{"id":"1d8be588-86c0-4489-822a-c96133d0dc34","arxiv_id":"2608.09300","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A graphene/3R-MoS2 hybrid photodetector shows persistence-free, temperature-insensitive photoresponse via photoinduced modulation of ferroelectric polarization.","lead":"This paper builds a photodetector from bilayer graphene on a ferroelectric form of bilayer MoS2 and shows it can detect repeated light pulses without the slow persistence that usually plagues such hybrids. The result points to a way to combine high sensitivity with fast, repeatable optical detection in atomically thin devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central mechanism—photoinduced polarization modulation in 3R-MoS2—is inferred indirectly; a control experiment and a quantitative Dirac-point-shift check would establish it.","rationale":"The reader's weakest assumption matches the central risk: the paper attributes the persistence-free, temperature-independent photoresponse to photoinduced polarization modulation in 3R-MoS2, but this is an inference rather than a direct observation. I agree that this is the load-bearing premise. Independent support includes the KPFM observation of polarization domains, the PL/Raman confirmation of 3R stacking, the asymmetric R_Dirac vs D, and the sign analysis connecting ΔR_PH to a gate-voltage change; however, none of these uniquely proves that illumination changes the polarization in the operating device. The alternative—reduced trap density and altered band alignment in 3R-MoS2—is plausible and not excluded. The paper itself notes uncertainties in absorption and optical transmission, and the close comparison between ΔV_op-m and ΔV_bg-A is only order-of-magnitude, so the quantitative case is not airtight. I therefore see no internal inconsistency that would reject the claim, but the evidence is insufficient for full acceptance without a decisive test. The proposed control experiment with 2H-MoS2 and the Dirac-point-shift measurement would directly address the concern. My recommendation is unchanged from the reader's CONDITIONAL verdict.","tokens_in":12387,"tokens_out":7135,"duration_ms":76206,"concrete_test":"Fabricate an otherwise identical control device with bilayer 2H-MoS2 and repeat the photoresponse measurements at 100 K over the same Vbg/Vtg range. If the 2H control shows persistent photoconductivity and the conventional sign of ΔR_PH, the 3R polarization mechanism is supported; if it also shows persistence-free response and an anomalous sign, the claim fails. In parallel, measure the shift of the Dirac point (minimum of R vs Vbg) under continuous illumination as a function of intensity, and compare with the shift predicted by the polarization-screening model using known absorption and gate capacitances. A quantitative match over at least one decade of intensity, with the correct sign, would directly substantiate the photoinduced polarization modulation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim of Section IV is that illumination lowers the out-of-plane polarization of 3R-MoS2, thereby reducing the interfacial field E_i,P and producing an optically controlled gating effect on the bilayer graphene channel. This mechanism is not directly observed in the device; it is inferred from the sign of the photoinduced resistance change relative to dR/dV_bg and from the order-of-magnitude agreement between ΔV_op-m and ΔV_bg-A. Because 3R stacking also changes the band alignment and, as the authors note (Refs. [19,36,37]), reduces the density of interfacial sulfur vacancies, the persistence-free temperature-independent response could plausibly be explained by a lower trap density and a different charge-transfer pathway rather than by dynamic polarization modulation. The slow (600 ms) component of the bi-exponential decay, and its attribution to deep traps in Section III, further suggests that trap processes are present even in the hole-doped regime, complicating the assignment. Thus the design principle—polarization as an optically controllable gate—is not uniquely established by the data.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a dual-gated bilayer graphene (BLG) / bilayer 3R-MoS2 van der Waals heterostructure photodetector and claims that photoinduced modulation of the spontaneous out-of-plane polarization in 3R-MoS2 produces an optically controlled gating effect on the BLG channel. The authors report a persistence-free photoresponse in the hole-doped regime, response times of tens of milliseconds, a gain-bandwidth product near 10^8 Hz, and a minimum detectable photon number of about 31 per pulse. The central mechanism is proposed in Section IV: illumination lowers the polarization of 3R-MoS2, reduces the interfacial field, changes the effective displacement field, and thereby shifts the Fermi level in BLG. The paper is primarily an experimental study with transport and optoelectronic characterization, supported by KPFM, PL, and Raman identification of the 3R phase.","tokens_in":12574,"tokens_out":7748,"duration_ms":79542,"significance":"If the proposed mechanism is correct, the work would be significant because it offers a route to decouple sensitivity from persistence in graphene/TMD photodetectors, a long-standing trade-off in trap-mediated photogating devices. The experimental dataset is rich: the authors show doping-dependent photoresponse asymmetry, power- and temperature-dependent timescales, and a systematic displacement-field tuning of the response. The device performance metrics, particularly the persistence-free operation at low temperature and the reported photon-counting capability, are noteworthy. However, the central mechanistic claim is not directly evidenced in the manuscript: no measurement under illumination directly probes the polarization state of 3R-MoS2, and the supporting quantitative comparison (Section IV) relies on quantities derived from the photoresponse itself. The paper would be substantially strengthened by a direct polarization probe or a control device without the ferroelectric layer.","major_comments":[{"comment":"The quantitative support for the polarization-modulation mechanism is not independent. The quantity ΔV_bg-A is obtained by dividing the measured resistance change R_A by dR/dVbg, so it is a rescaling of the photoresponse; the comparison with ΔV_op-m, which depends on the absorption coefficient and top-gate transmission (uncertainties acknowledged in the same section), therefore tests only internal consistency of the model. An order-of-magnitude match between two quantities that both derive from the same photoresponse does not uniquely establish a photoinduced change in polarization. A direct probe (e.g., KPFM under illumination, or polarization-sensitive optical measurement) or a control experiment with a 2H-MoS2 device of comparable interface quality is needed.","section":"Section IV"},{"comment":"Alternative non-ferroelectric explanations are not excluded. In Section IV the authors themselves state that there is \"a lower density of localized trap states at the Gr/TMD interface in bilayer configuration\" for 3R stacking, which could explain the persistence-free response through reduced trapping without invoking polarization modulation. Moreover, the bi-exponential decay in the hole-doped regime (Section III, Eq. 1) includes a 600 ms component attributed to deep traps, showing that trap processes are still active. The claim that the persistence-free response is governed by polarization modulation therefore needs a control measurement that separates the trap-density effect from the polarization effect, for example a 2H-MoS2 device fabricated and measured under identical conditions.","section":"Sections III and IV"},{"comment":"Eq. (1) is written as a rising exponential, ∆R = R_A(1−e^{(t−t0)/τ_A}) + R_B(1−e^{(t−t0)/τ_B}), but the text uses it to describe the decay after the pulse; this appears to be a sign error that prevents the reader from checking the fit. In addition, the statement that the rise time of 9.5±2 ms is \"instrument-limited\" is unsupported because no instrument response function or bandwidth characterization is provided. If the timescale is indeed instrument-limited, the observed flat power and temperature dependence of τ_A (Fig. 3c–e) cannot be used to argue for a thermally independent mechanism.","section":"Section III, Eq. (1) and GBW discussion"},{"comment":"The photon-number calibration uses ΔR_1e = dR/dVbg × e/Cbg, which assumes that the measured resistance change is equivalent to a change in gate voltage produced by a single electron. This equivalence is exactly the hypothesis under investigation, so the extracted values of 31 photons and IQE = 10% inherit model dependence. The reported uncertainty of ±10 photons from the Poisson fit does not include systematic errors from the absorption coefficient and top-gate transmission, which are acknowledged to be uncertain in Section IV. The photon-counting claim should be backed by an independent calibration or a stated error budget.","section":"Section III, Eqs. (2)-(3)"},{"comment":"The sign interpretation is internally inconsistent. In Section III, the sign of R_PH is said to suggest \"net transfer of electrons from BLG to the TMD layer,\" which is a charge-transfer picture. In Section IV, the same sign is attributed to a photoinduced reduction in polarization that changes the effective displacement field. Both mechanisms can produce the same resistance sign in this device geometry, so the sign alone does not discriminate between them. The paper should either provide a measurement that separates charge transfer from field gating (for example, a Hall measurement) or explicitly state that the sign is not a fingerprint of the polarization mechanism.","section":"Sections III and IV"}],"minor_comments":[{"comment":"The KPFM description says \"amplitude of 11 nm\" and \"AC bias of 2 V\"; the amplitude unit should be checked because KPFM amplitude is usually reported in volts or nanometers depending on the mode, and the current phrasing is ambiguous.","section":"Section II"},{"comment":"In the PL spectra description, \"neural\" should be \"neutral\" (A exciton).","section":"Section III"},{"comment":"The KPFM text reports an interlayer potential difference ∆V_KPFM of approximately 70 mV, while the transport section states the minimum resistance at CNP occurs at D = -0.058 V/nm and \"closely matches the interlayer potential of ≈55 meV.\" The relation between these two numbers (mV vs meV) should be clarified.","section":"Section III, Fig. 1c"},{"comment":"The caption lists two panels labeled (c) and uses (d) twice; this makes it difficult to follow which panel shows inverse rise time, fall time, and the extracted τ_A and τ_B. Please relabel the panels.","section":"Section III, Fig. 3 caption"},{"comment":"The gain-bandwidth product is stated as approximately 10^8 Hz in the abstract but as approximately 5 × 10^8 Hz in Section III; please make the numbers consistent or explain the difference.","section":"Abstract and Section III"},{"comment":"The sentence \"We ensure device operation in a regime where Vtg > 0 and Vbg is < 0, such that D is consistently directed downward\" is in Section IV, not Section III, so the cross-reference in the text should be corrected.","section":"Section III"}],"recommendation":"major_revision","confidential_remarks":"The experimental work is careful and the device performance is interesting, but the central mechanistic claim is not yet established with the presented evidence. I would consider acceptance if the authors add a direct probe of polarization modulation (e.g., KPFM under illumination) or a convincing control device without the ferroelectric 3R-MoS2 layer and with comparable trap density. The current manuscript is well within the journal's scope, but the mechanistic attribution needs strengthening before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear Colleague,\n\nThe headline: this paper reports a persistence-free photoresponse in a bilayer graphene/3R-MoS2 hybrid, with recovery times around 20 ms, and attributes it to photoinduced modulation of the sliding ferroelectric polarization. That is a genuinely new application of 3R-MoS2 and a real departure from the trap-mediated photogating story in Gr/TMD devices.\n\nWhat it does well: the 3R phase is well characterized (KPFM domain contrast, PL red shift, layer-breathing/shear Raman). The electrical data are internally consistent — sign reversal of the photoresponse with doping, persistence only in an electron-doped window, and a fast timescale that is nearly independent of power and temperature. The paper is honest about its limitations, noting the instrument-limited response and the slow 600 ms component.\n\nSoft spots: the central mechanism is inferred, not directly measured. The comparison between ΔV_op-m and ΔV_bg-A is partly circular because both derive from the photoresponse. The response speed is instrument-limited, so the claim is 'at least as fast as our setup,' not an intrinsic speed. The persistence-free regime still shows a small trap-related component, which complicates the clean separation from conventional photogating. And the photon-counting/IQE numbers carry uncertainties that are not fully quantified.\n\nOverall, this is a credible experimental report of a plausible design principle. It does not fully prove the mechanism, but the evidence is coherent and prior work on photoinduced polarization changes in sliding ferroelectrics gives it independent support. A control device with 2H-MoS2 or a direct measurement of the Dirac point shift under illumination would significantly strengthen the case.\n\nFor peer review: send it out. An expert referee can push for the control experiment and more rigorous error analysis, but this paper deserves referee time.\n\nBest,","headline":"A genuinely new application of sliding ferroelectricity to photodetection, with a plausible but not fully proven mechanism; deserves serious refereeing.","tokens_in":13168,"tokens_out":2620,"would_cite":true,"duration_ms":25676,"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":"This paper claims that in a bilayer graphene/bilayer 3R-MoS2 hybrid, illumination reduces the ferroelectric polarization of 3R-MoS2, creating an optically controlled gate that makes the photoresponse persistence-free and millisecond-fast.","keywords":["emergent ferroelectricity","rhombohedral stacking","sliding ferroelectricity","bilayer graphene","3R-MoS2","persistence-free photoresponse","photogating","van der Waals photodetector"],"falsifier":"Measure the out-of-plane polarization of the 3R-MoS2 bilayer in the operating device, for example with Kelvin probe force microscopy or optical second-harmonic generation, before, during, and after a 532 nm pulse at 100 K with the same gate voltages. If the polarization does not change while the fast, temperature-independent millisecond photoresponse remains, the proposed polarization-modulation mechanism is ruled out.","tokens_in":12176,"feed_emoji":"💡","tokens_out":9003,"duration_ms":80089,"temperature":0.7,"pith_summary":"This paper argues that the usual speed-versus-sensitivity trade-off in graphene/transition-metal-dichalcogenide photodetectors can be broken by using a ferroelectric form of molybdenum disulfide, 3R-MoS2. In a hybrid of bilayer graphene and bilayer 3R-MoS2, light lowers the spontaneous out-of-plane polarization of the ferroelectric, which acts as an optically controlled gate that changes the electrostatic environment of the graphene channel. The result is a persistence-free photoresponse in the hole-doped regime: the signal returns to baseline in roughly 20 ms rather than persisting for minutes, and the response time does not depend on temperature between 100 K and 165 K. If the mechanism is right, it converts a slow, trap-dominated detector into one that resolves repeated low-intensity pulses, with measured detection down to about 31 photons per pulse and a gain-bandwidth product near $10^{8}$ Hz.","feed_headline":"31-photon pulses detected with zero photodetector persistence","feed_subtitle":"A polarization switch in 3R-MoS2, not trap states, drives a millisecond-scale response at 100 K.","key_machinery":"The central object is sliding ferroelectricity in rhombohedrally stacked bilayer 3R-MoS2, where non-centrosymmetric AB/BA stacking creates a spontaneous out-of-plane polarization that persists at room temperature. The load-bearing identity is the polarization-modulation cycle: illumination dopes the MoS2, lowers that polarization, reduces the interfacial field $E_{i,P}$ at the BLG/MoS2 junction, and thereby boosts the effective displacement field $D$ seen by the bilayer graphene. Layer polarization under $D$ then routes the photoresponse: holes concentrate in the top graphene layer, away from the interface, giving a persistence-free response, while electrons flow in the bottom layer, closer to the interface, giving partial persistence.","core_discovery":"The central claim is that photoinduced doping lowers the polarization in 3R-MoS2, reducing the interfacial field at the bilayer graphene/MoS2 junction, so that the effective displacement field on the graphene increases and shifts its Fermi level; this optically controlled gating produces a temperature-independent photoresponse with response times on the order of tens of milliseconds, limited by the measurement instrument. In the hole-doped regime the response is persistence-free, with a rise time of 9.5±2 ms and fall components of 23±5.8 ms and 600±37.8 ms, whereas the electron-doped regime shows partial persistence because electrons transported in the bottom graphene layer sit closer to the interface and can be trapped. The paper reports a photogain of $3.1\\times10^{7}$, a gain-bandwidth product of about $5\\times10^{8}$ Hz, an internal quantum efficiency of about 10%, and a minimum detectable photon number of 31 in single-shot measurements, with the same device switchable to a non-volatile optical memory mode by tuning the Fermi level.","pith_inferences":["If the polarization-modulation mechanism is generic, the same persistence-free behavior should appear in graphene paired with other sliding ferroelectrics, such as 3R-WS2, 3R-WSe2, or twisted TMD bilayers, turning this from a single-device result into a family of detectors.","A direct in-operando measurement of the 3R-MoS2 polarization under illumination would close the inference gap the authors leave open, since their evidence for the mechanism is transport-based.","The electron-doped persistent mode, treated here as a side effect, could be developed into a single-pulse-write optical memory with the same device.","Engineering the ferroelectric's polarization magnitude, rather than passivating traps, is the implied lever for pushing response times below the current instrument limit."],"forward_implications":["Persistence-free, temperature-independent photoresponse should be reproducible in repeated low-intensity optical pulses in the hole-doped regime, operating as a linear detector.","Displacement field $D$ becomes a tunable knob: increasing $D$ amplifies responsivity, allowing detector performance to be optimized independently of material choice.","Because the fall time is instrument-limited rather than thermally activated, the measured bandwidth is a lower bound; faster electronics should reveal the intrinsic limit.","Tuning the Fermi level between hole- and electron-doped regimes switches the same device between repeatable detection and non-volatile optical memory.","A gain-bandwidth product near $10^{8}$ Hz together with 31-photon detection establishes a benchmark for low-light all-2D photodetectors."],"supporting_citations":[{"why":"Establishes interfacial ferroelectricity in rhombohedral-stacked bilayer TMDs, the foundation for the 3R-MoS2 polarization.","marker":"[10]"},{"why":"Shows interfacial ferroelectricity persists in 2D semiconductors, supporting room-temperature spontaneous polarization.","marker":"[11]"},{"why":"Demonstrates spontaneous-polarization-induced photovoltaic effect in 3R-MoS2, linking the polarization to photocarrier response.","marker":"[15]"},{"why":"Supplies the asymmetric interlayer coupling and type-II band alignment used to explain layer-selective carrier confinement.","marker":"[16]"},{"why":"Reports cumulative polarization in conductive interfacial ferroelectrics, supporting photoinduced polarization change under doping.","marker":"[24]"},{"why":"Predicts large photoinduced tuning of ferroelectricity in sliding ferroelectrics, the specific mechanism behind the optically controlled gate.","marker":"[35]"},{"why":"Defines the trap-mediated photogating and persistent photoconductivity baseline the paper sets out to overcome.","marker":"[2]"},{"why":"Provides the Gr/TMD power- and temperature-dependent timing comparison that distinguishes the new mechanism.","marker":"[5]"},{"why":"Argues defect-mediated carrier trapping in sliding ferroelectrics, supporting the lower interfacial trap density claim.","marker":"[19]"}],"fun_headline_variants":["Polarization switch gives fast, sensitive 2D photodetector","Trap-free photoresponse from ferroelectric MoS2 allows 31-photon readout","Millisecond response, no persistence: 31-photon detection in 2D hybrid","Ferroelectric 3R-MoS2 enables trap-free detection of 31 photons","Trap-free 2D photodetector sees down to 31 photons"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"That light actually changes the built-in electric polarization of the 3R-stacked MoS2 layer, and that this polarization change is what makes the photoresponse fast and persistence-free; the paper infers this from transport data and earlier literature rather than from a direct measurement of polarization in the device.","fun_headline_variants_meta":{"raw":{"variants":["Polarization switch gives fast, sensitive 2D photodetector","Trap-free photoresponse from ferroelectric MoS2 allows 31-photon readout","Millisecond response, no persistence: 31-photon detection in 2D hybrid","Ferroelectric 3R-MoS2 enables trap-free detection of 31 photons","Trap-free 2D photodetector sees down to 31 photons"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001784,"raw_usage":{"total_tokens":7103,"prompt_tokens":1086,"completion_tokens":6017,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":702,"completion_tokens_details":{"reasoning_tokens":5908}},"tokens_in":702,"tokens_out":6017,"duration_ms":39922,"temperature":1.0,"reasoning_tokens":5908,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T19:51:25.328793+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the out-of-plane polarization of the 3R-MoS2 bilayer in the operating device, for example with Kelvin probe force microscopy or optical second-harmonic generation, before, during, and after a 532 nm pulse at 100 K with the same gate voltages. If the polarization does not change while the fast, temperature-independent millisecond photoresponse remains, the proposed polarization-modulation mechanism is ruled out.","supporting_citations":[{"cited_title":"Weston, E","cited_arxiv_id":null,"evidence_quote":"Shows interfacial ferroelectricity persists in 2D semiconductors, supporting room-temperature spontaneous polarization."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates spontaneous-polarization-induced photovoltaic effect in 3R-MoS2, linking the polarization to photocarrier response."},{"cited_title":"Liang, D","cited_arxiv_id":null,"evidence_quote":"Supplies the asymmetric interlayer coupling and type-II band alignment used to explain layer-selective carrier confinement."},{"cited_title":"Gao and L","cited_arxiv_id":null,"evidence_quote":"Predicts large photoinduced tuning of ferroelectricity in sliding ferroelectrics, the specific mechanism behind the optically controlled gate."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the trap-mediated photogating and persistent photoconductivity baseline the paper sets out to overcome."},{"cited_title":"Parappurath, S","cited_arxiv_id":null,"evidence_quote":"Provides the Gr/TMD power- and temperature-dependent timing comparison that distinguishes the new mechanism."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Argues defect-mediated carrier trapping in sliding ferroelectrics, supporting the lower interfacial trap density claim."}],"review_version":1}