Mechanisms of blueshifts in organic polariton condensates

Organic semiconductors host Frenkel excitons with large binding energies (0.5–1 eV) and strong localization on single molecules, enabling room-temperature polaritonics. Polariton condensation and related coherent phenomena have been observed in various organic microcavities, yet the microscopic origin of nonlinear energy shifts (blueshifts) at condensation threshold remains unclear. In inorganic (Wannier–Mott) systems, blueshifts are attributed to repulsive Coulomb exchange interactions; such interactions should be dramatically suppressed for localized Frenkel excitons. The research question is: what mechanisms produce the ubiquitous step-like blueshift at condensation threshold in organic microcavities? The study aims to disentangle possible contributions including intracavity Kerr nonlinearity, gain-induced frequency pulling, polariton–polariton and polariton–exciton interactions, and saturation-induced effects (Rabi splitting quenching and cavity mode renormalization), and to explain the concomitant increase in degree of linear polarization at threshold.

Prior work established strong exciton–photon coupling in organics and room-temperature polariton condensation and lasing across the visible spectrum, including demonstrations of polariton transistors and superfluid propagation. In inorganic microcavities, blueshifts at threshold are commonly linked to interparticle Coulomb interactions. In organics, previous reports consistently observe step-like energy shifts and polarization behavior at threshold across diverse materials, but the microscopic origin has remained unresolved due to the localized nature of Frenkel excitons which weakens exchange interactions. Studies have also examined cavity-mode renormalization in weakly coupled microcavities and carrier-density dependent refractive index effects in inorganic ZnO systems, providing context for considering refractive index changes and saturation phenomena as contributors to energy shifts.

• Samples and strong coupling characterization: Fabricated planar microcavities with a ~λ/2 spin-cast film of BODIPY-G1 dye dispersed in polystyrene between two DBRs (bottom: 10 SiO2/Nb2O5 pairs; top: 8 pairs). Spin-casting produced a thickness gradient enabling a broad range of exciton–photon detunings. Angle-resolved reflectivity and non-resonant PL provided polariton dispersion; a coupled-oscillator fit yielded vacuum Rabi splitting Ω ≈ 116 meV and negative detuning (e.g., −160 meV), with δ spanning approximately [−240, −120] meV across the sample.
• Condensation measurements: Non-resonant, horizontally polarized excitation with single 2 ps pulses at 400 nm in transmission. Fourier-space imaging recorded dispersion below and above threshold. Extracted PL intensity, linewidth (FWHM), emission energy shift, and degree of linear polarization near k≈0 as functions of excitation density; condensation observed around ~6 mJ cm−2 excitation density (~120 µJ cm−2 absorbed fluence).
• Kerr nonlinearity (Z-scan): Measured closed- and open-aperture Z-scan on neat films (600 nm) using 140 fs pulses at 545 nm, 10 Hz, tightly focused (16 µm radius). Incident pulse energies 9.5 nJ (≈17 GW cm−2 at focus) and 438 nJ (≈779 GW cm−2). Extracted Im[χ(3)] = 1.71×10−20 m2 V−2 and Re[χ(3)] = 2.17×10−20 m2 V−2; derived n2 ≈ +1.89×10−14 cm2 W−1 at high intensity.
• Gain-induced frequency pulling test: Recorded ASE from a control film (no cavity) using 355 nm, 350 ps, 1 kHz excitation with stripe pumping (1470 µm × 80 µm). ASE peaked at 2.272 eV (545.8 nm). Employed single-shot dispersion imaging across the cavity thickness gradient to measure blueshift ΔE vs lower polariton energy, avoiding averaging over laser fluctuations. Compared ΔE distributions to ASE spectrum.
• Interaction analysis: Determined exciton fraction |X|2 vs energy from linear dispersion fits (Ω and δ maps). Analyzed ΔE vs |X|2 to test contributions from pair-polariton (∝|X|4) and polariton–exciton (∝|X|2) interactions.
• Saturation and Kramers–Kronig analysis: Modeled saturation-induced quenching of Rabi splitting Ω = Ω0√(1 − 2(nx+np)/n0). Used Kramers–Kronig relations to compute refractive index change Δn from reduced extinction k(ω) (absorption modeled by Gaussians at 2.446 and 2.548 eV); expressed n(ω) via Dawson functions. Estimated cavity mode shift ΔEc from Δn and combined with Rabi quenching to predict ΔELPB across detuning; evaluated relative contribution ratio p of cavity renormalization vs Rabi quenching.
• Intermolecular energy transfer model: Built coupled rate equations distinguishing populations aligned parallel (N∥) and perpendicular (N⊥) to pump polarization, with reservoirs and polariton states, including intermolecular energy transfer (γxx), stimulated relaxation to polaritons (γxp), polariton decay (γp), and nonradiative decay (γNR). Numerical simulations fitted PL intensity, blueshift, and degree of linear polarization vs pump using variable γxp and fraction of strongly coupled molecules fc.
• Comparative study: Repeated analysis for a lower dye concentration (~4% vs 10%), finding reduced Ω (72 meV) and altered detuning; compared blueshifts vs threshold power and vs exciton fraction, fitting with the saturation model to extract saturation parameters ξ.

• Strong coupling confirmed with vacuum Rabi splitting Ω ≈ 116 ± 1.5 meV (10% dye), detuning δ spanning −240 to −120 meV. Under non-resonant 2 ps pumping, condensation occurs at ~6 mJ cm−2 excitation density (~120 µJ cm−2 absorbed), with PL linewidth narrowing from 1.6 nm to 0.25 nm and a step-like increase in degree of linear polarization and emission energy (blueshift) at threshold.
• Optical Kerr effect is not responsible for the blueshift: At 9.5 nJ (≈17 GW cm−2), no nonlinear absorption or refractive index change observed; at 438 nJ (≈779 GW cm−2), measured n2 ≈ +1.89×10−14 cm2 W−1 indicates self-focusing that would cause a redshift, not a blueshift. Extracted Im[χ(3)] = 1.71×10−20 m2 V−2 and Re[χ(3)] = 2.17×10−20 m2 V−2.
• Gain-induced frequency pulling ruled out: Single-shot measurements across detuning show ΔE > 0 on both sides of the ASE peak at 2.272 eV (545.8 nm); no negative shifts on the low-energy side of gain, inconsistent with gain-pulling dominance.
• Interparticle interactions do not dominate: ΔE vs exciton fraction |X|2 shows a sub-linear power-law with exponent β ≈ 0.7, inconsistent with quadratic (pair-polariton) or linear (polariton–exciton) dependence expected from Coulomb exchange-driven mechanisms.
• Saturation-induced mechanisms identified: Saturation of molecular transitions induces (i) Rabi splitting quenching and (ii) cavity mode energy renormalization via Δn from Kramers–Kronig. Modeling shows the cavity renormalization contribution dominates the total blueshift (ratio p ≈ 1) across accessible detunings.
• Intermolecular energy transfer explains step-like threshold behavior: Below threshold, efficient energy migration depolarizes emission (low DLP). At threshold, stimulated relaxation to the ground-state polariton aligned with pump polarization outpaces energy transfer, producing a step-like increase in both DLP and blueshift via sudden saturation of optically aligned molecules.
• Rate-equation simulations reproduce experimental trends using γNR = 2.5×10^8 s−1, γp = 1×10^13 s−1, γxx = 3.33×10^10 s−1, and fitted γxp and fc; turning off energy transfer (γxx=0) removes the step-like behavior, yielding a linear shift that saturates above threshold.
• Concentration dependence corroborates mechanism: A 4% dye sample shows Ω ≈ 72 meV and ~27% lower absorbed threshold power; at equal exciton fraction, it exhibits larger blueshifts than the 10% sample. Fits yield saturation parameters ξ10% ≈ 0.03 and ξ4% ≈ 0.06 (≈2× stronger saturation), consistent with stronger cavity renormalization contribution at smaller |δ| and lower concentration. The trend persists across detuning/exciton fractions.

The study demonstrates that in organic microcavities, where Frenkel exciton localization suppresses Coulomb exchange interactions, the ubiquitous blueshift at polariton condensation threshold originates from saturation of molecular optical transitions. Two linked saturation effects—Rabi splitting quenching in strongly coupled molecules and cavity mode energy renormalization via refractive index reduction in weakly or uncoupled molecules—shift the lower polariton energy. Kramers–Kronig analysis and detuning-dependent measurements show that cavity renormalization dominates the total shift. The observed step-like blueshift and concomitant increase in degree of linear polarization at threshold are explained by competition between ultrafast intermolecular energy transfer (which depolarizes emission below threshold) and stimulated relaxation into the ground polariton mode aligned with the pump polarization (which becomes dominant at threshold). This framework accounts for the positive blueshift irrespective of gain spectral position and the sub-linear ΔE vs |X|2 dependence, reconciling experimental observations across materials and detunings. The comparative lower-concentration sample further supports the role of saturation and uncoupled molecules via stronger blueshifts and fitted higher saturation parameter. These findings clarify that blueshifts in organics reflect saturation physics and energy migration rather than interparticle Coulomb interactions, impacting how threshold behavior is interpreted in organic cavities.

This work identifies and quantifies the mechanisms behind blueshifts in organic polariton condensates. The blueshift arises from saturation-induced Rabi splitting quenching and, predominantly, cavity mode energy renormalization via refractive index changes caused by saturation of molecular transitions; intermolecular energy transfer dynamics produce the characteristic step-like increase at condensation threshold and concurrent rise in linear polarization. The model and measurements rule out intracavity Kerr effects, gain-induced frequency pulling, and dominant Coulomb-driven interparticle interactions as primary causes. An analytic framework (including Kramers–Kronig-based Δn and saturation parameter ξ) enables evaluation of relative contributions across detuning. Practically, blueshifts at the onset of nonlinear emission cannot alone distinguish strong-coupling condensation from weak-coupling lasing in organic microcavities, since cavity renormalization can occur without condensation. Future research directions include time-resolved quantification of energy transfer and stimulated relaxation rates, exploration across diverse molecular systems and cavity architectures to map parameter dependencies (e.g., linewidths, oscillator strengths, fc, thickness), and strategies to control energy migration and saturation to tailor nonlinear responses.

• The analysis relies on models with simplifying assumptions, including small saturation parameter (ξ ≪ 1), a small fraction of strongly coupled molecules (fc ≪ 1), Gaussian decomposition of absorption for Kramers–Kronig analysis, and effective-medium treatment of refractive index changes.
• Intermolecular energy transfer rates and stimulated relaxation are inferred via fitting of rate-equation models rather than directly measured with ultrafast time-resolved spectroscopy under identical conditions.
• The study focuses on BODIPY-G1 systems at two concentrations and specific cavity designs; generalization to other molecular systems, disorder levels, and cavity architectures may require additional validation.
• Single-shot and spatially varying detuning measurements mitigate averaging effects but residual spatial inhomogeneities could influence extracted dependencies.