{"id":"d0be025d-d126-4320-b337-9925387ddd0d","arxiv_id":"2412.13863","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Atomically thin ReS2 devices with graphite contacts show a polarization-dependent zero-bias photocurrent attributed to the intrinsic bulk photovoltaic effect, with about 1 mA/W responsivity.","lead":"Researchers measured a light-induced current in atomically thin ReS2 at zero applied voltage and attribute it to the intrinsic bulk photovoltaic effect. The result suggests a way to study and exploit this effect in other thin non-centrosymmetric van der Waals materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Unverified non-centrosymmetric stacking of the measured ReS2 flakes is the load-bearing assumption; the reported theory also does not match the measured polarization angle at 633 nm.","rationale":"Good-faith reading: the electrical data are internally consistent, and the lateral SPCM geometry is a reasonable approach to separating contact and channel signals; the linear power dependence, low contact resistance, and the absence of channel photocurrent in metallic-contact devices are real supporting observations. However, the BPVE attribution is contingent on a symmetry that the paper assumes rather than demonstrates. The reader's conditional verdict already captures this. My independent recalculation of the 633 nm maximum angle from the paper's own tensor components gives approximately -32 degrees, not within the claimed [-12, +2] degree range, which means the wording 'supported by theoretical calculations' is stronger than the numbers justify. Since neither the stacking assumption nor the angle discrepancy is disproved by the existing data, and both could be resolved by direct SHG verification of the measured flakes plus a corrected theory comparison, the appropriate outcome remains conditional acceptance pending the proposed symmetry check.","tokens_in":15766,"tokens_out":7695,"duration_ms":70495,"concrete_test":"Perform polarization-resolved second-harmonic generation (SHG) directly on device A (the bilayer flake measured in Fig. 2d) and, if possible, on device B, using a subgap fundamental wavelength such as 1064 nm so the SHG response reports structural symmetry rather than the same photocurrent process. A non-centrosymmetric A' bilayer should show a strong, anisotropic SHG polar pattern; a centrosymmetric AB bilayer should show no dipolar SHG or only a much weaker quadrupolar response. Comparing the SHG principal axes with the Raman-determined b-axis and with the photocurrent polarization maximum will settle whether the flakes are actually non-centrosymmetric in the measured configuration. Independently, recompute theta0 from Eq. 3 with the tabulated sigma components to check whether the stated [-12, +2] degree range is correct.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the channel photocurrent is the intrinsic bulk photovoltaic effect of ReS2 requires the measured exfoliated flakes to adopt the non-centrosymmetric A' (180-degree-rotated) bilayer stacking used in the calculation. This is stated in the main text (paragraph beginning 'While bulk ReS2 is centrosymmetric') and in SI Methods section 1, but no SHG, ferroelectric switching, or structural measurement is presented for the actual devices. Bulk ReS2 is centrosymmetric, and exfoliation from a bulk crystal will generally reproduce the centrosymmetric parent stacking unless a specific polytype is selected or verified; the cited proofs of non-centrosymmetry in ReS2 (Refs. 38, 39) were obtained on material that was explicitly characterized for that property. If the device flakes are centrosymmetric, the zero-bias, polarization-dependent channel signal cannot be the intrinsic shift-current BPVE of an A' bilayer, and the ab initio comparison is irrelevant. This concern is compounded by an internal inconsistency in the claimed theory-experiment agreement: using the reported components at 633 nm (sigma_xxx = 2.68, sigma_xyy = -1.03, sigma_xxy = -3.71 uA/V^2), Eq. 3 gives theta0 = 0.5*arctan(2*sigma_xxy/(sigma_xxx - sigma_xyy)) approximately -32 degrees, far outside the stated [-12, +2] degree range; thus the calculation does not actually reproduce the measured polarization response (maximum at the b-axis), making it a weak basis for inferring the symmetry of the measured flakes.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports fabrication of lateral, hBN-encapsulated ReS2 devices with graphite contacts, in which a zero-bias photocurrent is observed in the pristine channel under 633 nm and 532 nm illumination. The authors attribute this signal to the intrinsic bulk photovoltaic effect (BPVE) on the basis of linear power scaling, polarization dependence described by a second-order conductivity tensor, and scanning photocurrent microscopy showing channel-separated photocurrent. Ab initio shift-current calculations for a non-centrosymmetric (A', 180°-rotated) bilayer ReS2 give nonlinear conductivities of the same order of magnitude as the experimental fit, and the paper claims the calculated maximum-photocurrent angle lies within [-12°, +2°] of the b-axis.","tokens_in":16046,"tokens_out":5310,"duration_ms":45415,"significance":"If the interpretation holds, the work would be a valuable demonstration of an intrinsic BPVE in a lateral vdW device without strain or vertical geometry, with a clean contact scheme and an independent first-principles calculation. The paper includes detailed fabrication and characterization, and the shift-current calculation is independent of the measured data, which is a strength. However, the central claim depends on the measured flakes adopting the non-centrosymmetric bilayer stacking assumed in the calculation, and on the theory reproducing the polarization dependence; both assumptions currently need attention.","major_comments":[{"comment":"The non-centrosymmetric A′ stacking is assumed for the measured flakes, but no SHG, ferroelectric switching, or structural measurement is provided for the actual devices. Since the bulk parent crystal is centrosymmetric, exfoliation need not yield the A′ polytype, and the cited Refs. 38 and 39 characterized material that was specifically selected for those properties. Without direct verification that the measured bilayer and four-layer flakes are the A′ stacking, the assignment of the channel photocurrent to the intrinsic BPVE of that structure is not fully justified.","section":"Main text, 'While bulk ReS2 is centrosymmetric' paragraph; SI Methods §1"},{"comment":"Using the reported 633 nm tensor components (σ_xxx = 2.68, σ_xyy = −1.03, σ_xxy = −3.71 µA/V²) in the paper's own Eq. (3) gives θ0 = 0.5 arctan(2σ_xxy/(σ_xxx − σ_xyy)) ≈ −32°, which is outside the stated range [-12°, +2°] for the angle of maximum photocurrent. Thus the ab initio calculation does not reproduce the measured polarization response, whose maximum lies near the b-axis. The internal inconsistency in the claimed theory–experiment agreement needs to be resolved, either by correcting the tensor components or by re-evaluating the comparison.","section":"Main text, Fig. 4 and surrounding paragraph"},{"comment":"The ab initio calculation is performed for a bilayer, but the SPCM device (device B) is a four-layer flake. The text does not explain how the A′ stacking and the calculated shift-current tensors for the bilayer apply to a four-layer flake, which may have different stacking energetics. The comparison between the two devices and the theory should be made explicit.","section":"Main text, Fig. 3 and thickness determination (SI §2)"}],"minor_comments":[{"comment":"The figure caption states an optical power of 120 mW, but the text and SI §4 consistently use 120 µW; the units should be corrected.","section":"Fig. 2d caption"},{"comment":"The fitted conductivity values are given in mA/V² in the main text, while the SI reports the same quantities in µA/V²; the units should be made consistent.","section":"Main text near Eq. (3)"},{"comment":"The definition κ = <I_SC>/(r×t) divided by P/(π r²) uses an effective area r×t rather than the beam area; the geometric model should be clarified, and the sensitivity of κ to the assumed beam radius and refractive index should be stated.","section":"Main text, intrinsic responsivity definition"},{"comment":"Equation (4) introduces the interband dipole matrix and generalized derivative notation without explicit definitions; a brief explanation of the symbols would improve readability for non-specialists.","section":"Equation (4)"}],"recommendation":"major_revision","confidential_remarks":"The paper is potentially publishable after major revision. The unverified stacking assumption is a serious concern, but it is testable with SHG or ferroelectric switching on the same devices. The internal inconsistency in the angle prediction is fixable by recomputation or by correcting the reported tensor values. I would not reject at this stage, but the authors need to address both points convincingly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Hi — quick take on the ReS2 BPVE paper.\n\nThe genuinely new thing here is the device platform: a lateral, hBN-encapsulated ReS2 channel with graphite contacts, where a zero-bias photocurrent appears in the pristine channel, spatially separated from contact signals in SPCM, with linear power scaling and a clean polarization dependence peaking near the b-axis. That is a real experimental step beyond earlier ReS2 BPVE work at grain boundaries, engineered edges, and vertical stacks. The ~1 mA/W intrinsic responsivity is modest but consistent with the field. The device engineering (graphite contacts, encapsulation, pick-up assembly) looks careful, and the control with Ti/Au contacts showing no channel signal supports the claim that contact engineering matters.\n\nThe soft spots are real, and one of them is a straightforward internal inconsistency. The authors claim the ab initio calculation predicts a photocurrent maximum within [-12°, +2°] of the b-axis for all energies. But plugging their reported 633 nm components—σ_xxx=2.68, σ_xyy=-1.03, σ_xxy=-3.71 µA/V²—into their own Eq. 3 gives θ0 = 0.5 arctan(2σ_xxy/(σ_xxx−σ_xyy)) ≈ -32°. That is far outside the claimed range, and it is also far from the experimental maximum at ~0°. So the theoretical calculation, as presented, does not reproduce the measured polarization dependence. This is not a matter of opinion; it is arithmetic. Either the printed components, the angle plot, or the text is wrong. A referee will catch this immediately.\n\nThe second soft spot is the load-bearing assumption that the exfoliated flakes adopt the non-centrosymmetric (180°-rotated) bilayer stacking used in the calculation. The paper cites prior SHG and ferroelectricity work in few-layer ReS2, but does not verify the stacking of the actual devices. No SHG, no ferroelectric switching, no structural probe. Bulk ReS2 is centrosymmetric, and exfoliation from a bulk crystal typically reproduces the parent stacking. If these flakes are the common centrosymmetric phase, the channel photocurrent needs a different mechanism, and the whole theory comparison is moot.\n\nMinor issues: units typo in main text (mA/V² vs µA/V²), no explicit thermal artifact checks (though the polarization dependence somewhat mitigates), and no raw data or code. The experimental core is plausible and the platform is useful, but the paper currently overstates its theoretical support and leaves its central symmetry assumption unverified.\n\nVerdict: worth sending to a serious referee, but it should come back with major revision requests—verify the stacking (or at least discuss the uncertainty), correct or explain the angle calculation, and clean up the units. This is a conditional accept, not a desk reject.","headline":"A clean lateral ReS2 platform that likely shows a bulk photovoltaic signal, but the paper's own numbers contradict its claimed theory-experiment agreement and the non-centrosymmetric stacking is never verified.","tokens_in":16687,"tokens_out":6328,"would_cite":true,"duration_ms":53341,"reading_group":"maybe","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 atomically thin ReS2 flakes with graphite contacts and hBN encapsulation produce a zero-bias, polarization-dependent photocurrent in the pristine channel, which the authors attribute to the intrinsic bulk…","keywords":["bulk photovoltaic effect","intrinsic photocurrent","second-order conductivity","shift current","ReS2","broken inversion symmetry","van der Waals heterostructures","scanning photocurrent microscopy"],"falsifier":"A scanning second-harmonic generation map of the exact devices from Figs. 2 and 3 that shows no signal in the channel would remove the symmetry basis for the intrinsic-bulk-photovoltaic assignment; conversely, measuring a comparable channel photocurrent in a flake verified to be centrosymmetric would indicate that the current has a different origin.","tokens_in":15549,"feed_emoji":"🔆","tokens_out":15410,"duration_ms":122417,"temperature":0.7,"pith_summary":"The paper reports that atomically thin flakes of the semiconductor ReS2, connected to graphite electrodes and encapsulated in hexagonal boron nitride, produce a photocurrent at zero applied voltage when illuminated with visible light. The signal is measured in the pristine middle of the channel, away from the contacts, and its magnitude depends on the linear polarization of the light, which is the characteristic fingerprint of the bulk photovoltaic effect. The authors extract the in-plane second-order conductivity tensor from the polarization dependence and compare it with ab initio shift-current calculations for a non-centrosymmetric bilayer stacking, finding the same order of magnitude. If the claim holds, a single unstrained few-layer van der Waals material can convert light into current through a nonlinear symmetry-driven mechanism, and the lateral device geometry provides a clean way to separate that intrinsic response from interface photocurrents.","feed_headline":"ReS2 flakes convert light to current at zero bias, 1 mA/W","feed_subtitle":"Graphite contacts and hBN encapsulation expose the intrinsic effect, backed by shift-current calculations.","key_machinery":"The load-bearing object is the second-order conductivity tensor $\\sigma^{(2)}_{ijk}$, defined by $j_i=\\sigma^{(2)}_{ijk}E_jE_k$, which is nonzero only when inversion symmetry is broken. In the experiment only the in-plane components along the b-axis matter, so the polarization-angle fit uses $I_x = t r [\\sigma^{(2)}_{xxx}\\cos^2\\theta + \\sigma^{(2)}_{xyy}\\sin^2\\theta + 2\\sigma^{(2)}_{xxy}\\cos\\theta\\sin\\theta]$, with $t$ the flake thickness and $r$ the beam radius. The same tensor is computed from first principles for the non-centrosymmetric bilayer stacking in which the top monolayer is rotated 180 degrees relative to the bottom one. Microscopically, the calculation evaluates the shift-current contribution, where the coherent displacement of photoexcited carriers produces the DC current. The device architecture, graphite electrodes plus hBN encapsulation in a lateral geometry, is what lets the tensor be measured in the pristine channel with a roughly one-micron laser spot, separating it from the large Schottky photocurrents that dominate metal-contacted devices.","core_discovery":"The central claim is that a lateral ReS2 device with graphite contacts and hBN encapsulation shows an intrinsic bulk photovoltaic effect: at zero bias, focused illumination of the channel center produces a short-circuit current along the b-axis that grows linearly with optical power and is modulated by, but never extinguished as, the linear polarization is rotated. From the angular data the paper fits the second-order conductivity components $\\sigma_{xxx}^{(2)}=5.51\\,\\mu\\mathrm{A}/\\mathrm{V}^2$, $\\sigma_{xyy}^{(2)}=3.06\\,\\mu\\mathrm{A}/\\mathrm{V}^2$, and $\\sigma_{xxy}^{(2)}=-0.05\\,\\mu\\mathrm{A}/\\mathrm{V}^2$, and reports intrinsic responsivities of about 1.3 mA/W for a bilayer device and 1.6 mA/W at 633 nm and 1.0 mA/W at 532 nm for a four-layer device. Ab initio shift-current calculations for the 180-degree-rotated non-centrosymmetric bilayer give $\\sigma^{(2)}$ components of the same order of magnitude and predict the photocurrent maximum within $[-12^\\circ,+2^\\circ]$ of the b-axis. The paper concludes that the channel photocurrent is an intrinsic bulk photovoltaic response, not a Schottky-barrier or edge effect.","pith_inferences":["A decisive check would be to verify the stacking polarity of the measured devices by second-harmonic generation or ferroelectric switching, since the paper does not directly confirm the non-centrosymmetric stacking in the devices it reports.","The fitted tensor components give a polarization-dependent angular profile that could serve as an in-situ crystallographic alignment tool, because the photocurrent maximum encodes the b-axis orientation.","The same measurement on flakes of different thickness and stacking polytypes would map how the bulk photovoltaic response evolves with symmetry, connecting the bilayer calculation to the four-layer result.","If the observed signal instead came from a buried interface or strain at the hBN boundaries, devices with different encapsulation materials should show a different magnitude, a testable extension the paper does not report."],"forward_implications":["Zero-bias bulk photovoltaic currents can be observed in unstrained, few-layer ReS2 without vertical device stacks or external tuning knobs.","The order-of-magnitude agreement between measured and calculated $\\sigma^{(2)}$ supports assigning the channel photocurrent to the non-centrosymmetric bilayer stacking rather than to interface effects.","Because the channel photocurrent is nonzero at every polarization angle, the response is a direct electrical signature of the polar character of few-layer ReS2, matching the earlier observation of nonzero second-harmonic generation for all angles.","Graphite contacts suppress the interface photocurrent relative to Ti/Au contacts, making the intrinsic channel response visible and cleanly separable in scanning photocurrent maps.","The same fabrication and measurement protocol should apply to other non-centrosymmetric van der Waals materials without a perpendicular two-fold rotation axis, where in-plane bulk photovoltaic currents are generically allowed at normal incidence."],"supporting_citations":[{"why":"Documents non-vanishing second-harmonic generation for all polarization angles in ReS2, the symmetry precedent for a nonzero mean photocurrent.","marker":"(38)"},{"why":"Identifies the 180-degree-rotated non-centrosymmetric bilayer stacking of ReS2 with ferroelectricity, the structure used in the ab initio calculation.","marker":"(39)"},{"why":"Reports zero-bias photocurrent only at grain boundaries in polycrystalline ReS2, the prior observation the present work extends to the pristine channel.","marker":"(40)"},{"why":"Reports a vertical-device bulk photovoltaic effect in single-crystalline ReS2, the reference point for the lateral-geometry improvement and responsivity comparison.","marker":"(42)"},{"why":"Reports negligible bulk photovoltaic current in unstrained monolayer 2H-MoS2, a comparative baseline for the ReS2 result.","marker":"(43)"},{"why":"Shows that lateral 3R-MoS2 devices need strain for a sizeable bulk piezophotovoltaic effect, the comparison baseline for the unstrained ReS2 response.","marker":"(44)"},{"why":"Demonstrates a 180-degree-twisted bilayer of the sister compound ReSe2 as a non-centrosymmetric semiconductor, supporting the rotated-stacking symmetry mechanism.","marker":"(48)"},{"why":"Supplies the interpolation method used to compute the shift-current conductivity from first principles.","marker":"(49)"}],"fun_headline_variants":["Zero-bias photocurrent in atomically thin ReS2 is intrinsic","Intrinsic BPVE in ReS2: 1 mA/W without strain or bias","ReS2 flakes turn light into current at zero bias, no tuning","Clean ReS2 devices reveal intrinsic bulk photovoltaic effect","Atomically thin ReS2: intrinsic photocurrent at 1 mA/W"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim rests on the measured flakes having the non-centrosymmetric 180-degree-rotated bilayer stacking used in the calculation; the paper does not directly verify this stacking in the devices it measures, and a centrosymmetric flake would not produce the claimed intrinsic bulk photovoltaic effect.","fun_headline_variants_meta":{"raw":{"variants":["Zero-bias photocurrent in atomically thin ReS2 is intrinsic","Intrinsic BPVE in ReS2: 1 mA/W without strain or bias","ReS2 flakes turn light into current at zero bias, no tuning","Clean ReS2 devices reveal intrinsic bulk photovoltaic effect","Atomically thin ReS2: intrinsic photocurrent at 1 mA/W"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000323,"raw_usage":{"total_tokens":1818,"prompt_tokens":956,"completion_tokens":862,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":572,"completion_tokens_details":{"reasoning_tokens":763}},"tokens_in":572,"tokens_out":862,"duration_ms":7709,"temperature":1.0,"reasoning_tokens":763,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T12:42:50.059101+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A scanning second-harmonic generation map of the exact devices from Figs. 2 and 3 that shows no signal in the channel would remove the symmetry basis for the intrinsic-bulk-photovoltaic assignment; conversely, measuring a comparable channel photocurrent in a flake verified to be centrosymmetric would indicate that the current has a different origin.","supporting_citations":[],"review_version":1}