{"id":"e2921732-be51-4d6b-b1dc-eec747660907","arxiv_id":"2506.04979","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Room-temperature Mie-polaritons form in monolayer WSe2 on WS2 nanoantennas on a gold mirror, with 78 to 86 meV Rabi splitting and a reported order-of-magnitude nonlinearity enhancement.","lead":"Researchers coupled the exciton in a single monolayer of WSe2 to the Mie resonance of a tiny WS2 nanoantenna sitting on a gold mirror, observing room-temperature polaritons with a Rabi splitting above 80 meV. The result shows that hybrid van der Waals dielectric nanoantennas on metal can host strong light-matter interactions and large nonlinearity in a compact geometry.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The nonlinearity claim rests on bi-Lorentzian fits of unresolved polariton branches and ignores fluence-induced broadening, so the order-of-magnitude nonlinearity assertion is not established.","rationale":"The reader's verdict is already CONDITIONAL, and my analysis supports keeping it there rather than accepting or rejecting. The low-power strong-coupling evidence is genuinely solid: anti-crossing appears in both dark-field and reflectance-contrast spectra, the FDTD simulation reproduces the splitting, and the splitting exceeds the half-sum of linewidths. The weakest link is the fluence-dependent measurement that underpins the abstract's nonlinearity claim. The reader flagged the high-fluence branch-merging as fragile, but framed the weakest assumption mainly as the single-mode coupled-oscillator model for the strong-coupling analysis. My concern is more specific: even granting the mode assignment, the conversion from observed splitting reduction to oscillator-strength reduction is invalid unless linewidths are shown to be fluence-independent, because SI Eq. (5) makes the splitting explicitly dependent on the linewidth difference. The reported collapse of strong coupling and the 20 meV branch shifts are consistent with broadening alone. A single refit with free linewidths would settle whether the oscillator-strength reduction is real. Therefore the verdict remains CONDITIONAL, pending a robust nonlinearity analysis that disentangles coupling-strength reduction from broadening.","tokens_in":14843,"tokens_out":7106,"duration_ms":91664,"concrete_test":"Refit the fluence-dependent RC spectra of Fig. 4a using the full complex coupled-oscillator model of the SI (Eqs. 4-5) with g fixed to the low-fluence value and gamma_X and gamma_ph as free fluence-dependent parameters. If this model fits the high-fluence spectra as well as the published two-Lorentzian fits (e.g., comparable chi-squared), then the observed splitting reduction can be explained entirely by broadening and the claimed f/f0 ~ 0.49 oscillator-strength reduction, and with it the order-of-magnitude nonlinearity, is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central strong-coupling observation is supported by multiple independent measurements, but the headline nonlinearity claim is not. In the SI coupled-oscillator model, the observable splitting is hbar*Omega = sqrt(4g^2 - (gamma_ph - gamma_X)^2) (SI Eq. 5). Thus Omega shrinks if either the coupling g drops or the linewidth difference |gamma_ph - gamma_X| grows. The paper converts the fluence-dependent Omega/Omega_max in Fig. 4b directly into an oscillator-strength reduction via Omega proportional to sqrt(f), implicitly holding gamma_ph and gamma_X constant. However, Fig. 4a states that at the highest fluences the LP and UP branches 'cannot be resolved', meaning the two-Lorentzian fits used to extract the shifts in Fig. 4b are ill-constrained exactly in the regime that defines the claimed order-of-magnitude nonlinearity. No fluence-dependent linewidths, no bare-NA control under identical pulse conditions, and no goodness-of-fit comparison with a single-broadened-resonance alternative are reported. Since a fluence-induced increase in gamma_X (or gamma_ph from heating) would reproduce the observed decrease in peak separation with g fixed, the data do not uniquely demonstrate a reduction of exciton oscillator strength. The 'order of magnitude' nonlinearity is therefore an interpretation, not a demonstrated result.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports room-temperature exciton-polaritons formed by coupling the exciton in a monolayer WSe2 to the electric-dipole Mie resonance of individual WS2 nanoantennas placed on a gold mirror. Rabi splittings of 86±5 meV (dark-field), 78±6 meV (reflectance contrast), and 85.4±0.47 meV (FDTD) are extracted from anti-crossing fits. The authors further claim that the Mie-polaritons exhibit nonlinearity one order of magnitude larger than the bare monolayer exciton, with up to 20 meV lower-polariton blueshift at 145 µJ/cm2.","tokens_in":15167,"tokens_out":6630,"duration_ms":77600,"significance":"If the strong-coupling observation holds, the demonstration is significant: it extends Mie-polariton strong coupling to externally placed monolayers using a compact van der Waals dielectric-metal platform, and it is supported by two independent experimental techniques plus a forward FDTD simulation using literature optical constants. The central anti-crossing evidence is credible and the FDTD simulation is a strength because it is a forward model without fitting to the target result. The nonlinearity claim, however, is not established by the present data, and the order-of-magnitude statement should be moderated or supported by additional analysis.","major_comments":[{"comment":"The quantitative reduction of Ω/Ωmax to 0.7 is obtained from bi-Lorentzian fits of spectra in which, as stated in the text, 'the LP and UP peaks cannot be resolved' at high fluences. In this regime the two-peak fit is not uniquely constrained, so the extracted peak positions and the resulting Ω values are not reliable. The authors should either restrict the claim to fluences where the branches are resolved, or provide model-independent evidence, such as second-derivative analysis or a comparison of single-resonance versus two-resonance fits with goodness-of-fit metrics.","section":"Nonlinearity of Mie-polaritons, Fig. 4"},{"comment":"The conversion Ω/Ωmax ∝ sqrt(f/f0) implicitly assumes that the linewidths γph and γX remain constant with fluence. However, SI Eq. (5) shows that the observable splitting depends on |γph − γX| as well as on g, so an increase in the linewidth difference at high fluence (e.g., from local heating) would reduce Ω even if the coupling strength were unchanged. The paper does not report fluence-dependent linewidths of the polariton branches, nor a bare-NA control under identical pulsed excitation, nor a test against the alternative of a single broadened resonance. The attribution of the observed Ω reduction to a 51% reduction in exciton oscillator strength is therefore not unique, and this is load-bearing for the order-of-magnitude nonlinearity claim.","section":"Nonlinearity of Mie-polaritons and SI coupled-oscillator model"},{"comment":"For NA radii above approximately 127 nm, the bare NAs exhibit an additional Mie-plasmonic (MP) mode (Fig. 1e), but the coupled-oscillator model in the SI includes only one photonic mode. The authors should justify that the upper fitted feature assigned to the upper polariton branch is not the MP mode, for example by comparing fitted branch energies with the bare-NA spectra in Fig. S5 and with the FDTD spectra that include the monolayer. This is important for the anti-crossing interpretation for the largest radii, where the MP mode is present.","section":"Fig. 2 and SI coupled-oscillator model"},{"comment":"The paper states the strong-coupling condition as ħΩR > (γ0 + γMie)/2. For the RC measurement, the reported Ω = 78 ± 6 meV is compared with a threshold of approximately 74 meV, so the inequality is not satisfied within the reported uncertainty. The authors should propagate the uncertainties in Ω, γ0, and γMie and state explicitly which criterion (e.g., SI Eq. 5) is used to conclude strong coupling, since the RC measurement alone is borderline.","section":"Strong-coupling criterion, Fig. 2e-f"}],"minor_comments":[{"comment":"The SI text refers to 'Fig.3 of the main text' for the DF/RC fits, but those fits are shown in Fig. 2d/f; Fig. 3 is the FDTD simulation. The cross-reference should be corrected.","section":"SI coupled-oscillator model"},{"comment":"The Fig. S7 caption lists powers from 5 nW to 20 µW, whereas the main text reports incident average powers from 12 nW to 900 nW for the same fluence range. Please reconcile these numbers or clarify the difference.","section":"Fig. S7"},{"comment":"The Methods section contains minor typographical errors, such as 'using a electron lithography', which should read 'using an electron lithography process'.","section":"Methods"},{"comment":"The abstract states a Rabi splitting above 80 meV, but the RC value is 78 ± 6 meV. Please qualify the statement, for example by referring to the DF and FDTD values, or specify that the 'above 80 meV' claim refers to those measurements.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":"The paper builds directly on the authors' prior work (Ref. [38]) on Mie-plasmonic resonances in WS2 nanoantennas on gold. The new element is the demonstration of strong coupling to an externally placed monolayer and the claim of enhanced nonlinearity. The strong-coupling data are reasonably convincing, but the nonlinearity claim, which is a headline result, is currently supported only by an indirect and under-constrained analysis. I would ask the authors to either provide the missing linewidth and control data or to substantially moderate the nonlinearity claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The strong-coupling half of this paper is solid; the nonlinearity claim is not.\n\nWhat's new here is genuinely new: the first experimental demonstration of room-temperature Mie-polaritons formed between a Mie resonance in an individual van der Waals dielectric nanoantenna and an external monolayer TMD exciton. The trick of placing the WS2 nanoantenna on a gold mirror works: the field is pulled to the top surface, and the coupling to the overlying WSe2 monolayer is clear. The evidence for strong coupling is good: an anti-crossing in dark-field and reflectance-contrast spectra, with extracted Rabi splittings of 86±5 and 78±6 meV, and an FDTD simulation using literature optical constants gives 85.4±0.47 meV without fitting to the target result. The SiO2-substrate control simulation showing no strong coupling reinforces the role of the gold. The coupled-oscillator fit treats only one photonic mode, and the higher-energy Mie-plasmonic mode is ignored, but the anti-crossing is visible in two independent measurements, so this is a caveat, not a fatal one.\n\nThe soft spot is the high-fluence nonlinearity. The authors extract Rabi splitting from bi-Lorentzian fits at fluences where they state the branches 'cannot be resolved.' Those fits are ill-constrained exactly in the regime that produces the claimed order-of-magnitude nonlinearity. More importantly, the conversion from splitting reduction to oscillator-strength reduction uses Omega ∝ sqrt(f), which is only valid if the linewidth difference gamma_ph − gamma_X stays constant. But the observable splitting in their own Eq. S5 is sqrt(4g^2 − (gamma_ph − gamma_X)^2), so a fluence-induced broadening of either linewidth would shrink the splitting with g unchanged. No fluence-dependent linewidths, no bare-NA control under identical pulse conditions, and no single-broadened-resonance alternative fit are presented. The reversible shift of up to 20 meV is interesting, but the 'order of magnitude more nonlinear' conclusion is not established by these data.\n\nSo: the central strong-coupling result deserves to be published and is likely to be useful. The nonlinearity section needs either more data or a much more careful analysis before that claim is made.\n\nI'd send this to a serious referee; the weak nonlinearity section is fixable. The paper is worth citing for the strong-coupling platform.","headline":"A credible demonstration of room-temperature Mie-polaritons in a monolayer TMD, with a nonlinearity claim that needs more scrutiny.","tokens_in":15711,"tokens_out":2734,"would_cite":true,"duration_ms":30337,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Monolayer WSe2 on a 27-nm WS2 nanoantenna over a gold mirror forms room-temperature Mie-polaritons with a Rabi splitting above 80 meV and nonlinear polariton shifts up to 20 meV.","keywords":["exciton-polaritons","Mie resonances","van der Waals nanoantennas","strong coupling","monolayer WSe2","gold mirror","Rabi splitting","nonlinear polaritons"],"falsifier":"Measure dark-field spectra of the monolayer-covered antennas over a finer radius series and fit them with a three-mode model (exciton, electric-dipole Mie, Mie-plasmonic): if the upper fitted branch tracks the Mie-plasmonic mode's dispersion rather than the two-mode anticrossing, the strong-coupling claim fails.","tokens_in":14692,"feed_emoji":"⚛️","tokens_out":10417,"duration_ms":118644,"temperature":0.7,"pith_summary":"Strong coupling normally needs a cavity or a collective array; this paper claims to get it from a single, 27-nm-tall WS2 nanoantenna sitting on a gold mirror, with the monolayer WSe2 simply placed on top. The gold mirror is the key element: it pushes the electric-dipole Mie mode's field out of the antenna toward the top surface, giving about 130 times more field there than on SiO2, so an externally placed monolayer can couple. The result is an anti-crossing between the WSe2 exciton and the Mie mode, with a Rabi splitting of 86±5 meV in dark-field scattering, 78±6 meV in reflectance contrast, and 85.4±0.47 meV in FDTD simulation. The paper also reports that these Mie-polaritons are an order of magnitude more nonlinear than the bare exciton, with a ~20 meV blueshift of the lower polariton at 145 µJ/cm2 and collapse of the splitting to ~0.7 of its low-power value. If correct, this makes single dielectric nanoantennas on mirrors a compact platform for room-temperature strong light-matter coupling with arbitrary monolayers.","feed_headline":"27-nm antenna on gold gives 80 meV Mie-polaritons at room temperature","feed_subtitle":"Strong coupling needs no cavity array, and the polariton shift beats the bare exciton's by an order of magnitude.","key_machinery":"The central object is the electric-dipole Mie mode of a 27-nm-thick WS2 nanoantenna sitting on a gold film; the gold mirror hybridizes with it and redistributes the mode so that high field intensity appears above the antenna's top face. The argument is carried by the size-tunable Mie resonance: changing the nanoantenna radius tunes the electric-dipole mode through the WSe2 exciton energy, producing an anti-crossing that is fitted with a two-level coupled-oscillator Hamiltonian (one photonic mode, one exciton) whose Rabi splitting is the extracted coupling metric. In FDTD, the same geometry is modelled with a monolayer WSe2 wrapped around the antenna, reproducing the splitting (85.4±0.47 meV) and showing that the bare mode's field is split between upper and lower polaritons. A comparison structure on SiO2 shows no strong coupling, which the paper attributes to the gold mirror's field redistribution.","core_discovery":"On the paper's own terms, the discovery is that a hybrid van der Waals dielectric-metal nanoantenna can reach the strong-coupling regime at room temperature with an exciton in a monolayer that is placed on the antenna externally rather than grown into it. The structure is a 27-nm WS2 nanoantenna on a gold film, covered by a monolayer WSe2 that conforms to the antenna. The gold substrate both raises the Q factor of the electric-dipole Mie resonance and moves its near field to the top surface of the antenna, so the WSe2 exciton at about 1.668 eV couples to the Mie mode; as the antenna radius tunes the Mie energy through the exciton, the two measured spectral features anti-cross with a Rabi splitting above 80 meV. Under pulsed excitation the polariton branches shift up to 20 meV, the splitting falls to about 70% of its low-fluence value, and the inferred oscillator-strength change is at least an order of magnitude larger than in the bare monolayer, reversibly and without damage. The paper takes this as evidence that the hybrid WS2/gold nanoantenna is a viable building block for compact room-temperature polaritonics.","pith_inferences":["This reader's extension: because the emitter is placed on top after fabrication rather than patterned, the platform should extend to other exfoliated monolayers, defect emitters, or molecules, making it a general testbed for room-temperature strong coupling.","A testable extension of the nonlinearity result: a time-resolved pump-probe measurement of the branch positions and linewidths at 145 µJ/cm2 would separate a genuine drop in coupling strength from power-induced broadening of unresolved peaks.","The reversible ~30% reduction of the normalized Rabi splitting under pulsed excitation suggests an all-optical modulation scheme, although the merging of the branches at high fluence makes the usable switching contrast an open question.","Varying the WS2 thickness or the residual gold pedestal height should allow deliberate engineering of the mode volume and the field enhancement, giving a parameter knob for the Rabi splitting beyond the 80 meV demonstrated here."],"forward_implications":["Each individual WS2/gold nanoantenna acts as a stand-alone strongly coupled cavity for a monolayer placed on top, so no periodic array or collective resonance is required to reach a Rabi splitting above 80 meV.","The gold mirror is not just a reflector but the enabling element: replacing it with SiO2 removes the strong coupling in simulation, so the metal-backed van der Waals geometry is the essence of the platform.","The Mie-polariton nonlinearity is large enough for ~20 meV branch shifts at 145 µJ/cm2 and an inferred reduction of the effective exciton oscillator strength to about half its low-fluence value, while the bare monolayer shifts by less than 2 meV over the same excitation range.","Because the monolayer is transferred on top after the antenna is fabricated, the same platform can be applied to other monolayer or few-layer emitters without changing the nanoantenna fabrication."],"supporting_citations":[{"why":"Identifies the electric-dipole and Mie-plasmonic modes in WS2/gold nanoantennas and the gold-induced field redistribution this paper exploits.","marker":"[38]"},{"why":"Provides the WS2 refractive index and the electron-beam lithography/etching fabrication procedure for the nanoantennas.","marker":"[24]"},{"why":"Provides the monolayer WSe2 refractive index used in the FDTD model of the wrapped monolayer.","marker":"[45]"},{"why":"Gives the theoretical strong-coupling framework for Mie excitons in dielectric nanoparticles that the experiment realizes in a flat antenna geometry.","marker":"[37]"},{"why":"Supplies the gold optical constants used in the FDTD simulations.","marker":"[46]"},{"why":"Models resonance coupling between a dielectric nanoparticle and a monolayer TMD, the geometry this work replaces with a fabricated WS2 antenna on gold.","marker":"[36]"}],"fun_headline_variants":["Hybrid antenna traps light to make room-temperature polaritons","Gold-backed WS2 nanoantenna couples strongly to WSe2 exciton at 300 K","80 meV Rabi splitting: van der Waals antenna on gold does it at room temperature","Thin WS2 antenna on gold reaches strong coupling with monolayer exciton","Nanoantenna on mirror yields polaritons with 80 meV splitting at 300 K"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The strongest conclusion rests on the assumption that the two spectral features tracked across antenna radii and laser fluences are just the lower and upper polariton branches of one Mie mode coupled to one exciton, with the higher-energy Mie-plasmonic mode contributing nothing to the fits.","fun_headline_variants_meta":{"raw":{"variants":["Hybrid antenna traps light to make room-temperature polaritons","Gold-backed WS2 nanoantenna couples strongly to WSe2 exciton at 300 K","80 meV Rabi splitting: van der Waals antenna on gold does it at room temperature","Thin WS2 antenna on gold reaches strong coupling with monolayer exciton","Nanoantenna on mirror yields polaritons with 80 meV splitting at 300 K"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000268,"raw_usage":{"total_tokens":1680,"prompt_tokens":1071,"completion_tokens":609,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":687,"completion_tokens_details":{"reasoning_tokens":499}},"tokens_in":687,"tokens_out":609,"duration_ms":7070,"temperature":1.0,"reasoning_tokens":499,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T10:29:37.851197+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure dark-field spectra of the monolayer-covered antennas over a finer radius series and fit them with a three-mode model (exciton, electric-dipole Mie, Mie-plasmonic): if the upper fitted branch tracks the Mie-plasmonic mode's dispersion rather than the two-mode anticrossing, the strong-coupling claim fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Identifies the electric-dipole and Mie-plasmonic modes in WS2/gold nanoantennas and the gold-induced field redistribution this paper exploits."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the WS2 refractive index and the electron-beam lithography/etching fabrication procedure for the nanoantennas."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the monolayer WSe2 refractive index used in the FDTD model of the wrapped monolayer."},{"cited_title":"Tserkezis, P","cited_arxiv_id":null,"evidence_quote":"Gives the theoretical strong-coupling framework for Mie excitons in dielectric nanoparticles that the experiment realizes in a flat antenna geometry."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the gold optical constants used in the FDTD simulations."},{"cited_title":"Lepeshov, M","cited_arxiv_id":null,"evidence_quote":"Models resonance coupling between a dielectric nanoparticle and a monolayer TMD, the geometry this work replaces with a fabricated WS2 antenna on gold."}],"review_version":1}