operating regime combined a strong trion PL peak at about $Z_X$7, a weaker neutral exciton near $Z_X$8, and clear anticrossings of the cavity $Z_X$9 mode with both resonances. The resulting three-coupled-oscillator spectrum resolved lower, middle, and upper polariton branches, and at high drive power the trion anticrossing collapsed as the effective trion coupling was quenched (Emmanuele et al., 2019).

The MoSe$\Omega_{\mathrm{eff}} \simeq \Omega \sqrt{Z_X}$0 Tamm-plasmon realization established strong coupling specifically to the trion resonance by combining substrate-induced electron accumulation with a Tamm mode near $\Omega_{\mathrm{eff}} \simeq \Omega \sqrt{Z_X}$1. The fitted matter resonance $\Omega_{\mathrm{eff}} \simeq \Omega \sqrt{Z_X}$2 matched the PMMA-capped trion energy $\Omega_{\mathrm{eff}} \simeq \Omega \sqrt{Z_X}$3, and the lower branch showed $\Omega_{\mathrm{eff}} \simeq \Omega \sqrt{Z_X}$4 circular polarization at $\Omega_{\mathrm{eff}} \simeq \Omega \sqrt{Z_X}$5, substantially above the sub-1\% values reported for bare MoSe$\Omega_{\mathrm{eff}} \simeq \Omega \sqrt{Z_X}$6 under similar nonresonant conditions (Lundt et al., 2017).

Plasmonic implementations demonstrate that the phenomenon is not restricted to high-$\Omega_{\mathrm{eff}} \simeq \Omega \sqrt{Z_X}$7 dielectric resonators. In monolayer WS$\Omega_{\mathrm{eff}} \simeq \Omega \sqrt{Z_X}$8 coupled to individual Ag nanoprisms, cooling to $\Omega_{\mathrm{eff}} \simeq \Omega \sqrt{Z_X}$9 K and p-doping by a poly-lysine adhesion layer stabilized a positive trion $_2$0, transforming a two-peak plasmon–exciton spectrum into a three-peak plasmon–exciton–trion spectrum. The plasmon–trion coupling was highlighted as $_2$1, exceeding the $_2$2 trion binding scale (Cuadra et al., 2017).

3. Nonlinearity and interaction mechanisms

A defining feature of trion polaritons is their unusually strong optical nonlinearity. In monolayer MoSe$_2$3 microcavities this nonlinearity is not attributed primarily to a simple trion–trion contact interaction. Instead, it is dominated by phase-space filling and oscillator-strength saturation arising from the composite nature of trions and the limited free-electron reservoir. In the trion-dominated regime, the coupling obeys

$_2$4

and strong coupling collapses near $_2$5. Experimentally, the collapse occurred near $_2$6, consistent with $_2$7. The extracted effective nonlinear coefficient was $_2$8, compared with a theoretical estimate of $_2$9. The corresponding effective nonlinear refractive index per MoSe$4$0 flake was estimated as $4$1, and the nonlinearity was reported as $4$2 to $4$3 times larger than in other polariton systems and $4$4–$4$5 larger than in weak-coupling TMDCs, graphene, Si, or AlGaAs (Emmanuele et al., 2019).

A second mechanism is genuine Coulomb-induced trion–trion scattering. Microscopic exchange calculations for doped WSe$4$6 found that the low-momentum direct Coulomb repulsion is canceled by the Fermi-sea background, leaving exchange channels as the dominant contribution. In that analysis, electron–hole exchange is repulsive, whereas electron–electron and hole–hole exchange are attractive; for realistic parameters the attractive channels dominate, producing a net interaction

$4$7

to be compared with an exciton reference $4$8. This implies a trion-polariton interaction that is attractive, of opposite sign to the exciton-polariton reference, and about fivefold stronger in magnitude (Song et al., 2022).

At the few-quanta level, saturation nonlinearities alone are already sufficient to produce nonclassical light. A quantum theory of collective trion excitations in doped TMD microcavities predicts both unconventional blockade, driven by destructive interference of two-photon excitation pathways, and conventional blockade, driven by anharmonic dressed levels. For MoSe$4$9, estimated antibunching values were $\hbar\Omega_T \approx 5.8~\text{meV}$0 in an unconventional-blockade regime and $\hbar\Omega_T \approx 5.8~\text{meV}$1 under improved low-temperature linewidths in a conventional-blockade regime (Kyriienko et al., 2019).

4. Many-body theory and spectroscopic probes

Because the charged resonance is strongly entangled with the Fermi sea, spectroscopic interpretation requires many-body formalisms beyond a rigid two-level model. One route is the extended Chevy ansatz, which expands the trion-like polariton in a basis containing a coherent exciton component, a coherent photon component, and one particle–hole excitation of the electron gas. In that framework the trion-polariton effective coupling is again controlled by the excitonic residue, $\hbar\Omega_T \approx 5.8~\text{meV}$2, and rephasing two-dimensional coherent spectroscopy predicts three diagonal peaks and six off-diagonal cross-peaks associated with the three principal polariton branches. At a representative detuning $\hbar\Omega_T \approx 5.8~\text{meV}$3 meV, the predicted coherence periods were $\hbar\Omega_T \approx 5.8~\text{meV}$4 fs, $\hbar\Omega_T \approx 5.8~\text{meV}$5 fs, and $\hbar\Omega_T \approx 5.8~\text{meV}$6 fs for the three branch-pair combinations (Hu et al., 2022).

A complementary many-body line of work treats the optical response of a doped quantum well as a quench problem governed by the Fermi-edge singularity and the Anderson orthogonality catastrophe. In that description the absorption contains trion and exciton thresholds with power-law singularities, and the cavity photon hybridizes with this singular many-body continuum rather than with isolated Lorentzians. The same theory predicts substantial electron accumulation around the excitonic component of the lower polariton only when the Rabi frequency does not exceed the trion binding energy, because electrons otherwise do not have enough time to form the trionic screening cloud (Baeten et al., 2015).

In the ultra-strong-coupling regime, the problem changes qualitatively again. A Nambu-like $\hbar\Omega_T \approx 5.8~\text{meV}$7 Green-function formalism including counter-rotating and diamagnetic terms predicts four poles rather than the two branches of the rotating-wave approximation. In a doped cavity, the trion or polaron features survive but are reorganized by virtual-photon dressing, so that the lower, middle, and upper polaron-polariton structures are redistributed across an ultra-strong-coupling spectrum with additional branches (Bastarrachea-Magnani et al., 2024).

Microscopic three-particle diagonalization provides yet another perspective. For a doped TMDC monolayer in a high-finesse cavity, exact diagonalization of the two-electron–one-hole Hamiltonian combined with a cavity Dyson equation predicts low-density trion-polariton splittings of about $\hbar\Omega_T \approx 5.8~\text{meV}$8 at $\hbar\Omega_T \approx 5.8~\text{meV}$9, growth to about $_2$0 at $_2$1, cavity-brightened carrier-induced intermediate modes, and at high doping even strongly coupled excited-trion branches. The same framework predicts that strong coupling can lower a bright trion-polariton below a dark bare trion state, converting the effective ground state from dark to bright (Zhumagulov et al., 2021).

5. Tunability by carrier density, magnetic field, dielectric environment, and moiré structure

Carrier density is the primary tuning knob because it controls both the existence of trion resonances and their oscillator strength. In the low-density MoSe$_2$2 microcavity, the regime $_2$3 was used to justify a trion description rather than a Fermi-polaron one, while still maintaining enough electron density for appreciable trion absorption (Emmanuele et al., 2019). At higher doping, several theories instead emphasize brightening of carrier-dressed excitonic modes, excited trions, or attractive-polaron branches (Zhumagulov et al., 2021, Hu et al., 2022).

A perpendicular magnetic field adds a further layer of control. In $_2$4-doped WSe$_2$5, microscopic magnetotrion calculations found that orbital and spin effects give comparable contributions to the trion energy, while field-induced valley repopulation of the resident electron gas makes the trion-photon coupling polarization dependent. The effective coupling scales as $_2$6, so one circular polarization can strengthen while the other weakens with increasing field, producing an effective giant Zeeman splitting of the trion-polariton branches (Kudlis et al., 2024).

Dielectric screening can also tune exciton–trion hybridization. In monolayer WS$_2$7, systematic variation of the surrounding dielectric environment from suspended samples to hBN encapsulation reduced the $_2$8–$_2$9 energy separation from about $5.2~\text{meV}$0 meV to about $5.2~\text{meV}$1–$5.2~\text{meV}$2 meV, increased fitted polarizabilities, and narrowed the damping parameters from $5.2~\text{meV}$3 meV to $5.2~\text{meV}$4 meV and from $5.2~\text{meV}$5 meV to $5.2~\text{meV}$6 meV. Within the paper’s dielectric-response model, these trends sharpen a three-branch exciton–trion–photon structure and make it more robust against decoherence (Cao et al., 8 Jun 2025).

Moiré heterobilayers introduce an additional control axis. In $5.2~\text{meV}$7-doped $5.2~\text{meV}$8 moiré heterobilayers inside a metal-DBR cavity, strong coupling to a layer-hybridized trion $5.2~\text{meV}$9 with $4.2~\text{meV}$0 was combined with screening by dopant electrons and the reported absence of electron capture in the moiré lattice. The lower polariton first redshifted and then blueshifted with increasing density, giving a total shift of about $4.2~\text{meV}$1–$4.2~\text{meV}$2 meV and an extracted effective nonlinearity around $4.2~\text{meV}$3. The same system exhibited nominal diffusion lengths approaching $4.2~\text{meV}$4 for trion-derived hot polaritons (Loweimi et al., 5 Jun 2026).

6. Conceptual boundaries, controversies, and open problems

Not every trion-related spectral feature should be interpreted as a standalone trion-polariton normal mode. A microscopic scattering theory of doped microcavities shows that polariton–electron interactions become resonantly enhanced when the collision energy approaches the trion bound-state energy. In that analysis the crucial object is a trion resonance in the polariton–electron $4.2~\text{meV}$5-matrix rather than an independently quantized trion-polariton branch, and the resulting resonance is predicted to be near universal in the sense that it depends mainly on the ratio of light–matter coupling to trion binding energy (Kumar et al., 2023).

The field therefore contains a genuine conceptual split. One camp emphasizes direct trion-based pictures, especially in low-density regimes with well-resolved bound trions; another emphasizes attractive Fermi polarons or exciton-mediated coupling to dark four-body trion states. A plausible implication is that “trion-polariton” functions best as an umbrella term for cavity hybrids involving charged-exciton resonances, while the correct microscopic basis depends on density, screening, and the experiment used to interrogate the system (Hu et al., 2022, Koksal et al., 2021).

Several practical limitations remain. In the highly nonlinear MoSe$4.2~\text{meV}$6 microcavity, the cavity linewidth was only $4.2~\text{meV}$7, but the polariton linewidths remained much broader, $4.2~\text{meV}$8–$4.2~\text{meV}$9, and were attributed to coupling to an exciton–trion reservoir. The same work explicitly noted that the single-polariton quantum regime was not yet reached, that neutral-exciton nonlinearities were not captured by simple pairwise interaction theory over the full density range, and that higher-order nonlinearities become important at high density (Emmanuele et al., 2019).

These open issues define the current frontier. The combination of tunable doping, valley-selective magnetism, screening engineering, and moiré localization suggests that trion polaritons are not a single fixed quasiparticle species but a family of charged polaritonic states whose precise microscopic character remains regime dependent. That interpretive flexibility is not a weakness of the topic; it is a direct reflection of the fact that strong light–matter coupling, few-body binding, and Fermi-sea dressing become comparable on the same energy scale in doped two-dimensional semiconductors.