{"id":"e4e5f461-3b84-45f7-a4e1-7c9d1d83564b","arxiv_id":"2505.10506","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"In magic-angle twisted bilayer graphene, continuous uniaxial strain produces a gauge factor up to around 400, two orders of magnitude above conventional metals, with Curie-Weiss-like divergence near half filling.","lead":"This paper measures how the electrical resistance of magic-angle twisted bilayer graphene changes when a piezoelectric device continuously stretches or compresses it. It finds a resistance response with gauge factor above 400, and near half filling a Curie-Weiss-like temperature divergence, pointing toward strain as a probe of correlated electronic states.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Curie-Weiss claim near ν=-2 is not cleanly separated from strain-induced domain-wall and resistance-knee artifacts in the same filling regime.","rationale":"The first claim (giant elastoresistance, GF ~400) is well supported by direct ρxx(ν,ε) data in three TBG devices and the BBG control, and the authors conservatively treat GF as a lower bound. The second claim (Curie-Weiss divergence near ν=-2) is the fragile part because it is a three-parameter fit in one device, in a filling region where the same device exhibits a strain-tunable orbital-magnetic state. The authors themselves flag irreversibility and Barkhausen jumps in ρxy (SI Fig. 12) and attribute the AHE strain response to domain reconfiguration, but they do not show the analogous control for ρxx. Because GF is a longitudinal strain derivative taken near this regime, any strain-dependent domain-wall scattering or strain-shifted resistivity knee contaminates GF(T) and can masquerade as 1/(T-Θ). A saturating-field control and a hysteresis check would settle this cleanly: if the Curie-Weiss form persists in a single-domain state with no ρxx hysteresis, the microscopic interpretation is restored; if not, the paper should be revised to present the Curie-Weiss behavior as preliminary or extrinsic. The direct transport observation and the strain platform remain valid either way, so CONDITIONAL (or UNCHANGED relative to the reader's verdict) is the right disposition; the request for raw data and additional controls is a reasonable condition.","tokens_in":16341,"tokens_out":6634,"duration_ms":70495,"concrete_test":"In the θ=1.20° device, measure ρxx(Vp) in both sweep directions at ν≈-2 and base T; if ρxx shows hysteresis matching the irreversible ρxy jumps of SI Fig. 12, the gauge factor includes non-equilibrium domain-wall contributions. Then repeat the GF(T) extraction at ν≈-2 in a small magnetic field that saturates the anomalous Hall loop (single-domain state): if the Curie-Weiss form survives with comparable C and Θ, the divergence is intrinsic; if it flattens, shifts, or the fit degrades, the claimed Curie-Weiss behavior is a magnetic-domain artifact.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The gauge factor GF=(ΔR/R)/Δε is measured as a single longitudinal component near ν=-2, the same filling at which the authors report a strain-tunable anomalous Hall effect with Barkhausen jumps and irreversible sign switching attributed to orbital-magnetic domain-wall motion (Fig. 4a inset; SI Fig. 12). The Curie-Weiss form GF=C/(T-Θ)+GF0 (Fig. 4c) is the evidence for a divergent electronic susceptibility, but the paper does not establish that ρxx is free of domain-wall contributions: the main text states that the strain response of the AHE 'likely arises from a microscopic reconfiguration of orbital-magnetic domains,' and the SI data show irreversible ρxy jumps during Vp sweeps. A strain-dependent domain configuration would enter ρxx through domain-wall scattering and would produce a strain derivative that is not an equilibrium electronic susceptibility. In addition, GF(T) tracks the strain-induced spread of ρxx(T) through the sharp low-temperature knee (Fig. 4b), so a strain-shifted knee alone can generate the peaking/downturn structure without any Curie-Weiss criticality. The giant elastoresistance itself is supported by multiple devices and a BBG control, but the second, microscopic claim is not.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports in-situ uniaxial strain transport measurements on magic-angle twisted bilayer graphene (MATBG) and Bernal bilayer graphene (BBG). The authors find that MATBG devices exhibit a very large elastoresistance, with gauge factors approaching 400, two orders of magnitude above conventional metals, and strongly doping- and temperature-dependent behavior. Near half-filling of the moiré valence band (ν = -2), the gauge factor is reported to follow a Curie-Weiss form GF = C/(T - Θ) + GF0, which the authors interpret as evidence of a divergent electronic susceptibility, possibly related to nematic or isospin fluctuations. The same filling also shows a strain-tunable anomalous Hall effect with Barkhausen jumps and irreversible switching, attributed to reconfiguration of orbital-magnetic domains. The paper includes three TBG devices with different twist angles, a BBG control, detailed strain-calibration discussion in the Supplementary Information, and a discussion of possible microscopic mechanisms.","tokens_in":16614,"tokens_out":4002,"duration_ms":42490,"significance":"If the central claims hold, the paper establishes uniaxial strain as a powerful in-situ probe of correlated physics in moiré materials, with a quantitatively large elastoresistance that is unusual for a near-magic-angle graphene system. The strength of the paper is the direct transport observation in multiple TBG devices and a BBG control, together with transparent Supplementary Information that explains the strain calibration, acknowledges the lower-bound nature of the reported gauge factors, and provides fitting criteria for the Curie-Weiss analysis. The device-to-device consistency and the BBG comparison make the giant elastoresistance claim credible. The Curie-Weiss interpretation is a more delicate claim, and the manuscript itself contains an important caveat: the same filling and device show strain-dependent domain-wall behavior that could contaminate the longitudinal resistance used to extract the gauge factor. The significance of the microscopic claim therefore depends on resolving this contamination, but the experimental platform and the main elastoresistance observation are valuable regardless.","major_comments":[{"comment":"The Curie-Weiss interpretation of GF(T) near ν=-2 is not cleanly separated from domain-wall and resistance-knee artifacts that the paper itself documents at the same filling. The main text attributes the strain response of the anomalous Hall effect to a probable microscopic reconfiguration of orbital-magnetic domains, and SI Fig. 12 shows irreversible jumps in ρxy during Vp sweeps that are attributed to domain-wall motion. If strain changes the domain configuration, then domain-wall scattering can contribute to ρxx, and the strain derivative of ρxx would not be a purely equilibrium electronic susceptibility. In addition, Fig. 4b shows that GF(T) tracks the strain-induced spread of ρxx(T) through the sharp low-temperature knee; a strain-shifted knee alone can produce a peaking and downturn in GF(T) that mimics a Curie-Weiss form over a limited temperature window. I ask the authors to provide evidence that ρxx in this filling range is free of strain-dependent domain-wall contributions—for example, by showing that Rxx(Vp) sweeps are reversible in the same conditions, or by comparing GF extracted in a configuration where the AHE/domain effects are suppressed—and to explicitly model the strain-shifted knee scenario as a null hypothesis for the Curie-Weiss claim.","section":"Strain-tunable anomalous Hall effect and Curie-Weiss law behavior near ν=-2 (Fig. 4 and SI Fig. 12)"},{"comment":"The Curie-Weiss fit uses three parameters (C, Θ, and GF0) over a narrow doping range in a single device. The linearity of (GF - GF0)^{-1} versus T is not an independent confirmation of the functional form because GF0 is a fitted parameter chosen to optimize this linearity. The SI criteria (R² close to 1, GF0 close to zero, and fit over at least a decade in T) are reasonable, but the authors should demonstrate additional robustness: for example, fits with GF0 fixed to zero, the sensitivity of C and Θ to the fitting interval, and a comparison of the residuals against the alternative scenario of a strain-shifted low-temperature knee. Without this, the claim that the divergence is a genuine Curie-Weiss electronic response remains under-supported.","section":"SI 'Assessment of the Curie-Weiss fitting' and Fig. 4c inset"}],"minor_comments":[{"comment":"The phrase 'in the the anomalous Hall loop' contains a duplicated article; it should read 'in the anomalous Hall loop'.","section":"Supplementary Information Fig. 12 caption"},{"comment":"The phrase 'zero strain state of the the sample' contains a duplicated 'the'; please correct it.","section":"Main text, first paragraph of 'Giant elastoresistance in MATBG'"},{"comment":"'95% confident interval' should be '95% confidence interval'.","section":"Fig. 4 caption"},{"comment":"The symbol ν is used both for the band filling factor and for the Poisson ratio (νP); although the subscript helps, the distinction could be made more explicit near Eq. (1) to avoid confusion for readers who encounter both in the same paragraph.","section":"Eq. (1) and notation"},{"comment":"The authors state that Vp = -20 V corresponds to zero external strain based on cryogenic Raman spectroscopy in one prior device. This is an important calibration assumption; it would be helpful to state explicitly in the main text that this reference point is not measured in the present devices, even though the SI is transparent about it.","section":"SI 'Strain calibration'"}],"recommendation":"major_revision","confidential_remarks":"The paper fits the journal's scope and the giant elastoresistance observation is likely significant. The main risk is the Curie-Weiss claim, which is not cleanly separated from the domain-wall/knee artifacts that appear in the same device and filling. If the authors can rule out or quantitatively bound the domain-wall and knee contributions, the revised manuscript could be suitable for publication. The current version, however, overstates the strength of the Curie-Weiss evidence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the experimental core is solid, and the Curie-Weiss interpretation is the part to be careful with. I would send this to referees.\n\nWhat is actually new: in-situ continuous uniaxial strain on dual-gated MATBG devices, with elastoresistance mapped across doping and temperature. Gauge factors approaching 400, two orders of magnitude above ordinary metals, are directly supported in three TBG devices and are absent in a Bernal bilayer control. The strain apparatus is self-cited, but Ref. 20 is an independent engineering paper with published calibration, so that is not a problem. The authors also deserve credit for being explicit that strain values are relative and that the gauge factor is a lower bound given the assumed perfect wafer-to-graphene transmission.\n\nThe soft spot is exactly where the stress-test note lands. The Curie-Weiss claim rests on a three-parameter fit in one device (1.20 degrees) over a narrow doping window. To their credit, the SI defines explicit criteria (R2 close to 1, small GF0, at least a decade of temperature) and shows they are met only near nu=-2. But that is also the filling where the same device shows a strain-tunable anomalous Hall effect with Barkhausen jumps and irreversible domain-wall motion. The paper does not demonstrate that rho_xx is free of domain-wall scattering contributions. Separately, GF(T) is extracted from rho_xx(T) through a sharp low-temperature knee; a strain-shifted knee alone can produce the peaking/downturn structure without any critical divergence. So the Curie-Weiss behavior is not yet cleanly separated from these artifacts. The authors do discuss nematic fluctuations, entropy, and electron-phonon coupling as alternatives, and they explicitly call for future tensor measurements, which limits the damage. But calling the section 'Curie-Weiss law behavior near nu=-2' overstates what is currently demonstrated.\n\nThe data availability statement says source data are available only 'upon request.' For a paper built on a few pristine datasets, depositing raw rho_xx and rho_xy traces plus the strain calibration would make the Curie-Weiss analysis auditable. That is a fair referee request rather than a fatal flaw.\n\nWho this is for: anyone working on strain tuning of moire materials, and the MATBG correlated-state community broadly. The giant elastoresistance is the headline; the microscopic origin is secondary. My verdict: conditional on either ruling out the domain-wall/knee contamination or softening the Curie-Weiss claim to 'consistent with' rather than 'law.' Definitely worth full peer review.","headline":"The continuous strain-tuning platform and the giant elastoresistance look real; the Curie-Weiss criticality near nu=-2 is suggestive but not yet separated from domain-wall and knee artifacts.","tokens_in":17150,"tokens_out":1944,"would_cite":true,"duration_ms":19465,"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":"The paper reports that uniaxial strain applied in-situ to magic-angle twisted bilayer graphene produces a gauge factor approaching 400—hundreds of times larger than in conventional metals—and that near half-filling of the moiré valence…","keywords":["magic-angle twisted bilayer graphene","elastoresistance","gauge factor","uniaxial strain","Curie-Weiss law","nematic fluctuations","isospin moments","anomalous Hall effect"],"falsifier":"Record the longitudinal and Hall resistance simultaneously while cycling strain near half-filling at a fixed small magnetic field; if every Barkhausen jump in the Hall signal coincides with a step or kink in the longitudinal signal, or if the divergence disappears when repeated strain sweeps are confined to one reproducible magnetic-domain configuration, the Curie-Weiss claim would be shown to be a domain artifact rather than an intrinsic electronic susceptibility.","tokens_in":16130,"feed_emoji":"⚡","tokens_out":13601,"duration_ms":121440,"temperature":0.7,"pith_summary":"This paper aims to establish uniaxial strain as a sharp, continuously tunable probe of correlated physics in magic-angle twisted bilayer graphene—two graphene sheets rotated by about 1.1° to make flat moiré bands that amplify electron interactions. The key observable is the elastoresistance, the fractional resistance change per unit strain (the gauge factor), which reaches values near 400 in devices with twist angles 1.10°–1.31°—two orders of magnitude larger than the geometric gauge factor of ordinary metals. Near $\\nu=-2$, in the device whose ground state shows a strain-tunable anomalous Hall effect, the gauge factor grows on cooling and follows $GF = C/(T-\\Theta)+GF_0$ over a decade in temperature. The authors interpret this Curie-Weiss divergence as a growing electronic susceptibility—plausibly nematic fluctuations or entropy from fluctuating isospin moments—and argue that in-situ strain can expose correlated behavior in moiré materials that has previously been studied mainly in bulk crystals.","feed_headline":"Stretching magic-angle graphene changes resistance 400-fold","feed_subtitle":"Near half-filling the gauge factor diverges like a Curie-Weiss law, signaling incipient electronic order in the moiré flat band.","key_machinery":"The central object is the gauge factor $GF = (\\Delta R/R)/\\Delta\\epsilon = 1+2\\nu_P + (\\Delta\\rho/\\rho)/\\epsilon$, which separates the geometric response of the lattice from the electronic response of the material. The load-bearing identity is the Curie-Weiss form $GF = C/(T-\\Theta)+GF_0$, checked as a linear plot of $1/(GF-GF_0)$ versus $T$, which is the paper's evidence for a diverging electronic susceptibility. The enabling mechanism is a three-piezostack strain cell that applies continuous in-situ uniaxial stress to a dual-gated van der Waals device, with strain calibrated through commercial and evaporated gauges; all reported gauge factors are lower bounds because strain transmission from wafer to graphene is assumed to be perfect. Near $\\nu=-2$ the divergence is tied by the authors to coupling of strain to the isospin and orbital-magnetic degrees of freedom of the correlated state.","core_discovery":"The paper's central claim is that continuous uniaxial strain produces a giant elastoresistance in twisted bilayer graphene near the magic angle: gauge factors approaching 400, with the electronic contribution $(\\Delta\\rho/\\rho)/\\epsilon$ dominating the geometric contribution $1+2\\nu_P$. The gauge factor depends sharply on band filling, exhibits features at integer moiré fillings, and generally increases on cooling. The most consequential result is near $\\nu=-2$ in the 1.20° device, where $GF(T)$ follows $GF = C/(T-\\Theta)+GF_0$ over a decade in temperature with a Weiss temperature near zero; the authors take this as a Curie-Weiss-like divergence of a strain-coupled electronic susceptibility. They also find that compressive strain suppresses a half-filling anomalous Hall effect and can irreversibly switch its sign, which they attribute to strain-induced reconfiguration of orbital-magnetic domains, and they discuss nematic fluctuations and heavy-fermion-like fluctuating isospin moments as candidate mechanisms.","pith_inferences":["A testable extension of the strain-cell approach: measuring all in-plane resistivity tensor components ($\\rho_{xx}$, $\\rho_{yy}$, $\\rho_{xy}$) in a device with contacts both parallel and perpendicular to the stress axis would separate the isotropic strain coupling (entropy/heavy-fermion mechanism) from the anisotropic coupling (nematic mechanism), exactly the step the paper identifies as necessary","If the Curie-Weiss divergence is intrinsic, a similar continuous-strain measurement on a device without hBN alignment—or on magic-angle twisted trilayer graphene—should reveal whether the divergence is generic to flat-band correlation physics or specific to the valley-ordered orbital-magnetic state of this sample.","The irreversible, sign-changing strain response of the anomalous Hall effect raises a caution that elastoresistance sweeps near $\\nu=-2$ may move magnetic domain walls; pairing strain sweeps with simultaneous Hall readout and checking reproducibility across repeated cycles would test whether any part of the giant gauge factor is a domain-reorganization artifact."],"forward_implications":["Uniaxial strain becomes a continuously tunable, quantitative probe of correlated moiré bands: the electronic part of the gauge factor dwarfs the geometric part, so resistance changes report on strain-induced modifications of band structure and correlations.","The Curie-Weiss divergence near $\\nu=-2$ places the normal state of this device at the brink of an electronically ordered state, with a near-zero Weiss temperature, consistent with an incipient nematic or isospin instability.","The strain tunability of the anomalous Hall effect shows that orbital-magnetic domain states near half-filling can be manipulated in-situ, with compressive strain reducing the AHE amplitude and even flipping its sign.","The filling- and temperature-dependence of the gauge factor—dome-like with steps at integer $\\nu$—parallels thermodynamic entropy measurements, connecting elastoresistance to the crossover between fluctuating local moments at high temperature and a Fermi liquid at low temperature."],"supporting_citations":[{"why":"Supplies the piezoelectric strain-cell apparatus and calibration procedure that make the in-situ uniaxial strain measurements possible.","marker":"[20]"},{"why":"Provides the precedent that a diverging elastoresistance signals an electronic nematic susceptibility, the interpretive template for the Curie-Weiss gauge factor.","marker":"[15]"},{"why":"Establishes the standard resistance-bump features of MATBG used to anchor the filling and temperature axis.","marker":"[22]"},{"why":"Supplies the interpretation that the half-filling anomalous Hall effect in this device comes from valley-polarized order related to hBN alignment.","marker":"[26]"},{"why":"Provides the basis that heterostrain strongly modifies the flat bands, a candidate source of the large gauge factor.","marker":"[30]"},{"why":"Supplies the heavy-fermion model in which flat-band local moments hybridize with itinerant carriers, one of the two mechanisms proposed for the elastoresistance.","marker":"[33]"},{"why":"Shows that isotropic strain coupling to electronic entropy produces large temperature-dependent elastoresistance in an iron-based superconductor, the analogue invoked for the entropy mechanism.","marker":"[42]"},{"why":"Demonstrates that symmetric strain response can track electronic entropy, supporting the interpretation of the gauge factor dome and its turnover.","marker":"[43]"}],"fun_headline_variants":["Stretching magic-angle graphene spikes resistance 400x","Stretching magic-angle graphene yields 400x resistance change","Strain makes magic-angle graphene's resistance jump 400-fold","Magic-angle graphene shows 400x elastoresistance under strain"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper assumes that near a doping of roughly half-filling of the moiré valence band ($\\nu=-2$), the measured longitudinal resistance changes smoothly with strain; if instead strain moves magnetic domain walls and those motions leak into the resistance reading, the Curie-Weiss divergence could be a domain artifact rather than an intrinsic electronic response.","fun_headline_variants_meta":{"raw":{"variants":["Stretching magic-angle graphene spikes resistance 400x","Stretching magic-angle graphene yields 400x resistance change","Strain makes magic-angle graphene's resistance jump 400-fold","Magic-angle graphene shows 400x elastoresistance under strain"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001252,"raw_usage":{"total_tokens":5110,"prompt_tokens":899,"completion_tokens":4211,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":515,"completion_tokens_details":{"reasoning_tokens":4141}},"tokens_in":515,"tokens_out":4211,"duration_ms":32950,"temperature":1.0,"reasoning_tokens":4141,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:08:06.463419+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Record the longitudinal and Hall resistance simultaneously while cycling strain near half-filling at a fixed small magnetic field; if every Barkhausen jump in the Hall signal coincides with a step or kink in the longitudinal signal, or if the divergence disappears when repeated strain sweeps are confined to one reproducible magnetic-domain configuration, the Curie-Weiss claim would be shown to be a domain artifact rather than an intrinsic electronic susceptibility.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the piezoelectric strain-cell apparatus and calibration procedure that make the in-situ uniaxial strain measurements possible."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the precedent that a diverging elastoresistance signals an electronic nematic susceptibility, the interpretive template for the Curie-Weiss gauge factor."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the standard resistance-bump features of MATBG used to anchor the filling and temperature axis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the interpretation that the half-filling anomalous Hall effect in this device comes from valley-polarized order related to hBN alignment."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the basis that heterostrain strongly modifies the flat bands, a candidate source of the large gauge factor."},{"cited_title":"& Bernevig, B","cited_arxiv_id":null,"evidence_quote":"Supplies the heavy-fermion model in which flat-band local moments hybridize with itinerant carriers, one of the two mechanisms proposed for the elastoresistance."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows that isotropic strain coupling to electronic entropy produces large temperature-dependent elastoresistance in an iron-based superconductor, the analogue invoked for the entropy mechanism."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates that symmetric strain response can track electronic entropy, supporting the interpretation of the gauge factor dome and its turnover."}],"review_version":1}