Pith. sign in

REVIEW 4 major objections 5 minor 13 references

At half filling of the moiré band, weak long-wavelength photons collapse the correlated gap in magic-angle twisted bilayer graphene, driving an insulator-to-metal transition and a bolometric response near 10^7 V/W.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-05 00:51 UTC pith:CVALS35X

load-bearing objection Impressive device result, but the hot-electron mechanism rests on an extrapolated thermal-decoupling ratio that likely fails at the operating point. the 4 major comments →

arxiv 2608.00488 v1 pith:CVALS35X submitted 2026-08-01 cond-mat.mes-hall cond-mat.mtrl-scicond-mat.othercond-mat.str-el

Correlated Insulator Moir\'e Bolometer

L. Elesin , A. L. Shilov , M. Kravtsov , X. Zhou , M. Lukianov , A. Kuksov , S. Jana , I. Iorsh
show 9 more authors
R. Izmaylov K. Shein I. Gayduchenko T. Taniguchi K. Watanabe K. S. Novoselov G. N. Goltsman A. Principi D. A. Bandurin
This is my paper
classification cond-mat.mes-hall cond-mat.mtrl-scicond-mat.othercond-mat.str-el
keywords correlated insulatormagic-angle twisted bilayer graphenehot-electron bolometerterahertz photoresponsefar-infrared detectionmoiré flat bandsintervalley coherencebolometric detection
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper claims that a correlated insulator can be turned into a metal by weak, long-wavelength photons alone, without structural change or above-gap excitation. In magic-angle twisted bilayer graphene at half filling of the moiré band, the absorbed radiation heats the low-heat-capacity electron system by roughly ten kelvin while the lattice barely moves, collapsing the ~1 meV correlated gap and sharply lowering the resistance. The effect is broadband (85–2140 µm), equivalent to DC Joule heating at the same absorbed power, and yields an internal voltage responsivity of (1.7±0.4)×10^7 V/W with thermal-fluctuation noise near 2×10^-15 W/√Hz, robust to magnetic fields of several tesla. If correct, it establishes correlated flat bands as ultrasensitive detectors of faint far-infrared and terahertz radiation and as a direct probe of many-body order under weak thermal perturbation.

Core claim

The discovery is that at moiré filling ν=2, a weak beam of far-infrared or millimetre-wave photons collapses the correlated insulating state of MATBG through electronic heating rather than photocarrier generation. The resistance at ν=2 rises by over an order of magnitude on cooling from 10 K to 1.7 K, tracing a gap of about 1 meV that softens and vanishes near 10 K; illumination deposits power that raises the electronic temperature, restoring density of states at the Fermi level and driving the sample metallic. The paper shows that the photoresistance measured under 0.14 and 3.5 THz illumination matches the change in differential resistance produced by DC Joule heating at the same absorbed p

What carries the argument

The load-bearing object is the many-body correlated gap at half filling of the flat moiré band—an intervalley-coherent insulating state with an energy scale of about 1 meV, the smallest energy in the system—and its extreme temperature sensitivity: it closes continuously as the electronic temperature approaches roughly 10 K. Radiation absorbed by the graphene is thermalized on femtosecond timescales into a hot Fermi–Dirac distribution at temperature T_e; because the electronic heat capacity is tiny and electron–phonon coupling is weak, T_e rises by ~10 K while T_L rises by less than ~100 mK, collapsing the gap and producing a large negative photoresistance. The argument is carried by a DC-hea

Load-bearing premise

The main device's electrons are thermally decoupled from the lattice: deposited THz power raises electron temperature by about 10 K while the lattice rises less than ~100 mK, a hierarchy measured on a different reference device and extrapolated to the main sample.

What would settle it

On the same ν=2 device at T0=1.7 K, measure T_e by Johnson-noise thermometry and T_L with an independent phonon thermometer under 0.14 THz irradiation strong enough to visibly reduce Rxx; observing ΔT_L of order 1 K when ΔT_e is ~10 K, or no ΔT_e at all, would overturn the hot-electron hierarchy and the electronic-melting interpretation.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

If this is right

  • A correlated insulator can serve as a practical bolometer in the 85–2140 µm window, with internal voltage responsivity above 10^7 V/W and a thermal-fluctuation-limited NEP around 2×10^-15 W/√Hz.
  • Because the response is thermal, it is inherently broadband and polarization-insensitive in the present geometry; spectral flatness follows from heating, not resonant absorption.
  • The device remains sensitive in perpendicular magnetic fields up to several tesla, unlike superconducting hot-electron bolometers, and its intrinsic electron–phonon cooling time is picosecond, with the measured speed set by the RC time constant of the high-resistance geometry.
  • The same heating mechanism produces weaker but same-sign photoresistance at ν=-2, and the effect size tracks the quality of the correlated insulating state, so optimized stacks should show proportionally larger responses.
  • Time-resolved photoresistance measurements could interrogate the intrinsic suppression-and-recovery dynamics of the correlated order without high-energy interband excitation.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the hot-electron hierarchy holds generally, any moiré system with a steep temperature-dependent resistance—Chern insulators, fractional states, or other correlated gaps—should show a similar bolometric collapse, making detector sensitivity a proxy for the fragility of the order.
  • The DC-heating equivalence implies the detector is self-calibrating in absorbed-power units, which could make cross-device sensitivity comparisons possible without absolute optical power calibration.
  • Because the transduction is thermal, engineering absorption (cavities, plasmonic metasurfaces) should raise external responsivity toward the internal 10^7 V/W value while preserving nanosecond-scale response, potentially enabling photon-counting in the far infrared.
  • Bolometric readout also offers a thermodynamic thermometer for the correlated phase: Rxx at ν=2 maps T_e, so the same device can measure electronic heat capacity and electron-phonon coupling in situ.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper reports a giant photoresistive response in magic-angle twisted bilayer graphene (MATBG) tuned to half filling of the moiré band. Under continuous-wave far-infrared and millimetre-wave illumination (λ = 85–2140 μm), the resistance at ν = 2 drops dramatically, and the authors attribute this to selective overheating of the low-heat-capacity electronic subsystem, which suppresses the correlated insulating gap and drives an insulator-to-metal transition. Using dual-modulation transport and a DC Joule-heating comparison, they extract an absorbed-power voltage responsivity of R_V^abs = (1.7 ± 0.4) × 10^7 V/W and a thermal-fluctuation-limited NEP ~ 2 × 10^-15 W/√Hz. The response is broadband, polarization-independent, and robust to magnetic fields up to several tesla. The central mechanistic claim is that this is a hot-electron bolometric effect, distinct from single-particle photoconductivity and dual to superconducting hot-electron bolometers.

Significance. If the hot-electron mechanism is correct, the paper demonstrates a new class of bolometric detector based on a many-body correlated insulator, with internal responsivity far exceeding typical commercial bolometers and with magnetic-field compatibility. The work also provides a transport-based probe of the fragility of the correlated gap in MATBG. The paper has clear strengths: the photoresistance is measured with a dual-modulation technique; the DC-heating comparison in Fig. 3 and Supplementary Figures 7–8 provides a consistency check across filling, temperature, and magnetic field; the polarization dependence is explicitly tested; two additional MATBG devices are measured; and the data for the main figures are provided as Source Data. The main weakness is the extrapolation of the electron–phonon/substrate thermal hierarchy from a 200 mK reference device to the operating conditions of the main device, on which the selective electronic-overheating interpretation rests.

major comments (4)
  1. [Supplementary Note 11, Eqs. S17, S33–S34] The hot-electron interpretation relies on the hierarchy G_e-ph ≪ G_⊥, quantified as ~10^-4 from a reference device at T0 = 200 mK. Using the paper's own heat-balance model P = Σ(T_e^δ − T_0^δ) with δ ≈ 5 (Eq. S17), the differential conductance G_e-ph = dP/dT_e scales as T_e^{δ−1}. Moving from 0.2 K to the claimed T_e ≈ 10 K raises G_e-ph by roughly (10/0.2)^4 ≈ 6×10^6, and even relative to the main device's 1.7 K base by ~1200. Unless G_⊥ grows by a comparable factor, the bound ΔT_L < 100 mK (Eq. S34) is not justified. The reference device also uses superconducting Al contacts; the insensitivity to suppressing superconductivity was demonstrated on that reference device, not on the main device, whose contacts are not specified as superconducting. This extrapolation is load-bearing for the claim of selective electronic overheating: without it, the photoresponse could still be bolometric bu
  2. [Supplementary Note 5, Fig. 3a–b] The incident-to-absorbed power conversion is determined by matching R_THz(ρ) with Δ(dV/dI)(P_abs) from DC Joule heating at ν = 2. This is a calibration convention rather than an independent measurement of absorbed power. The agreement across fillings, temperatures, and fields (Supplementary Fig. 8) establishes consistency of the two perturbations, but only under the assumption that both act through the same thermal variable (T_e). If a substantial component of the DC-heating response is lattice-mediated, the calibration would fold that same component into the THz channel. Consequently, the headline internal responsivity R_V^abs = (1.7 ± 0.4) × 10^7 V/W and NEP_TF ≈ 2 × 10^-15 W/√Hz are contingent on this assumption. The manuscript should state this explicitly and, ideally, provide an independent calorimetric or phonon-thermometry check.
  3. [Supplementary Note 10] The noise thermometry is performed on a separate TBG device at T0 = 3.9 K with a silicon lens, not on the main device. It measures T_e only and does not report T_L. The sublinear T_e(P) dependence and δ ≈ 5 fit are consistent with electron–phonon cooling but do not directly verify the T_e ≫ T_L hierarchy in the main device. The main-text statement that external illumination selectively overheats the electronic subsystem is therefore supported by inference from auxiliary devices and the DC-heating comparison, not by a direct measurement in the main device. Please clarify which quantities are measured in the main device and which are transferred from reference devices.
  4. [Main text, 'Performance', and Supplementary Table 1] The practical external responsivity for ideal focusing is R_V^{NA=1} ≈ 75–370 V/W, orders of magnitude below the internal R_V^abs. The abstract and Discussion emphasize the internal responsivity and compare it with commercial bolometers, without making clear that the external figure of merit in the present device is set by diffraction-limited optics and poor absorption. While the distinction is stated in the Results, the abstract's 'millivolts per nW' phrasing refers to absorbed, not incident, power. This should be clarified to avoid overstating the device-level performance.
minor comments (5)
  1. [Methods, first paragraph] Typo: 'T ransport and photoresponse measurements' should be 'Transport and photoresponse measurements'.
  2. [Supplementary Figure 17 caption] The caption reads 'MA TBG' with an extra space; correct to 'MATBG'.
  3. [Figure 1h caption] The caption notes the curve was acquired during a different cooldown and absolute values differ. Please state the scale or normalize to dark resistance so readers can compare with Fig. 1g without confusion.
  4. [Supplementary Note 8, Eq. S14] The expression for NEP_JN^F contains a ratio with V*/I* and dV/dI. The notation (dV/dI)* vs V*/I* is not defined explicitly; define these operating-point quantities.
  5. [Data Availability] The statement says source data for Figs. 1–3 are provided, but not for the supplementary figures. If the main quantitative claims (e.g., DC-heating calibration, noise thermometry) are shown in supplementary figures, please include those source data as well.

Circularity Check

1 steps flagged

One calibration step is self-referential at the operating point, but the central mechanism has independent cross-checks; no load-bearing circularity.

specific steps
  1. fitted input called prediction [Supplementary Note 5 ('Comparison of photoresistance with DC heating'); main text Fig. 3a]
    "The conversion between incident THz power and effective absorbed power was determined by matching the R_THz(ρ) with Δ(dV/dI)(P_abs) for ν=2. This calibration was performed at the base temperature and B=0, and provides an independent conversion between the incident THz power density and the effective absorbed power responsible for electronic overheating."

    At the calibration point (ν=2, T0=1.7 K, B=0), the 'close correspondence' between R_THz(ρ) and Δ(dV/dI)(P_abs) is enforced by construction: the ρ→P_abs conversion is defined by that match. The main text's claim that this correspondence demonstrates that both mechanisms act via the same T_e is thus true at ν=2 by definition, and the headline R_V^abs=(1.7±0.4)×10^7 V/W, quoted at the same operating point, is the DC-calibrated electrothermal response rescaled by the matched P_abs, not an independent prediction. Partial circularity: the fixed conversion factor is tested at other ν, B, T (Supp. Fig. 8), and noise thermometry (Supp. Note 10) independently shows electronic heating.

full rationale

The paper's central claim—that THz photons selectively heat the electronic subsystem and melt the correlated insulator—is supported by multiple independent lines of evidence: the broadband, spectrally flat photoresponse; the polarization independence pointing to direct absorption; the Johnson noise thermometry of Supplementary Note 10, which directly measures a T_e rise under THz illumination; and the cross-checks in Supplementary Figure 8 where a single conversion factor fixed at ν=2, T0=1.7 K, B=0 reproduces the DC-heating comparison at other fillings, temperatures, and magnetic fields. The single genuinely self-referential element is the electrical-substitution calibration of absorbed power: matching R_THz(ρ) to Δ(dV/dI)(P_abs) at ν=2 makes the correspondence at that operating point true by construction, and the headline internal responsivity quoted there is consequently the DC-calibrated electrothermal response rather than an independent prediction. This is a standard bolometric calibration, and it does not by itself prove the hot-electron mechanism, but it is not the only evidence. The lattice-temperature bound in Supplementary Note 11 extrapolates G_e-ph/G⊥ from a 200 mK reference device and does not explicitly apply the paper's own δ≈5 scaling of G_e-ph; that is a correctness risk for the selective-heating hierarchy, but it is an extrapolation issue, not circularity. No load-bearing self-citations, imported uniqueness theorems, or ansatz-smuggling via citation were found. Overall, the central claim retains independent content, so the circularity score is modest.

Axiom & Free-Parameter Ledger

5 free parameters · 5 axioms · 0 invented entities

The central claim rests on five fitted or estimated quantities (power conversion, saturation parameters, cooling exponent δ, G_th, and the conductance ratio G_e-ph/G_⊥) and on standard domain assumptions about MATBG's correlated gap and graphene hot-carrier thermalization. No invented physical entities. The most consequential fit is the absorbed-power conversion, because it defines the headline responsivity.

free parameters (5)
  • Incident-to-absorbed power conversion factor = P_abs ~ 100-500 pW at ρ0 = 70 μW/mm²
    Set by matching R_THz(ρ) to Δ(dV/dI)(P_abs) at ν=2, B=0, T0=1.7 K (Supplementary Note 5). The headline internal responsivity 1.7x10^7 V/W and the saturation power P0=100 pW depend on this conversion.
  • Saturation fit parameters V0 and P0 = V0 = 2.1 mV, P0 = 100 pW
    Fit of |V_ph| = V0·P_abs/(P0+P_abs) to the photovoltage saturation curve (Fig. 3e). P0 sets the linear-to-saturation crossover used in the dynamic-range estimate.
  • Electron-phonon cooling exponent δ = δ ≈ 5
    Fitted to the T_e(P) calibration curve and to the noise-thermometry power dependence (Supplementary Notes 8 and 10); used to extrapolate G_th and NEP_TF.
  • Thermal conductance G_th(T0) = 12x10^-12 W/K
    Determined from the slope ΔP/ΔT_e at the lowest temperature (Supplementary Note 8); scaled as (T_e/T0)^(δ-1) to give G_th ≈ 2x10^-9 W/K at the operating point.
  • Electron-phonon to substrate conductance ratio G_e-ph/G_⊥ = ~10^-4
    Order-of-magnitude estimate from comparing direct Joule heating with graphite-heater heating in a separate reference device at T0=200 mK (Supplementary Note 11); underpins the ΔT_L < 100 mK claim.
axioms (5)
  • domain assumption The ν=2 state in MATBG is a fragile correlated insulator with a gap close to 1 meV that closes near 10 K
    Imported from prior STM/transport literature (Refs 8, 33-35); the steep dR/dT that powers the bolometer rests on this.
  • domain assumption The electronic subsystem thermalizes to a Fermi-Dirac distribution with effective temperature T_e on femtosecond timescales
    Standard hot-carrier assumption in graphene (Supplementary Note 9); needed for the hot-electron description.
  • domain assumption Absorbed THz photons heat electrons rather than directly exciting phonons; graphene optical phonons are nonpolar and hBN phonon energies are much larger than the photon energy
    Supplementary Note 9; needed to attribute absorption primarily to the electronic subsystem.
  • domain assumption Wiedemann-Franz heat leakage to contacts is negligible
    Supported by the insensitivity of overheating to magnetic field in the reference device (Supplementary Note 11).
  • domain assumption The fabricated twist angle of 1.01° realizes the magic-angle regime
    Twist angle set during assembly via micromanipulators; no post-fabrication angle confirmation is reported.

pith-pipeline@v1.3.0-alltime-deepseek · 22987 in / 15789 out tokens · 153638 ms · 2026-08-05T00:51:21.486099+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of Correlated Insulator Moir\'e Bolometer." pith.science (2026). https://pith.science/paper/CVALS35X

@misc{pith2026260800488,
  author       = {Pith},
  title        = {Pith review of: Correlated Insulator Moir\'e Bolometer},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CVALS35X}},
  note         = {Machine review of arXiv:2608.00488}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Light incident on an insulator is generally not expected to turn it into a metal without invoking intense ultrafast excitation that leads to transient structural transitions. Here we show that magic-angle twisted bilayer graphene tuned to half filling of the moir\'e band provides a notable exception to this expectation. We find that weak beam of long-wavelength photons, with energies comparable to the flat-band width, selectively heat the low-heat-capacity electronic subsystem, thereby suppressing the correlated gap. This produces a giant resistance change governed not by a persistent photocarrier population, but by the extreme sensitivity of a many-body correlated gap to weak electronic heating. The resulting photon-driven insulator-to-metal transition produces a broadband low-noise photoresponse with voltage responsivity exceeding millivolts per nW of absorbed power. The mechanism is dual to superconducting hot-electron response: radiation-heated electrons suppress a many-body order, but in reverse the correlated insulator melts into a metal, providing robustness to magnetic fields of several tesla and a sharp insulator-to-metal resistive contrast. Our results establish correlated flat-band systems as a platform for ultra-sensitive detection of faint long-wavelength radiation.

Figures

Figures reproduced from arXiv: 2608.00488 by A. Kuksov, A. L. Shilov, A. Principi, D. A. Bandurin, G. N. Goltsman, I. Gayduchenko, I. Iorsh, K. Shein, K. S. Novoselov, K. Watanabe, L. Elesin, M. Kravtsov, M. Lukianov, R. Izmaylov, S. Jana, T. Taniguchi, X. Zhou.

Figure 1
Figure 1. Figure 1: FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

13 extracted references · 13 canonical work pages

  1. [1]

    Extended transport data

    work page

  2. [2]

    Incident power calibration and responsivity estimates

    work page

  3. [3]

    Polarization dependence of a reference device

    work page

  4. [4]

    Polarization independence of photoresistance in MATBG sample

    work page

  5. [5]

    Comparison of photoresistance with DC heating

    work page

  6. [6]

    Photoresistance atν=−2

    work page

  7. [7]

    Measurements of additional MATBG samples

    work page

  8. [8]

    NEP, DR, andG th estimation

    work page

  9. [9]

    THz-induced electron heating and heat transfer balance

    work page

  10. [10]

    Noise thermometry in TBG

    work page

  11. [12]

    Electromagnetic simulations

    work page

  12. [13]

    Correlated insulator vs superconducting bolometry

    work page

  13. [14]

    EXTENDED TRANSPOR T DA T A Supplementary Figure 1.T ransport data in a voltage-bias scheme.a, Longitudinal resistance map as a function of temperature and moir´ e filling factor

    CI phase diagram 3 Supplementary Note 1. EXTENDED TRANSPOR T DA T A Supplementary Figure 1.T ransport data in a voltage-bias scheme.a, Longitudinal resistance map as a function of temperature and moir´ e filling factor. The measurement scheme employs a small AC bias voltage U, which prevents additional electron heating, since in the correlated state the d...

    work page 2020