{"id":"c23e548b-0a0f-436f-beca-0b3e97824a29","arxiv_id":"2511.08932","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"In SrTiO3, the phonon thermal Hall signal is strong in clean crystals, weak in disordered ones, partially restored by air annealing without changing thermal conductivity, and identical with metallic or insulating contacts—so it is intrinsic and disorder/strain-sensitive.","lead":"This paper shows that disorder and internal strain suppress the phonon thermal Hall effect in strontium titanate: clean crystals show a measurable transverse heat signal, dirty ones show almost none, and air annealing partially restores it without improving heat flow. The result matters because it pins down which experimental variables control a widely debated signal and clears the effect of metallic-contact artifacts.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Annealing results do not uniquely implicate strain; unmeasured carrier/defect changes could explain THE recovery without supporting decoupling from mean free path.","rationale":"The reader identified the same weakest assumption as the most load-bearing: the annealing experiment infers strain relief from the recovery of κxy without directly measuring strain or carrier content. My review confirms this is the central soft spot because the paper's headline conclusions of 'decoupled from phonon mean free path' and 'internal strain as a suppressor' both hinge on this inference. The contact experiment adds a useful control, but it only addresses parasitic heat paths, not the microscopic origin of the bulk signal. The paper's own introductory discussion of carrier-mediated phonon-drag THE (ref 44) and resonant-scattering extrinsic mechanisms (ref 30) makes the alternative interpretation concrete. The raw data—the sample dependence, annealing recovery, and contact independence—are valuable and coherent, but the interpretive leap to 'intrinsic' and 'strain-controlled' is not uniquely forced. Thus the reader's conditional verdict remains appropriate; I see no reason to strengthen or weaken it. The missing direct characterization is a solvable experimental gap, and the proposed test would settle whether the concern lands.","tokens_in":8754,"tokens_out":4376,"duration_ms":52467,"concrete_test":"On the same SrTiO3 samples before and after the 1000 °C air annealing used in Fig. 2, measure (a) the Hall coefficient (or low-temperature resistivity) to quantify carrier density changes and (b) the X-ray rocking-curve width (mosaic spread) to quantify strain/dislocation content. If the carrier density changes by more than ~10% while κxy recovers by a factor of several, the annealing effect cannot be uniquely attributed to strain relief, and the decoupling claim is not supported. A complementary control would be to vacuum-anneal a nominally identical disordered sample (introducing oxygen vacancies) and show the opposite or distinct κxy response, directly testing the carrier-confound.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the phonon THE is intrinsic and decoupled from the phonon mean free path rests primarily on the annealing experiments (Figs. 2a–d). The authors assume air annealing at 1000 °C relieves internal strain without altering oxygen-vacancy or charge-carrier content, but no direct strain or carrier-density measurements are presented. This assumption is load-bearing because κxx is phonon-dominated in these crystals; the flat κxx after annealing does not constrain the carrier or dilute-defect content. An equally consistent scenario is that air annealing reduces oxygen vacancies or changes dilute resonant scatterers, altering a carrier-mediated phonon-drag or skew-scattering contribution to κxy while leaving the phonon-dominated κxx nearly unchanged. Such a scenario would not imply that the THE is intrinsic or that its amplitude is decoupled from the phonon mean free path. The paper itself acknowledges that vacuum annealing introduces carriers that can produce a phonon-drag THE (ref 44), and it cites extrinsic resonant-scattering mechanisms (ref 30) that could suppress or cancel κxy without affecting κxx. Without ruling out these channels, the 'decoupling' conclusion and the attribution of the recovery to internal strain are underdetermined. The ambiguity directly undermines the strongest interpretation, not the raw data.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports a systematic thermal Hall study of insulating SrTiO3 single crystals. Four crystals with different disorder levels are compared: high-κxx samples show a clear phonon thermal Hall signal (Hall angle up to 0.3% at 9 T), while low-κxx samples show essentially no signal. Air annealing of the disordered samples partially restores the thermal Hall signal without notably changing κxx. Measurements with metallic silver-paste contacts and insulating grease contacts on the same sample give indistinguishable transverse signals. The authors conclude that the phonon thermal Hall effect is intrinsic to the ideal lattice, that its amplitude is decoupled from the phonon mean free path, that disorder and internal strain are the dominant suppressors, and that metallic-contact artifacts are ruled out.","tokens_in":9048,"tokens_out":2498,"duration_ms":28638,"significance":"If the central interpretation holds, the paper makes an important contribution to the phonon thermal Hall debate: it identifies sample quality and internal strain as controlled experimental variables, which could explain the long-standing sample-to-sample irreproducibility and provide a sharp guide for theory. The experimental strengths are the multi-sample comparison, the annealing protocol, the use of two independent contact types on the same sample, and the raw field-dependent Hall-angle data. However, the paper's strongest conclusions go beyond what the present measurements can establish: the annealing result is interpreted as strain relief without any direct strain or carrier-density probe, and the 'definitive' exclusion of parasitic signals is based on a single sample at two temperatures. The empirical correlations are valuable and likely correct, but the decoupling claim and the intrinsic-lattice claim require additional supporting measurements.","major_comments":[{"comment":"The central claim that annealing restores κxy without changing the phonon mean free path rests on the assumption that 1000 °C air annealing relieves internal strain without altering oxygen-vacancy/charge-carrier content. This is not measured. The flat κxx does not constrain carrier or dilute-defect content, because κxx is phonon-dominated. A carrier-mediated phonon-drag contribution (ref. 44) or a change in resonant scatterers (ref. 30) could also produce a restored κxy with little κxx change. Without direct strain and carrier-density data (e.g., dielectric loss, Hall carrier density, or an O2-annealing control), the 'decoupling from mean free path' and 'internal strain' conclusions are underdetermined. The stress-test concern therefore lands on a load-bearing point.","section":"Annealing Dependence (Figs. 2a–d; Eq. S2)"},{"comment":"No error bars or uncertainty estimates are shown for the thermal Hall angles or for κxy. The distinction between 'virtually absent' (≤0.02% at 9 T in Sample #2) and 'recovered' (~0.1%) depends on the noise floor and on systematic uncertainties in the thermocouple geometry and calibration. Propagating these uncertainties through Eq. (S2) is essential to assess whether the annealing recovery and the contact comparison are statistically significant. As written, the reported differences could be affected by contact-position or baseline-offset effects, particularly at the 0.02% level.","section":"Figs. 1b, 2b–d, and Fig. S1; Eq. (S2)"},{"comment":"The phrase 'definitively rules out parasitic signals' is stronger than the evidence supports. The contact comparison is performed on one sample at two temperature points (28.4 K and 64.0 K) with one contact geometry. The authors themselves acknowledge that metallic-contact effects may become non-negligible for very small Hall angles or high sample resistance. A single demonstration on a high-quality sample with a relatively large signal does not exclude contact artifacts in all regimes where the phonon THE is measured. Either soften the conclusion or provide additional tests, such as varying the contact size/geometry, measuring a poor-thermal-contact configuration, or comparing with the 'bypass-suppressed' setup proposed in ref. 40.","section":"Contact Geometry (Figs. 3c–d)"}],"minor_comments":[{"comment":"The abstract states the effect is 'virtually absent' in disordered samples; the quantitative limit (≤0.02% at 9 T) should appear in the main text, not only in the supplementary figure caption, so readers can gauge the detection threshold.","section":"Abstract and Fig. S1"},{"comment":"The term 'internal-strain mosaic domains' is introduced as a speculation. It should be explicitly labeled as a hypothesis, not a conclusion, and the supportive analogy to NiPS3 should be presented as a suggestion rather than evidence.","section":"Results: Sample Dependence"},{"comment":"Typo: 'alimented' should be 'powered' or 'supplied'. Also, the information on thermocouple attachment, contact size, and thermal radiation shielding would help reproducibility.","section":"Methods"},{"comment":"Reference 21 gives a DOI '10.1103/r572-5dfm' that appears to be a placeholder or malformed; it should be corrected.","section":"References"},{"comment":"The figures lack error bars and, in some panels, clear axis labels for the Hall angle (e.g., the units are given as ratios but the scaling is not defined). Adding a scale bar or explicit numeric ticks would improve readability.","section":"Figs. 1–3"}],"recommendation":"major_revision","confidential_remarks":"The raw data and the qualitative correlations are likely to be of interest to the thermal Hall community, and the paper addresses a timely reproducibility crisis. The main risk is the interpretive overreach: the annealing experiment is presented as proof of strain-controlled decoupling from the mean free path without direct strain or carrier measurements, and the contact test is called definitive based on a single sample. Both points are fixable with additional experiments or by substantially tempering the conclusions. I would not reject the manuscript, but I would not accept it in its current form."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Take a look at this paper if you care about the phonon thermal Hall mess. It reports something concrete: in SrTiO3, crystals with high κxx show a sizable transverse signal, low-κxx crystals show almost none, air annealing partially restores the transverse signal without moving κxx, and metallic vs insulating contacts give the same answer on the same sample. That last bit is the most direct test of the contact-artifact worry I've seen, and it's clean. The sample-dependence result also gives a plausible handle on why different groups see different sizes of the effect. That's real progress.\n\nThe paper also does some things right methodologically. They deliberately annealed in air rather than vacuum because they knew vacuum introduces carriers that can produce a phonon-drag thermal Hall effect (ref 44). They used four-thermocouple configurations to check homogeneity, and they cross-checked with multiple samples.\n\nNow the soft spots. The annealing experiment is the keystone for the claim that the amplitude is decoupled from the phonon mean free path. But air annealing at 1000°C doesn't only relieve strain; it can also change oxygen-vacancy or dilute-defect populations, and the paper cites mechanisms—resonant scattering (ref 30) and phonon drag from dilute carriers (ref 44)—that could suppress or enhance κxy without leaving a visible mark on the phonon-dominated κxx. They don't measure strain or carrier density, so the inference is real but not unique. The phrase 'definitively rules out parasitic signals' is also a bit strong given it's one sample and two temperatures; the paper itself admits contact artifacts may matter below 0.02%, which is exactly the range of the small signals they are comparing. Finally, the raw data aren't public and there are no error bars on the tiny Hall angles, which makes independent verification harder.\n\nNone of that kills the empirical core. The correlations are coherent, the contact comparison is a genuine control, and the practical advice—check crystal quality before interpreting a null or a huge effect—is worth taking seriously. The interpretation needs more support.\n\nFor peer review: yes, send it out. It deserves referee time. Push the authors to add uncertainty estimates, make data available, and either measure carriers/strain or soften the decoupling language.","headline":"Useful experimental control of the SrTiO3 phonon thermal Hall effect, but the 'decoupled from mean free path' claim outruns the annealing evidence.","tokens_in":9520,"tokens_out":3152,"would_cite":true,"duration_ms":27738,"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":"This paper establishes that the phonon thermal Hall effect in SrTiO3 is an intrinsic property of an ideal crystal lattice, suppressed by disorder and internal strain, and not caused by metallic contacts.","keywords":["phonon thermal Hall effect","strontium titanate","thermal Hall conductivity","disorder","internal strain","annealing","metallic contacts","thermal conductivity"],"falsifier":"Measure strain (e.g., via x-ray diffraction peak widths or birefringence) and carrier density (e.g., via thermopower or optical spectra) on the same disordered SrTiO3 sample before and after the 24-hour air anneal; if the restored κxy appears while strain maps are unchanged, or if carrier density changes track the restoration, the strain-relief interpretation is falsified.","tokens_in":8659,"feed_emoji":"🧲","tokens_out":4676,"duration_ms":48678,"temperature":0.7,"pith_summary":"The paper sets out to settle whether the phonon thermal Hall effect is a genuine bulk property of insulators or an artifact of measurement. Using strontium titanate as a test bed, it shows that high-quality crystals show a thermal Hall angle up to 0.3% at 9 T, while disordered crystals show almost nothing. Air annealing partially restores the transverse signal while leaving the longitudinal thermal conductivity flat, so the effect's size is not controlled by the phonon mean free path. Identical results from metallic and insulating contacts rule out spurious contact signals. A sympathetic reader comes away with an intrinsic, disorder-sensitive phonon response that any correct theory must reproduce.","feed_headline":"Phonon thermal Hall effect is intrinsic to clean crystals","feed_subtitle":"Disorder and internal strain suppress the signal without changing heat conduction; metallic contacts are ruled out.","key_machinery":"The experimental lever is a set of controlled perturbations on the same material: four crystals with different intrinsic disorder, air annealing at 1000 °C to relieve internal strain, and simultaneous metallic/insulating contacts to probe spurious effects. The central observable is the thermal Hall angle ∇Ty/∇Tx, converted to κxy via κxy = (∇Ty/∇Tx)κxx. The load-bearing comparison is the annealing panel: κxx stays flat while κxy recovers, decoupling the transverse response from the phonon mean free path.","core_discovery":"On its own terms, the paper claims that the phonon thermal Hall effect in SrTiO3 is an intrinsic property of an ideal crystal lattice. The evidence is a three-part correlation: thermal-Hall amplitude tracks crystal quality rather than mean free path; annealing disordered crystals restores κxy even though κxx is unchanged; and the transverse signal is identical with metallic and insulating contacts. The authors conclude that disorder and internal strain are potent suppressors, that the amplitude is decoupled from phonon scattering, and that parasitic effects cannot be the origin.","pith_inferences":["The domain-cancellation analogy (random strain domains canceling the signal) implies a size effect: samples whose strain domains are larger than the phonon mean free path might show stronger κxy; this can be tested by comparing crystals with different domain sizes.","If the decoupling is general, then the spread of reported phonon thermal Hall magnitudes in other insulators may be dominated by uncharacterized strain and defect distributions, making κxx alone an insufficient sample-quality metric.","A controlled bending or epitaxial-strain experiment on SrTiO3 could directly test whether strain suppresses κxy while leaving κxx flat, turning the paper's annealing inference into a mechanistic proof."],"forward_implications":["Future thermal Hall experiments on insulators should characterize crystal quality (κxx peak height) before interpreting absent signals; below a sharp κxx peak the effect may simply be suppressed.","Any microscopic theory of the phonon thermal Hall effect must predict a signal that survives in a perfect lattice but is strongly reduced by static disorder or strain, and that is independent of the overall phonon mean free path.","The measured κxy in clean SrTiO3 is a bulk lattice response, so contact-based corrections are not needed to explain it in the regime studied.","Air annealing is a practical recipe for restoring the intrinsic thermal Hall response in disordered SrTiO3 without adding charge carriers, unlike vacuum annealing.","The decoupling between κxy and κxx implies that the transverse response cannot be modeled simply as an extrinsic scattering correction proportional to the phonon scattering rate."],"fun_headline_variants":["Disorder suppresses thermal Hall without changing heat flow","Annealing restores thermal Hall in disordered SrTiO3","No contact artifacts in phonon thermal Hall","Thermal Hall is decoupled from heat conduction","Phonon thermal Hall is intrinsic to clean crystals"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"That air annealing relieves internal strain without meaningfully changing oxygen-vacancy or carrier content—the paper never measures strain or carrier density, and if annealing alters carriers, the restored signal could have a different origin.","fun_headline_variants_meta":{"raw":{"variants":["Disorder suppresses thermal Hall without changing heat flow","Annealing restores thermal Hall in disordered SrTiO3","No contact artifacts in phonon thermal Hall","Thermal Hall is decoupled from heat conduction","Phonon thermal Hall is intrinsic to clean crystals"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001388,"raw_usage":{"total_tokens":5432,"prompt_tokens":696,"completion_tokens":4736,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":440,"completion_tokens_details":{"reasoning_tokens":4663}},"tokens_in":440,"tokens_out":4736,"duration_ms":31060,"temperature":1.0,"reasoning_tokens":4663,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T22:42:42.191819+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure strain (e.g., via x-ray diffraction peak widths or birefringence) and carrier density (e.g., via thermopower or optical spectra) on the same disordered SrTiO3 sample before and after the 24-hour air anneal; if the restored κxy appears while strain maps are unchanged, or if carrier density changes track the restoration, the strain-relief interpretation is falsified.","supporting_citations":[],"review_version":1}