{"id":"c7899b3b-1367-4c84-8ba0-d81f05805074","arxiv_id":"2607.23868","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"A multi-Gaussian single-phonon plus triangle–delta–triangle multiphonon dispersion model with Drude holes reproduces IR spectra and Fano-type TO asymmetry in Zn-doped GaAs from 300–440 K.","lead":"The authors fit temperature-dependent infrared reflectance, transmittance, and ellipsometry of intrinsic and Zn-doped GaAs with a coupled phonon–carrier dispersion model, producing consistent far- and mid-IR optical constants from 300–440 K. The work supplies a practical dataset and modeling recipe for GaAs photonic and thermo-optical design.","discovery_kind":"extension","skeptic_critique":{"model":"moonshotai/kimi-k3","headline":"The qualitative Fano attribution rests on the total phonon response and is credible, but the channel-resolved decomposition — single-phonon vs. multiphonon Fano terms that \"oppose each other and partially cancel\" — is demonstrably unconstrained exactly where it matters: TDT elements j=13–18 sit in/n","rationale":"The reader's weakest_assumption correctly identified the same soft spot — the temperature-only separation of single- and multiphonon response over a narrow 140 K window, plus the linear doping interpolation (Eqs. 15–16, 34). I agree, and would sharpen it: Table VII itself provides the evidence, with reststrahlen-region TDT parameters (j=13–18) showing relative uncertainties of 10²–10⁴, meaning the near-TO multiphonon channel is numerically unconstrained in the region where the claimed Fano cancellation occurs. The paper is unusually candid about this (Sec. IV D acknowledges intrinsic ambiguity, over-parameterization, and the value of a wider temperature range), and importantly the total optical constants and the net interaction function are stable under these correlations, so the practical deliverable (a consistent 300–440 K constant set) and the qualitative Fano attribution survive the concern. What does not survive unscathed is the channel-resolved quantitative story — partial strengths, p_ph vs p_mph doping ratios, and the specific claim that the two channels' Fano terms oppose and cancel. One further caveat worth recording: the \"no true TO shift\" conclusion is partly enforced by the shared-energy parameterization rather than extracted from the data, though the physical-prior argument (10⁻⁴ substitution level) makes it the more plausible reading. This does not warrant moving the verdict: CONDITIONAL with the reader's stated conditions (data release, uniqueness tests / wider temperature range) already prices in exactly this risk, and the proposed refit is a concrete, no-new-data way to check whether the condition actually bites.","tokens_in":31146,"tokens_out":3920,"duration_ms":94667,"concrete_test":"Refit the full dataset with all TDT elements restricted to E_mph,j < 250 cm⁻¹ (no multiphonon strength or coupling inside the reststrahlen), forcing the entire near-TO asymmetry into the single-phonon channel. Compare χ² per data point and the net interaction curve (Fig. 11c) against the published fit. If χ² is statistically indistinguishable while the recovered M_ph,j and the single/multi split shift materially, the \"opposing, partially canceling\" decomposition is not data-determined and only the total interaction is supported; if χ² degrades significantly, the decomposition carries real information.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central interpretive claim has two tiers. Tier 1: the net TO-peak asymmetry in heavily Zn-doped GaAs is phonon–free-carrier (Fano-type) interference, not a true shift of the TO resonance. Tier 2: the specific decomposition of this effect into single-phonon coupling (M_ph,j, Table VI) and multiphonon coupling (M_mph,j, Table VII), which the paper says \"oppose each other and partially cancel\" (Sec. IV D, Figs. 10b vs 11b).\n\nTier 1 is reasonably supported: the total response is well constrained (Fig. 4), the net interaction (Fig. 11c/12b) depends only on the difference between doped and undoped total phonon susceptibilities, and the physical prior (only ~4×10⁻⁴ of lattice sites substituted, so a ~1% eigenfrequency shift is implausible) is a fair — though not airtight — argument against the Drude–Lorentz \"apparent shift\" reading. Note, however, that \"coupling does not change the resonance frequency\" is partly built in by construction: all three samples share the same E_ph,j (Eqs. 15–16), so a genuinely shifted TO mode cannot be represented by the model at all. The no-shift conclusion is therefore conditional on that parameterization, and the paper's own admission that a Drude–Lorentz model with a shifted peak fits \"reasonably well\" means the data alone do not exclude it.\n\nTier 2 is the load-bearing weak point. In Table VII, the TDT elements located inside/at the edge of the reststrahlen region (j=13–18, E ≈ 258–294 cm⁻¹) have parameter uncertainties orders of magnitude larger than the values themselves (e.g., j=14: N_d = 139(571563); j=16: N_d = 400(939021); j=17: N_l = 2601(1349571); j=14: M = −14941(114566)). These are the very elements carrying the multiphonon-side Fano coupling near the TO frequency. The text concedes \"the largest uncertainties occur in the spectral interval where the samples are opaque, i.e., in the reststrahlen region\" and that the single/multi separation \"relies solely on their different temperature dependences and is, therefore, intrinsinsin","agreement_with_reader":"agree"},"referee_report":{"model":"moonshotai/kimi-k3","summary":"The manuscript presents a combined ellipsometric/reflectance/transmittance study (300–440 K) of one undoped and two Zn-doped GaAs wafers, analyzed within the authors' previously developed coupled-system dispersion framework. The dielectric function is decomposed into Sellmeier electronic poles, six Gaussian-broadened single-phonon sub-modes with complex (N+iM) coupling amplitudes, 39 triangle–delta–triangle (TDT) multiphonon elements, and two Drude terms for light/heavy holes, all tied together by f-sum-rule normalization and a shared Bose–Einstein temperature scaling. The principal physical claim is that the pronounced asymmetry of the TO-phonon peak in heavily doped GaAs is a Fano-type interference between phonon excitations and the free-hole continuum, not a genuine shift of the TO resonance; single-phonon and multiphonon coupling contributions are said to oppose and partially cancel. A by-product is a temperature-dependent set of GaAs infrared optical constants claimed to be more internally consistent than existing literature data.","tokens_in":31703,"tokens_out":2875,"duration_ms":80902,"significance":"If the results hold, the paper delivers (i) a validated, temperature-dependent set of GaAs infrared optical constants (300–440 K) with external anchors to independent literature n and α data, filling a real gap relative to existing datasets; (ii) a demonstration that Lorentzian-lineshape models fail on transparent-region transmittance where the Gaussian/TDT construction succeeds (Fig. 13); (iii) sum-rule-normalized response functions, making the temperature scaling falsifiable and the f-sum checks (Eqs. 2, 12, 13, 25, 35) internally verifiable; and (iv) a concrete, physically motivated interpretation of the heavily doped TO-peak asymmetry as phonon–free-carrier (Fano) interference, with consistency cross-checks (p_ph^#2 ≈ 5.7% vs N_fc strength ratios 5.47%/6.17%, Table VIII; E_LO from Eq. (36) matching the loss-function maximum). The authors are unusually candid about over-parameterization and ambiguity, and the constants are promised for refractiveindex.info, which raises practical impact. The main caveats are that the channel-resolved decomposition is weakly constrained and the no-shift conclusion is partly built into the parameterization.","major_comments":[{"comment":"The central interpretive claim — that the apparent TO-peak shift in sample #3 is a Fano-type asymmetry rather than a true resonance shift — is partly enforced by construction. Equations (15)–(16) constrain all three samples to share the same E_ph,j (and Eq. (17) applies one common temperature scaling), so the model cannot represent a genuinely shifted TO mode at all. The authors themselves state that a conventional Drude–Lorentz model with a ~1% shifted TO peak fits the reflectance 'reasonably well,' so the data do not discriminate between the two readings. The argument against the shifted reading (only ~10^-4 of lattice sites substituted, making a 1% eigenfrequency shift implausible) is a reasonable physical prior, but it should be labeled as such, and the no-shift conclusion should be stated as conditional on the shared-energy parameterization. As written, Sec. IV C presents the nois a","section":"Sec. IV C, Eqs. (15)–(16), discussion around Eq. (19)"},{"comment":"The channel-resolved claim that single-phonon and multi-phonon Fano contributions 'oppose each other and partially cancel' (Figs. 10(b) vs 11(b)) rests on parameters that are statistically undetermined precisely where the cancellation occurs. In Table VII, the TDT elements located in/at the edge of the reststrahlen band (j=13–18, E ≈ 258–294 cm^-1) have uncertainties orders of magnitude larger than the fitted values — e.g., M_mph,j entries −14941(114566) and 10768(110865)×10^-7 eV² for j=14,15, and N-type entries with errors ~10²–10⁴× the value. The sign and magnitude of the multiphonon coupling term at the TO frequency is therefore not constrained by the fit. The authors partially acknowledge this ('if the results are to be believed') and offer the total interaction (Figs. 11(c), 12(b)) as the robust quantity, which is the right instinct. However, the abstract and Sec. IV D still still","section":"Sec. IV D, Table VII (j=13–18); Figs. 10(b), 11(b)"},{"comment":"The model admits to strong over-parameterization (39 TDT blocks, each with N_d,l,r and M, plus six complex-amplitude phonon peaks), and the stated uncertainties are local-curvature regression errors that the authors correctly caution are 'not a full metrological uncertainty analysis.' Given this, the Conclusion's claim of 'a level of accuracy not previously available' is not supported by an uncertainty budget. Fig. 4 shows regression-based uncertainty bands only. The claim should be tempered, or supported by a propagated/comparison-based accuracy estimate in the transparent region (e.g., against the Marple/Hambleton/Skauli anchor data, which the authors estimate at ~10^-3 absolute). This matters because the deliverable most readers will use is the optical-constant set, and its accuracy statement is the load-bearing claim for that use.","section":"Secs. III, IV D (over-parameterization discussion); Conclusion; Fig. 4"}],"minor_comments":[{"comment":"The caption of Fig. 10 labels both panels '(a)'; the second should be (b). The text references Fig. 10(b), so this is presumably a typo.","section":"Fig. 10 caption"},{"comment":"Typo: 'optical+ constants'; and 'In this work, it was applied for the first time a recently published approach' is grammatically broken.","section":"Conclusion, first and last paragraphs"},{"comment":"Ref. 24 (companion Paper I) has an incomplete citation ('J. Appl. Phys. 139, (2026)') — volume/article number should be completed at proof stage.","section":"References"},{"comment":"The native surface layer is modeled as 3.31 nm amorphous gallium oxide with identical optical constants for all three samples, including the single-side-polished, heavily doped #3. One sentence justifying that the backside (unpolished) surface of #3 does not require a separate roughness/interface layer would strengthen the ellipsometric analysis.","section":"Sec. IV C (native layer assumption); Sec. II"},{"comment":"In Table VII several t_j values converge to 0.00 with uncertainties exceeding the physically meaningful range (e.g., t_5=0.00(7), t_13=1.0(6), t_14=0(3)); since t_j is bounded in [0,1] by construction, it would be cleaner to state whether bounds were imposed in the regression and, for the j with t_j indeterminate, whether fixing t_j=0 changes the result.","section":"Table VII; Eq. (27)"},{"comment":"The two E_LO values (Eq. (23), 36.003 meV vs Eq. (36), 36.1297 meV) and the explanation of why the total phonon strength must be used in the HO formula are physically reasonable but compressed; a short clarifying remark that Eq. (23) fails because N_ph alone no longer represents the full lattice polarizability in the coupled model would help readers.","section":"Sec. IV C, Eq. (23) vs Sec. IV D, Eq. (36)"},{"comment":"The data-availability statement promises deposition in refractiveindex.info; given that the tabulated optical constants are the most reusable output, a supplementary table of n,k(E,T) at representative temperatures (or a DOI for the deposited dataset) would substantially increase the paper's utility.","section":"Data Availability"}],"recommendation":"minor_revision","confidential_remarks":"The manuscript is closely tied to the authors' own program: the coupling framework is from Refs. 23, 24, 40 (companion Paper I), the fitting software is the authors' newAD2, and roughly a third of the methodology references are self-citations. This is appropriate for the Special Topic on ellipsometry methods, and the work is competently executed, but the editor may wish to note that Ref. 24 (Paper I) is cited with incomplete bibliographic data and that independent validation of the framework by other groups does not yet exist. The promised deposition of optical constants in refractiveindex.info is a genuine community service and should be verified at proof stage."},"author_rebuttal":{"model":"moonshotai/kimi-k3","summary":"We thank the referee for a careful, substantive report and for the recommendation of minor revision. The referee correctly identifies three places where the manuscript's interpretive claims outrun what the data or the uncertainty analysis strictly support: (i) the \"no true TO shift\" conclusion is conditional on the shared-energy parameterization of Eqs. (15)-(17); (ii) the channel-resolved partial-cancellation statement between single- and multiphonon Fano terms rests on TDT parameters whose regression uncertainties, in the reststrahlen region, vastly exceed their fitted values; (iii) the Conclusion's accuracy claim lacks a comparison-based uncertainty budget. We accept all three points and have revised accordingly: the no-shift statement is now explicitly labeled as conditional on the parameterization and supported by the stated physical prior (10^-4 substitution level) plus the internal cross-checks the referee noted; the partial-cancellation wording is demoted from the abstract and qualified in Sec. IV D, with the total interaction (Figs. 11c, 12b) presented as the robust observable; and the Conclusion's accuracy claim is tempered and tied to the ~10^-3 anchor-based estimate in the transparent region. We respond to each point in detail below.","responses":[{"response":"The referee is correct, and we have revised Sec. IV C accordingly. The shared-energy constraint is a modeling choice, not a fitting outcome, and we now state explicitly that the no-shift conclusion is conditional on this parameterization and on the physical prior that motivates it. The revised text reads (in essence): within our coupled-system framework, in which the lattice eigenfrequencies are assumed independent of dopant concentration (justified a priori by the ~10^-4 Zn substitution fraction, which makes a 1% eigenfrequency shift implausible), the apparent shift is naturally explained as Fano-type asymmetry. We also note, however, that the data are not entirely mute on the question: (a) the consistency of p_ph^#2 = 5.7% with the Drude strength ratios 5.47%/6.17% (Table VIII) shows the coupling amplitude scales with free-carrier density as the coupled reading predicts, whereas a genuine eigenfrequency shift would have no reason to track carrier density in this way; (b) E_LO from Eq. (36), built from the unshifted E_TO and total phonon strength, reproduces the loss-function maximum. These are consistency checks rather than proof, and the revised text presents them as such.","revision_made":"yes","referee_comment":"[Sec. IV C] The no-shift / Fano-asymmetry claim is partly enforced by construction: Eqs. (15)-(17) constrain all samples to share E_ph,j and one temperature scaling, so a genuinely shifted TO mode cannot be represented. A Drude-Lorentz fit with a ~1% shifted TO peak also fits reasonably well. The 10^-4 substitution argument is a physical prior and should be labeled as such; the no-shift conclusion should be stated as conditional on the shared-energy parameterization."},{"response":"We accept this point. The individual N and M parameters of TDT elements j=13-18 are effectively unconstrained by the fit, so neither the sign nor the magnitude of the multiphonon coupling term at the TO frequency is data-determined, and a quantitative 'partial cancellation' statement cannot be defended at the channel level. In revision we have (i) removed the channel-resolved cancellation claim from the abstract; (ii) reworded the relevant passage in Sec. IV D so that the opposition of the two contributions is presented as a qualitative feature of the model decomposition, explicitly qualified by the large uncertainties and by the intrinsic ambiguity of the temperature-based separation (which the manuscript already acknowledges); and (iii) made Figs. 11(c) and 12(b) — the total phonon-free-carrier interaction, which is robust because the sum is insensitive to the correlated redistribution among TDT elements — the basis of the phonon-plasmon coupling interpretation. We also added a sentence noting that extending the temperature range is the concrete route to constraining this decomposition, as it changes the single-/multiphonon statistical factors differentially.","revision_made":"yes","referee_comment":"[Sec. IV D, Table VII j=13-18] The claim that single-phonon and multiphonon Fano contributions oppose and partially cancel rests on TDT parameters that are statistically undetermined exactly where the cancellation occurs (M entries with uncertainties 10^2-10^4 times the values). The total interaction is the right robust quantity, but the abstract and Sec. IV D still assert the channel-resolved cancellation."},{"response":"We agree and have tempered the claim. The quoted parameter errors are local-curvature regression uncertainties, as stated in Sec. III, and Fig. 4 reflects only those; we do not have a full metrological uncertainty budget, and we now say so explicitly in the Conclusion. The revised Conclusion replaces the unqualified accuracy statement with a comparison-based one: in the transparent region the model reproduces the Marple MDM, Hambleton, and Skauli anchor data to within approximately 10^-3 absolute in n (consistent with the declared accuracy of those datasets, Fig. 2(b)), and we state this level of agreement as the accuracy estimate for the optical constants in the transparency range. We also note that the strongest external evidence for improvement over existing datasets is precisely the internal consistency across the far-IR phonon region and the transparent region — a combination existing tabulations do not provide — rather than a claim of superior absolute accuracy everywhere. In the reststrahlen region, where the samples are opaque and the fit is least constraining, we now explicitly caution that the constants carry substantially larger uncertainty (as Fig. 4 shows).","revision_made":"yes","referee_comment":"[Secs. III, IV D, Conclusion, Fig. 4] Uncertainties are local-curvature regression errors, not a metrological budget; Fig. 4 shows regression bands only. The Conclusion's claim of 'a level of accuracy not previously available' is unsupported. Temper it or provide a comparison-based accuracy estimate in the transparent region against the Marple/Hambleton/Skauli anchors (~10^-3)."}],"tokens_in":31333,"tokens_out":1445,"duration_ms":44264,"standing_objections":[]},"desk_editor":{"model":"grok-4.5","letter":"The useful core here is a joint multi-instrument fit (R, T, ellipsometry, plus literature n and α) that delivers temperature-dependent far/mid-IR optical constants for undoped and Zn-doped GaAs from 300–440 K, with sum-rule-normalized responses and a clear demonstration that plain Lorentz/Voigt tails fail once the wafer is partially transparent. That practical constant set, and the honest comparison to Skauli and classical reststrahlen fits, is what most readers will actually use.\n\nWhat is new is not the existence of multiphonon structure or phonon–plasmon coupling in p-GaAs—those are known—but the application of their coupled Gaussian + TDT + two-Drude framework (Part I / prior Si work) with explicit complex amplitudes, shared thermal factors, and a doping-mix parameter. They make a fair physical case that the heavy-doping TO asymmetry is interference with the hole continuum rather than a real ~1% lattice eigenfrequency shift: only ~10⁻⁴ of sites are Zn, and the total doped-minus-undoped phonon response is well constrained. Static conductivities also land in a sensible range relative to the wafer specs.\n\nThe soft spot is real but localized. The channel-resolved story—single-phonon and multiphonon Fano terms that “oppose and partially cancel”—sits on TDT blocks inside the reststrahlen band whose fitted N and M have uncertainties orders of magnitude larger than the values, and on a single/multi split that rests only on a narrow temperature window plus linear doping interpolation. The paper itself flags over-parameterization and opacity-limited errors there. So trust the net optical constants and the qualitative Fano attribution; treat the detailed M_ph vs M_mph decomposition as illustrative, not unique. Free-hole density held T-independent and neglected multiphonon–single-phonon coupling in the undoped crystal are ordinary modeling choices at this stage, not hidden landmines.\n\nMath and citation pattern look fine: Kramers–Kronig structure, f-sum rules, and the usual GaAs/IR literature are in place; self-citation is to their own framework, which is expected for a Part II. Data path to RefractiveIndex.info is the right move; full spectra/fit files would strengthen uniqueness claims.\n\nThis is for people who need consistent GaAs IR constants for thermo-optical or photonic modeling, and for anyone tracking practical coupled-dispersion fitting. It deserves a serious referee, not a desk reject. I would engage with the constants and the net coupling result; I would not lean hard on the fine-grained single/multi Fano split without wider T or fewer multiphonon DOF.","headline":"Solid applied IR metrology paper: usable GaAs optical constants and a credible net Fano picture, with an over-resolved single/multi-phonon decomposition that the data do not uniquely pin down.","tokens_in":32429,"tokens_out":667,"would_cite":true,"duration_ms":23234,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"A generalized IR dispersion model attributes the TO-phonon asymmetry in heavily Zn-doped GaAs to Fano interference with free holes, not a true resonance shift, and yields consistent temperature-dependent optical constants.","keywords":["GaAs","phonon-plasmon coupling","infrared dispersion model","Fano resonance","Zn-doped GaAs","multiphonon absorption","optical constants","spectroscopic ellipsometry"],"falsifier":"Extend the same simultaneous reflectance/transmittance/ellipsometry fits to a much wider temperature range (or to isotopically pure GaAs); if the extracted single- versus multi-phonon decomposition or the doping-linear coupling parameters change substantially, or if independent low-frequency conductivity disagrees with the optical static conductivity, the Fano attribution and partial strengths are not unique.","tokens_in":31789,"feed_emoji":"📡","tokens_out":974,"duration_ms":28001,"temperature":0.7,"pith_summary":"This paper validates a generalized dispersion framework for coupled phonon and free-carrier excitations on intrinsic and Zn-doped GaAs. Simultaneous fits of reflectance, transmittance, and ellipsometry from 300 to 440 K use Gaussian-broadened multi-peak single-phonon responses and triangle–delta–triangle multiphonon bands, plus two Drude terms for light and heavy holes. The model separates lattice and carrier contributions while preserving sum rules and volume expansion, and it reproduces how phonon–plasmon coupling evolves with doping and temperature. In heavily doped material the TO peak becomes strongly asymmetric; the authors show this is Fano-type interference with the free-hole continuum rather than a genuine shift of the TO resonance energy. The resulting far- and mid-IR optical constants are physically consistent and supply a quantitative tool for thermo-optical modeling of GaAs photonic structures.","feed_headline":"TO-peak asymmetry in doped GaAs is Fano, not a frequency shift","feed_subtitle":"Coupled phonon–hole dispersion fits yield consistent IR constants from 300 to 440 K for photonic modeling","key_machinery":"The generalized coupled dispersion model: complex transition strengths (N + iM) on Gaussian-broadened phonon peaks and on triangle–delta–triangle multiphonon bands, linearly interpolated with doping, plus two classical Drude terms for light and heavy holes, all constrained by f-sum rules and a shared Bose–Einstein temperature dependence.","core_discovery":"Simultaneous multi-temperature IR fits with Gaussian-broadened single-phonon distributions, TDT multiphonon bands, and dual-hole Drude terms produce self-consistent GaAs optical constants and demonstrate that the pronounced TO-peak asymmetry in heavily Zn-doped GaAs is Fano interference between lattice vibrations and the free-hole continuum, not a true change of the TO resonance frequency.","pith_inferences":["A wider temperature campaign would likely tighten the single/multi-phonon separation and reduce the strong parameter correlations noted in the reststrahlen region.","Isotopically engineered GaAs would cleanly test how much of the residual TO asymmetry is mass-disorder versus multiphonon coupling.","The same N+iM coupling construction should transfer directly to n-type GaAs or other III–Vs once light/heavy electron (or hole) Drude terms are adjusted.","Publishing the optical constants in a public refractive-index database would immediately enable thermo-optic simulations of GaAs/AlGaAs supermirrors and pump diodes."],"forward_implications":["Temperature-dependent far- and mid-IR optical constants of GaAs (300–440 K) become available for thermo-optical device simulation.","Phonon–plasmon coupling in p-GaAs can be modeled without invoking an artificial TO frequency shift.","The same coupled bound–free framework can be applied to other doped zinc-blende semiconductors and crystalline coatings.","Static conductivity emerges from the coupled response rather than being inserted as an independent parameter.","GaAs wafers can be treated as better-characterized IR reference substrates once the model is further refined."],"fun_headline_variants":["Zn-doped GaAs TO asymmetry is Fano, not a TO frequency shift","Fano phonon-hole interference shapes TO peak in heavily doped GaAs","Multi-T IR fits: TO asymmetry from Fano, not resonance-frequency change","Dual-hole Drude plus phonons explain Fano TO lineshape in Zn:GaAs","Self-consistent IR constants show Fano TO asymmetry in doped GaAs"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The split between single- and multi-phonon strengths, and the linear doping scaling of the coupling amplitudes, are assumed reliable even though the temperature window is only 300–440 K and multiphonon–single-phonon coupling in the undoped crystal is neglected.","fun_headline_variants_meta":{"raw":{"variants":["Zn-doped GaAs TO asymmetry is Fano, not a TO frequency shift","Fano phonon-hole interference shapes TO peak in heavily doped GaAs","Multi-T IR fits: TO asymmetry from Fano, not resonance-frequency change","Dual-hole Drude plus phonons explain Fano TO lineshape in Zn:GaAs","Self-consistent IR constants show Fano TO asymmetry in doped GaAs"]},"model":"grok-4.5","effort":"low","cost_usd":0.00533,"raw_usage":{"total_tokens":1426,"prompt_tokens":752,"num_sources_used":0,"completion_tokens":89,"cost_in_usd_ticks":53304000,"prompt_tokens_details":{"text_tokens":752,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":585,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":752,"tokens_out":89,"duration_ms":10873,"temperature":1.0,"reasoning_tokens":585,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-30T14:00:52.713728+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Extend the same simultaneous reflectance/transmittance/ellipsometry fits to a much wider temperature range (or to isotopically pure GaAs); if the extracted single- versus multi-phonon decomposition or the doping-linear coupling parameters change substantially, or if independent low-frequency conductivity disagrees with the optical static conductivity, the Fano attribution and partial strengths are not unique.","supporting_citations":[],"review_version":2}