{"id":"15683b82-80a6-444b-a99a-125886e429cc","arxiv_id":"1908.08738","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Nonhermitian defect states form at an interface in a coupled-resonator chain when the mode pair's lifetimes split but their frequencies stay aligned, as realized by nanoparticle perturbations on symmetric microresonators.","lead":"This paper shows that chains of light-carrying resonators can host localized defect states purely from differences in how long the two internal modes of each resonator live, without any asymmetric backscattering. The finding suggests a simpler recipe for nonhermitian photonic devices such as mode-selecting lasers.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The causal claim relies on a uniform, off-diagonal-only inter-resonator coupling (Eq. 3) that is not verified at the β/π−β interface; if the interface hopping differs, the defect states may not stem from lifetime differences.","rationale":"The reader correctly identifies the two-mode, uniform-hopping assumption as the weakest point. I agree; the full-wave observation of four localized modes is strong, but the abstract's causal claim ('due to lifetime differences, without asymmetric backscattering') depends on the tight-binding interface being exactly the H vs H^T domain wall. The untested interface hopping is the sharpest version of this vulnerability. If the dimer test showed a substantially different T_interface, the matching in Fig. 6 could be coincidental or the defect states could be dominated by the hopping defect, not by the lifetime-induced A=-B condition. Even so, the full-wave defect states themselves would remain, so I do not change the ACCEPT verdict; the concern is about the explanatory claim, not the existence claim. I also note an internal inconsistency in Eq. (7): for A=-B the radicand must be 4W^2 cos^2 k - A^2, not A^2 - 4W^2 cos^2 k; the text below the equation (imaginary-axis branches for |A|>=2W) matches the corrected form, so this is likely a typographical error but should be fixed in a revision.","tokens_in":10451,"tokens_out":27924,"duration_ms":290834,"concrete_test":"Perform a full-wave dimer simulation with the two interface orientations (β_r=19π/64 and β_l=π−β_r) at the same spacing a/R=0.43. Fit the complex eigenfrequencies to a general two-site, two-mode Hamiltonian to extract the full 2×2 hopping matrix T_interface, including diagonal and cross terms. Compare it with the hopping matrix from a dimer of two β_r resonators and with the W=0.00076 used in Fig. 6. If T_interface deviates by more than a few percent in phase or magnitude, or acquires diagonal terms comparable to W, the uniform-T assumption underpinning the A=-B defect-state prediction is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"To support 'without the need for asymmetric backscattering,' the paper must show that the defect states arise from the lifetime-induced A=-B condition, not from some other feature of the interface. The two-mode tight-binding model (Eq. 3) assumes a single WGM pair per resonator and a real, uniform, off-diagonal-only hopping W on every bond. The physical interface, however, flips the nanoparticle orientation by π−2β across one bond. The overlap of the rotated WGM evanescent fields on that bond need not equal the intra-half hopping; it can acquire a phase or develop diagonal/same-direction terms. The paper neither extracts nor justifies this interface hopping. Its discrepancy statement (Sec. IV B) attributes all deviations to the two-mode approximation, but this specific assumption is untested. If the interface hopping differs substantially from W, then the effective Hamiltonian at the interface is not the simple H vs H^T domain wall, and the defect states could be caused by the hopping defect rather than by lifetime differences. The full-wave modes demonstrate localization, but the load-bearing causal claim is the mechanism, so this gap is significant.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a chain of coupled dielectric microresonators (a CROW) with a central interface created by flipping the orientation of a nanoparticle attached to each resonator. Within a two-mode tight-binding model, the authors show that if the nanoparticle splits only the lifetimes (imaginary diagonal splitting δ) of the standing-wave mode pair and if the orientation satisfies the condition 2mβ = π/2 + nπ, the effective resonator Hamiltonian acquires real, opposite off-diagonal elements A = -B. This produces the same exceptional-point mechanism for interface defect states as earlier reciprocal-asymmetric-backscattering models, but without asymmetric backscattering. The authors extract δ from single-resonator finite-element simulations and W from dimer simulations, and compare the tight-binding spectrum of a 12-resonator chain to full-wave COMSOL simulations. The full-wave calculation shows a quadruplet of interface-localized states with complex eigenfrequencies in reasonable agreement with the tight-binding prediction.","tokens_in":10715,"tokens_out":13450,"duration_ms":140462,"significance":"If correct, this result broadens the practical design space for nonhermitian defect-state photonics by removing the need for engineered asymmetric backscattering. The work's strengths are the clean symmetry derivation in Sec. III A, the fact that the chain spectrum is a genuine prediction (δ and W are not fitted to the chain spectrum), and the independent full-wave corroboration of the quadruplet of localized states. The paper is also careful to place the nanoparticle far from the coupling regions (Sec. IV A), which supports the uniform-hopping assumption. The demonstration is numerical rather than experimental, but the geometry is simple and practically accessible, making the concept readily testable.","major_comments":[],"minor_comments":[{"comment":"The claim that the interface hopping equals the bulk coupling W is justified by the choice of β far from the coupling regions, but the manuscript would be more convincing if it stated explicitly whether the dimer used to extract W was composed of two resonators with the same nanoparticle orientations as in the bulk, and whether a dimer with the flipped interface orientations was checked. As written, the reader must infer this from the later chain agreement; a one-sentence clarification would remove residual ambiguity.","section":"Sec. IV B"},{"comment":"The notation switches from A0 and B0 in Eq. (14) to A and B in Eqs. (17)-(18) without comment; please align the notation or explicitly define the relationship between the two sets of coefficients.","section":"Sec. III A"},{"comment":"There is a typo in the text: 'oberservation' should be 'observation'.","section":"Sec. III A"},{"comment":"The distinction between 'non-circular symbols' and 'unfilled symbols' is described only in the caption; adding a legend directly on the figure would improve readability.","section":"Fig. 6"},{"comment":"The phrase 'lossy coupled-resonator optical waveguide' could be misread as implying material loss; the nonhermiticity here is radiative leakage, and a brief clarification in the text would be helpful.","section":"Abstract and Sec. II"}],"recommendation":"minor_revision","confidential_remarks":"The manuscript is within the journal's scope, and the self-citation of Ref. [15] is appropriate because the present mechanism is a variant of that earlier exceptional-point mechanism. The main novelty is the explicit demonstration that reciprocal asymmetric backscattering can be replaced by a purely imaginary diagonal splitting in a simple, realistic geometry. The numerical evidence is convincing but not experimental; the editor may wish to consider whether a proof-of-concept experiment or a systematic parameter scan is required for the strongest causal claims, though I do not regard this as necessary for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead Bosch et al. The useful new thing: they show that the A=-B condition for nonhermitian defect states can be met without any asymmetric backscattering, using only a split in lifetimes between degenerate standing-wave modes. The symmetry derivation in Sec. III is clean and the tight-binding prediction is genuinely predictive—they extract δ and W from single-resonator and dimer calculations, then predict the chain spectrum. The full-wave COMSOL simulations of a 12-resonator CROW with nanoparticle perturbations show a localized quadruplet, and the complex eigenvalues line up with the tight-binding model well enough to support the mechanism. No code or data is shared, but the two-method comparison is credible.\n\nThe real soft spot is the one the stress-test flagged: the model assumes a single, real, uniform inter-resonator hopping W, including across the interface. At the interface the nanoparticle orientation flips, so the mode phases on the two sides differ; the overlap integral between left and right WGM tails can pick up a phase or develop diagonal terms. The paper neither computes nor bounds that interface hopping. It attributes the spectral mismatch to the two-mode approximation, but this specific assumption is untested. If the interface hopping differed significantly from W, the localized states could in principle come from a hopping defect rather than from the lifetime-induced A=-B. The authors do place the nanoparticles far from the coupling regions, which makes a large deviation unlikely, but 'unlikely' isn't verification. I'd want a paragraph showing the interface bond's effective coupling is close to W, either by a dimer calculation with the rotated geometry or by an argument that the phase drops out.\n\nOther weaknesses are minor: the two-mode approximation is standard and the discrepancies are small; the lack of shared data is annoying but common. The citation to Ref. [15] is appropriate—it's the mechanism they extend.\n\nWho gets value: anyone working on nonhermitian topological photonics, especially experimental groups looking for a simple way to make defect-state lasers. The design is practically accessible and the result is likely reproducible.\n\nRecommendation: worth a serious referee. The central claim is important and mostly supported; the interface-hopping gap should be addressed in a revised version. I'd accept with that request.","headline":"A clean symmetry argument plus full-wave confirmation that lifetime differences alone produce nonhermitian defect states; the interface-hopping assumption deserves a check but doesn't sink the central result.","tokens_in":11194,"tokens_out":2493,"would_cite":true,"duration_ms":28890,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Lifetime differences alone can create non-Hermitian defect states","keywords":["non-Hermitian photonics","defect states","coupled-resonator optical waveguide","whispering-gallery modes","lifetime differences","exceptional points","microresonators","tight-binding model"],"falsifier":"Simulate or fabricate the same twelve-resonator chain with the nanoparticle parameters tuned so that $\\mathrm{Re}\\,\\delta\\neq 0$ (for instance, a different $r/R$ at fixed $d/R$): if the spectrally isolated quadruplet at the interface does not disappear or shift as the tight-binding model predicts, the claim that lifetime differences alone drive the defect states is wrong.","tokens_in":10273,"feed_emoji":"💡","tokens_out":9647,"duration_ms":88468,"temperature":0.7,"pith_summary":"The paper shows that interface-localized defect states in open optical systems need not rely on asymmetric backscattering, whether reciprocal or nonreciprocal. Instead, they can arise purely from lifetime differences between otherwise degenerate resonator modes. The authors demonstrate this in a lossy coupled-resonator optical waveguide: a chain of circular microresonators, each perturbed by a nanoparticle, with the perturbation position flipped across an interface. Tight-binding analysis predicts a spectrally isolated quadruplet of defect states at the interface, and full-wave simulations of a twelve-resonator chain confirm the localization and the complex-eigenfrequency pattern. If correct, the result means a simple, symmetric, purely passive resonator chain can host non-Hermitian defect states without the elaborate coupling engineering previously thought necessary.","feed_headline":"Lifetime differences alone can create non-Hermitian defect states","feed_subtitle":"Simulations show interface-localized resonances in a simple lossy microresonator chain.","key_machinery":"The load-bearing object is the $2\\times 2$ non-Hermitian Hamiltonian that describes a single whispering-gallery mode pair, extended to a chain by evanescent inter-resonator coupling $T=\\begin{pmatrix}0&W\\\\W&0\\end{pmatrix}$. The key step is what the paper calls lifetime backscattering: when a symmetric perturbation makes the standing-wave combinations of these modes differ only in lifetime ($\\mathrm{Re}\\,\\delta=0$, $\\mathrm{Im}\\,\\delta\\neq 0$), the backscattering coefficients become equal and purely imaginary; rotating the resonator's symmetry axis by an angle $\\beta$ obeying $2m\\beta=\\pi/2+n\\pi$ turns these into real, opposite values $A=-B$. That effective term is what opens the gap and produces defect states localized at the interface where the rotation is inverted.","core_discovery":"The central discovery is that non-Hermitian defect states can be generated by lifetime (linewidth) differences alone, with no asymmetric backscattering inside or between resonators. In the two-mode description of a whispering-gallery mode pair (the clockwise and counterclockwise circulating modes of a circular resonator), a reflection-symmetric perturbation that keeps the real frequencies of the standing-wave components aligned but splits their imaginary parts yields purely imaginary backscattering coefficients; rotating the perturbation axis relative to the chain axis converts this into the real condition $A=-B$. A chain in which the perturbation angle is $\\beta$ on one side of an interface and $\\pi-\\beta$ on the other therefore realizes the same exceptional-point defect mechanism (the point at which two eigenmodes coalesce) previously tied to asymmetric coupling. The paper verifies this with finite-element full-wave simulations of a chain of twelve nanoparticle-perturbed circular resonators, obtaining a quadruplet of interface-localized eigenmodes whose complex eigenfrequencies agree with the tight-binding prediction, with discrepancies attributed to the two-mode approximation.","pith_inferences":["By extension, any local mechanism that splits only the lifetimes of a degenerate mode pair—surface roughness, material absorption, or boundary deformation—could replace the nanoparticle, broadening the set of experimental platforms where this effect should appear.","Time-resolved ringdown measurements at the interface should show the two long-lived defect states outlasting extended states, a dynamical signature the paper does not explicitly report.","Varying the inter-resonator spacing to tune the coupling $W$ across the threshold $|A|=2W$ should switch the defect states on and off, providing a direct test of the predicted phase boundary."],"forward_implications":["Non-Hermitian defect states can be engineered in symmetric resonator shapes with only local perturbations, removing the need for engineered asymmetric backscattering.","The design works with passive loss (openness) alone; no gain or nonreciprocal elements are required.","The interface-localized quadruplet is spectrally isolated and shows exponential decay away from the interface, making it detectable in scattering or transmission measurements.","Because the perturbation is local and the argument is perturbative, the concept transfers to other resonator geometries and platforms beyond the nanoparticle-decorated chain."],"supporting_citations":[{"why":"Supplies the paradigmatic reciprocal lossy resonator chain in which defect states arise via the exceptional-point mechanism with reciprocal asymmetric coupling; the present setting is formally equivalent without that coupling.","marker":"[15]"},{"why":"Describes the reciprocal lossy resonator chain under reciprocity-conserving asymmetric coupling, the assumption the paper removes.","marker":"[33]"},{"why":"Provides the two-mode effective Hamiltonian for almost-circular resonators and the constraints relating backscattering coefficients to boundary perturbations.","marker":"[36]"},{"why":"Extends the two-mode approximation to a chain of evanescently coupled resonators, giving the block-coupling form used in the tight-binding model.","marker":"[38]"},{"why":"Establishes that a nanoparticle perturbation on a circular resonator does not induce internal asymmetric backscattering, the key property exploited in the design.","marker":"[39]"},{"why":"The finite-element wave-optics solver used for all full-wave eigenfrequency and field-distribution calculations.","marker":"[42]"},{"why":"The perfectly matched layer treatment used to simulate the open radiative boundary conditions, making the computed lifetimes meaningful.","marker":"[43]"}],"fun_headline_variants":["Defect states without backscattering: just lifetimes","Pure lifetime contrast yields defect modes","No backscatter needed: lifetime-split defect states","Lifetime differences spawn non-Hermitian defects","Defect states from linewidth contrasts alone"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument hinges on the assumption that only one clockwise/counterclockwise circulating mode pair per resonator matters and that neighboring resonators couple exclusively through a real, uniform inter-mode coupling $W$ with no extra backscattering; if higher-order modes or coupling-induced backscattering become significant, the $A=-B$ condition that creates the defect states can be disrupted.","fun_headline_variants_meta":{"raw":{"variants":["Defect states without backscattering: just lifetimes","Pure lifetime contrast yields defect modes","No backscatter needed: lifetime-split defect states","Lifetime differences spawn non-Hermitian defects","Defect states from linewidth contrasts alone"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000641,"raw_usage":{"total_tokens":2935,"prompt_tokens":914,"completion_tokens":2021,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":530,"completion_tokens_details":{"reasoning_tokens":1951}},"tokens_in":530,"tokens_out":2021,"duration_ms":14847,"temperature":1.0,"reasoning_tokens":1951,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:30:41.466847+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate or fabricate the same twelve-resonator chain with the nanoparticle parameters tuned so that $\\mathrm{Re}\\,\\delta\\neq 0$ (for instance, a different $r/R$ at fixed $d/R$): if the spectrally isolated quadruplet at the interface does not disappear or shift as the tight-binding model predicts, the claim that lifetime differences alone drive the defect states is wrong.","supporting_citations":[{"cited_title":"Malzard, C","cited_arxiv_id":null,"evidence_quote":"Supplies the paradigmatic reciprocal lossy resonator chain in which defect states arise via the exceptional-point mechanism with reciprocal asymmetric coupling; the present setting is formally equivalent without that coupling."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Describes the reciprocal lossy resonator chain under reciprocity-conserving asymmetric coupling, the assumption the paper removes."},{"cited_title":"Wiersig, S","cited_arxiv_id":null,"evidence_quote":"Provides the two-mode effective Hamiltonian for almost-circular resonators and the constraints relating backscattering coefficients to boundary perturbations."},{"cited_title":"Schomerus and J","cited_arxiv_id":null,"evidence_quote":"Extends the two-mode approximation to a chain of evanescently coupled resonators, giving the block-coupling form used in the tight-binding model."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The finite-element wave-optics solver used for all full-wave eigenfrequency and field-distribution calculations."},{"cited_title":"Berenger, J","cited_arxiv_id":null,"evidence_quote":"The perfectly matched layer treatment used to simulate the open radiative boundary conditions, making the computed lifetimes meaningful."}],"review_version":1}