{"id":"4b00288f-2509-4e04-93b4-67aa65bdd2cb","arxiv_id":"2608.10174","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"4H-SnS2 is shown to be a room-temperature ferroelectric whose transistor channel current and photocurrent respond to gate bias and to linear and circular light polarization.","lead":"Researchers show that the 4H crystal form of tin disulfide is ferroelectric at room temperature, while the common 2H form is not. They build a transistor from the 4H material whose current and photoresponse depend on gate voltage and on the polarization of light.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The ferroelectric-FET claim rests on a ~500x coercive-field mismatch: the paper's own PFM data (3 V on 50 nm) imply ~600 kV/cm, not the 1.2 kV/cm used to justify 1.5 kV/cm gate switching.","rationale":"Good-faith reading: the paper makes two coupled claims — intrinsic room-temperature ferroelectricity in 4H-SnS2, established mainly by PFM/KPFM with 2H controls, and ferroelectric control of transistor and photoresponse, established by transport and photocurrent measurements on a 100 micrometer-thick three-terminal device. The PFM evidence is substantial: off-field phase and amplitude loops, retention, endurance, thickness dependence, in-plane response weaker, 2H control non-ferroelectric. I give credit for those controls. The weakest link is the quantitative bridge from PFM coercive voltage to transport coercive field. The reader identified the geometric mismatch; I would go further and note that the paper's own numbers make the gate-field argument internally inconsistent, since 3 V on 50 nm implies about 600 kV/cm, a factor of roughly 500 larger than the quoted 1.2 kV/cm. That factor is what actually supports the claim that a 15 V gate on 100 micrometers can influence polarization. Without it, the transistor hysteresis and memory states are not tied to ferroelectric switching, and the gate-modulated photoresponse could be a conventional semiconductor effect. This does not disprove the PFM-based ferroelectricity claim, but it removes the load-bearing part of the transistor claim, so the manuscript should not be accepted before a uniform-field coercive-field measurement (P-E/PUND) is supplied. This aligns exactly with the reader's CONDITIONAL verdict; my concern is the same one, so no verdict change is needed.","tokens_in":20293,"tokens_out":7647,"duration_ms":75186,"concrete_test":"Perform room-temperature macroscopic P-E and PUND measurements on bulk 4H-SnS2 crystals with uniform electrodes on opposite c-faces, and extract the coercive field from the P-E loop. If Ec comes out near the ~600 kV/cm implied by the PFM data rather than the 1.2 kV/cm quoted in Section 2.2, then the 1.5 kV/cm gate field at Vg = 15 V cannot switch the polarization, and the ferroelectric-FET interpretation must be replaced by a charge-trapping/interface-state model. This single measurement settles whether the gate can actually influence the polarization.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 2.2's argument that electrostatic gating modulates the built-in polarization depends on the statement that the coercive field of 4H-SnS2 is approximately 1.2 kV/cm and that a 15 V gate on a 100 micrometer crystal produces Eg = 1.5 kV/cm, 'comparable' to that field. This is not internally consistent. The only PFM coercive voltage reported in the paper is about 3 V applied to a ~50 nm flake, which gives a nominal uniform-field coercive field of roughly 600 kV/cm (3 V / 50 nm), not 1.2 kV/cm. The 1.2 kV/cm figure appears instead to be the lateral source-drain field (3 V / 25 micrometers), which acts in-plane and cannot switch the out-of-plane polarization probed by PFM. Consequently, the coincidence between the Vd approx 3 V transport threshold and the PFM coercive voltage is a dimensional accident rather than evidence of polarization switching. If the true coercive field is hundreds of kV/cm, the 1.5 kV/cm gate field is orders of magnitude too small to switch or appreciably modulate the out-of-plane polarization, and the observed programming-state memory, Id-Vg hysteresis, and gate-modulated photocurrent can be explained by charge trapping, interface states, or ion migration without invoking ferroelectric switching. The paper itself defers macroscopic P-E/PUND measurements, so the polarization-controlled transistor part of the central claim is currently unsupported, even though the local PFM evidence for ferroelectricity may survive.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports room-temperature ferroelectricity in the 4H polytype of SnS2 and claims that the spontaneous polarization controls the electrical transport and photoresponse of a three-terminal transistor built on the same crystal. The evidence for ferroelectricity includes Raman identification of the 2H and 4H polytypes, XRD, PFM phase and amplitude hysteresis, domain writing with retention up to 60 hours, endurance over 20 cycles, thickness-dependent PFM, and a 2H PFM control. The transport section reports drain-current asymmetries under programming voltages, hysteretic transfer curves with temperature-dependent memory windows, time-dependent relaxation, and gate-dependent photocurrents under linearly and circularly polarized light, fitted with CPGE and LPGE terms. The central claim is that 4H-SnS2 is the first pristine binary van der Waals semiconductor combining switchable out-of-plane polarization with polarization-sensitive transistor action and photoresponse at room temperature.","tokens_in":20638,"tokens_out":6349,"duration_ms":64714,"significance":"If the central claim holds, the paper identifies a structurally simple binary van der Waals semiconductor with room-temperature ferroelectric order and demonstrates coupling of that order to transistor and photoresponse functionality, which would be of substantial interest for nonvolatile memory and photoferroelectric devices. The local PFM evidence is a clear strength: domain writing with opposite phase contrast, off-field phase and amplitude loops, long retention, repeated switching endurance, thickness-dependent switching, and the direct absence of response in the centrosymmetric 2H control together make a credible case for local ferroelectric switching. The transport and photocurrent data are extensive and, if properly connected to polarization, would be significant. However, the quantitative link between the gate field and ferroelectric switching is currently based on an inconsistent coercive-field estimate, and the absence of a 2H transistor control weakens the attribution of the electrical hysteresis to polarization. These issues are load-bearing for the transistor-action part of the central claim.","major_comments":[{"comment":"The coercive-field estimate used to justify gate control of polarization is internally inconsistent. The PFM data in Fig. 1g show polarization switching at a tip bias of about 3 V on a roughly 50 nm thick flake, which under a uniform-field estimate gives an out-of-plane coercive field of order 600 kV/cm, not 1.2 kV/cm. The 1.2 kV/cm value quoted in Section 2.2 instead matches 3 V divided by the 25 micrometer source-drain spacing, which is a lateral in-plane field and cannot switch the out-of-plane polarization probed by PFM. Consequently, the statement that a gate field of 1.5 kV/cm is 'comparable' to the coercive field is not supported by the paper's own data, and the claim that electrostatic gating significantly modulates the ferroelectric polarization is quantitatively unsupported.","section":"Section 2.2, Fig. 1g and Fig. 3c,e"},{"comment":"The coincidence between the Vd ≥ 3 V transport threshold and the PFM coercive voltage is presented as evidence of polarization switching, but the two voltages are applied in different geometries and directions: the drain voltage acts laterally across a 25 micrometer channel, whereas the PFM voltage is applied vertically across a roughly 50 nm flake. A voltage coincidence alone cannot establish that the same switching field is involved. The observed transport threshold and the resulting hysteresis could equally arise from bias-dependent carrier injection, charge trapping, or ion migration. Direct evidence of gate- or drain-induced polarization switching, for example by performing PFM on the channel after applying the relevant bias, is needed to support the ferroelectric interpretation.","section":"Section 2.2, Fig. 3c and Fig. 3e"},{"comment":"No transport or photoresponse measurements are reported for the 2H-SnS2 control crystal. The 2H polytype is used as a PFM control, which is appropriate, but the transistor hysteresis, gate dependence, and polarization-dependent photocurrent are not compared between 2H and 4H. Without that control, the polytype-specificity of the electrical and optoelectronic claims is not established, and the observed Id-Vg hysteresis and gate-tunable photoresponse could be caused by sample-specific defects, contacts, or surface adsorbates rather than by ferroelectric polarization.","section":"Section 2.2, Figs. 3-5"},{"comment":"The paper itself states that 'future macroscopic P-E and PUND measurements will be valuable for quantitatively determining the intrinsic polarization and coercive field.' In the absence of such measurements, the quantitative coercive-field value used in Section 2.2 for the gate-field comparison is unverified. The paper should either obtain a direct measurement of the out-of-plane coercive field relevant to the gate geometry or substantially weaken the quantitative claims about gate-induced polarization modulation.","section":"Section 2.1, last paragraph"}],"minor_comments":[{"comment":"The abstract says 'establish a room-temperature ferroelectric state in SnS2' without specifying 4H-SnS2 in the first sentence; the scope should be clarified immediately to avoid implying that the common 2H phase is ferroelectric.","section":"Abstract and Introduction"},{"comment":"If the 1.2 kV/cm value is retained, the derivation must be stated explicitly; as written, the text moves from a PFM voltage threshold and a lateral drain-voltage threshold to a field value without showing how the geometry conversion was made.","section":"Section 2.2, 'coercive field' sentence"},{"comment":"The caption says the measurement was performed 'over a duration of 20 mins,' while the text says 'monitored over 30 minutes time'; these should be reconciled.","section":"Figure 3 caption"},{"comment":"The power-law fit Iph = A Pb reports A and b without a goodness-of-fit metric; an R2 value or residual analysis should be provided to assess the fit quality.","section":"Section 2.3, Eq. (2)"},{"comment":"The text refers to 'fitting Iph(phi) in Figure 6(b)' when the data are in Figure 7(b); the figure reference is incorrect.","section":"Figure 7 caption and text"},{"comment":"The sentence 'The corresponding in-plane and out-of-plane resistivities were calculated using the measured device geometry (Figure 2b)' refers to the wrong figure; the device geometry is shown in Figure 3b.","section":"Experimental Section"}],"recommendation":"major_revision","confidential_remarks":"The PFM evidence for local ferroelectricity in 4H-SnS2 is credible and likely publishable, but the transistor-action portion of the central claim currently rests on a coercive-field comparison that is quantitatively inconsistent with the paper's own PFM data. If the authors cannot provide direct evidence that the applied gate field switches or substantially modifies the out-of-plane polarization (for example, PFM imaging after gate pulsing, or an independent determination of the out-of-plane coercive field), the manuscript's conclusions about ferroelectric control of transport and photoresponse will need to be substantially narrowed. The absence of a 2H transistor control is another gap that should be addressed in revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know before reading. First, the local PFM evidence for room-temperature ferroelectricity in 4H-SnS2 is the strongest part of the paper: box-in-box domain writing, 180° phase contrast, butterfly amplitude loops, retention for tens of hours, endurance for 20 cycles, thickness-dependent switching, and a clean 2H control. That is a credible experimental claim for a theoretically predicted ferroelectric. Second, the transport and photogalvanic interpretations are not yet supported. The paper compares the PFM coercive voltage of ~3 V to a drain-voltage threshold of ~3 V and then calls the corresponding field 1.2 kV/cm. But the PFM coercive voltage was applied across a ~50 nm flake, so the field is ~600 kV/cm; 1.2 kV/cm is the lateral source-drain field (3 V over 25 µm). Those are different geometries with different field directions. The gate field of 1.5 kV/cm at Vg = 15 V on a 100 µm crystal is therefore orders of magnitude too small to switch an out-of-plane polarization that needs ~600 kV/cm. So the programming-state memory, Id-Vg hysteresis, and gate-modulated photocurrent can be fully explained by charge trapping, interface states, or ion migration. The paper even acknowledges these alternatives and then dismisses them on the basis of the bad coercive-field estimate. That is a load-bearing flaw, not a cosmetic one.\n\nWhat else? The CPGE/LPGE analysis in Fig. 7 is underpowered: no non-ferroelectric control, no error bars on the fit coefficients, and the sign reversal with temperature is not tied to any specific band structure calculation. The Raman correlation with hysteresis is suggestive but not cause-and-effect.\n\nThe good news: the local ferroelectric claim may survive peer review. The 2H control, retention, endurance, and thickness dependence are exactly the checks you want. The paper honestly says macroscopic P-E/PUND are still needed. If those are added, and the transport section is reframed as polarization-modulated in a device where the actual coercive voltage is measured, the paper becomes steady.\n\nWho is this for? Workers on 2D ferroelectrics and photoferroelectric devices. It deserves serious referee time, but the authors need to fix the coercive-field analysis and add macroscopic switching evidence before the transport claims are credible.","headline":"Local PFM evidence for ferroelectric 4H-SnS2 is solid; the transistor and photogalvanic claims rest on a coercive-field estimate that mixes a lateral drain field with a tip-based PFM threshold.","tokens_in":21201,"tokens_out":5660,"would_cite":false,"duration_ms":51736,"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 provides the first direct experimental evidence for room-temperature ferroelectricity in the 4H polytype of SnS2, and shows that this polarization controls transistor behavior and photoresponse.","keywords":["van der Waals ferroelectricity","4H-SnS2","SnS2 polytypes","ferroelectric field-effect transistor","circular photogalvanic effect","piezoresponse force microscopy","polarization-sensitive photoresponse","room-temperature ferroelectric"],"falsifier":"Measure the polarization–electric-field (P–E) or positive-up–negative-down (PUND) hysteresis directly on the same 4H-SnS2 crystal used for the transistor, and repeat the transport measurements on a device where the channel thickness is known and the gate field is calculated from the actual crystal thickness; if the coercive field comes out much larger than about 1.2 kV/cm, or if the I_d–V_g hysteresis persists unchanged when the gate field is kept below any plausible coercive field, the ferroelectric explanation fails.","tokens_in":20073,"feed_emoji":"⚡","tokens_out":5529,"duration_ms":47819,"temperature":0.7,"pith_summary":"This paper sets out to establish that the layered semiconductor 4H-SnS2 is ferroelectric at room temperature and that this spontaneous polarization controls both its transistor behavior and its response to polarized light. The authors argue that ferroelectricity is a property of the crystal polytype: the polar 4H stacking breaks inversion symmetry, while the centrosymmetric 2H stacking does not, and they support this by contrasting the two polymorphs. They report that a three-terminal device made from a single 4H-SnS2 crystal, which serves as both channel and gate dielectric, shows gate-tunable, hysteretic, p-type transport and a photoresponse that depends on laser power, temperature, and the linear or circular polarization of the light. A sympathetic reader would care because 4H-SnS2 would be a simple, two-element van der Waals ferroelectric platform for nonvolatile memory and polarization-sensitive optoelectronics.","feed_headline":"Layered 4H-SnS2 shows room-temperature ferroelectric switching","feed_subtitle":"The same crystal acts as transistor channel and gate, and polarized light tunes its photocurrent.","key_machinery":"The central object is the 4H polytype of SnS2, a hexagonal stacking with four sulfur layers per unit cell (space group P63mc) that breaks inversion symmetry and produces a net out-of-plane electric polarization; the same crystal serves simultaneously as semiconducting channel and as the ferroelectric gate dielectric in the transistor. The argument is carried by the polarization-induced internal field: bound polarization charges create band bending and hole accumulation that shows up as p-type, hysteretic transport, and the same field separates photogenerated electron-hole pairs, producing polarization-dependent photocurrent. The machinery includes the Raman Eg-mode doublet as the polytype fingerprint, PFM and KPFM domain writing as the proof of switchable polarization, and the CPGE/LPGE fitting formula I(φ) = D + C sin(2φ) + L1 sin(4φ) + L2 cos(4φ) that separates circular and linear photogalvanic contributions.","core_discovery":"The paper's central claim is that the 4H polytype of SnS2 hosts intrinsic, switchable, out-of-plane ferroelectric polarization at room temperature, as predicted for the P63mc stacking, and that this polarization is the controlling internal field in a three-terminal transistor. Evidence offered: piezoresponse force microscopy shows 180-degree phase reversal, butterfly amplitude loops with a ~3 V coercive voltage, long retention up to 60 hours, and endurance over 20 write-erase cycles on exfoliated flakes, while identical measurements on 2H-SnS2 show no ferroelectric response. In transport, the drain current becomes asymmetric and enhanced when programming or gate voltages exceed the ~3 V switching threshold, the I_d–V_g curves are hysteretic with a memory window that tracks the PFM coercive field, and the photocurrent under 532 nm light changes sign and magnitude with gate voltage, temperature, and light helicity. The authors conclude that the built-in polarization modulates carrier density and photocarrier separation, making 4H-SnS2 the first pristine binary van der Waals semiconductor with room-temperature ferroelectricity tied to polytype and demonstrated in transistor action.","pith_inferences":["Editorial inference: If room-temperature ferroelectricity in 4H-SnS2 is confirmed by macroscopic P–E or PUND measurements, then stacking polytype engineering becomes a general route to hidden ferroelectricity in other binary layered dichalcogenides, not just SnS2.","Editorial inference: The helicity-dependent photocurrent could serve as an optical readout of the ferroelectric polarization state, since the CPGE and LPGE amplitudes and the LCP–RCP separation track the hysteresis and reverse sign near 250 K.","Editorial inference: The reported coercive field of about 1.2 kV/cm is unusually low; if it holds, 4H-SnS2 would merit testing in low-voltage ferroelectric field-effect transistors, but that claim needs a direct measurement of coercivity on the same crystal used for transport.","Editorial inference: A decisive control experiment would compare nominally identical devices under light of opposite helicity at fixed gate voltage to separate true photogalvanic current from thermal or bolometric contributions; the paper's oblique-incidence geometry is consistent with CPGE but does not by itself exclude helicity-dependent absorption artifacts."],"forward_implications":["4H-SnS2 can be used as a nonvolatile ferroelectric memory element: programmed current states remain separated for at least 300 s, and written domains persist for tens of hours.","Polytype, not just composition, decides ferroelectric functionality in SnS2: the 2H phase remains nonpolar while the 4H phase switches, so growth conditions that select polytype control device behavior.","The all-SnS2 transistor architecture works because the roughly 142–250x transport anisotropy confines current to the top layers and keeps gate leakage below about 1%.","The photoresponse is polarization-tunable: gate voltage, light helicity, and temperature can reverse or enhance the photocurrent, and the CPGE and LPGE amplitudes peak near the temperature where ferroelectric hysteresis is largest."],"supporting_citations":[{"why":"Theoretical prediction of spontaneous out-of-plane ferroelectric polarization in hexagonal 4H-SnS2, which this paper experimentally tests.","marker":"[31]"},{"why":"Provides the Raman mode assignment used to distinguish 2H and 4H polytypes, especially the Eg-mode doublet.","marker":"[32]"},{"why":"Establishes SnS2 as a two-element indirect band gap semiconductor and provides reference Raman and FET context.","marker":"[29]"},{"why":"CuInP2S6 van der Waals ferroelectric used as a comparative framework for PFM domain switching and ferroelectric behavior.","marker":"[14]"},{"why":"Gives the coercive field of CuInP2S6 (~25-30 kV/cm) against which the lower 4H-SnS2 threshold is compared.","marker":"[45]"},{"why":"Supplies the fitting expression I(φ) = D + C sin(2φ) + L1 sin(4φ) + L2 cos(4φ) used to separate CPGE and LPGE contributions.","marker":"[64]"},{"why":"Numerical simulation of transport anisotropy showing how cross-plane resistance suppresses gate leakage, supporting the all-SnS2 device geometry.","marker":"[40]"},{"why":"Provides the out-of-plane dielectric constant εr = 6.98 used to estimate gate capacitance and gate field.","marker":"[39]"}],"fun_headline_variants":["Light-polarized control of a ferroelectric SnS2 transistor","4H-SnS2: first binary vdW crystal with room-T ferroelectric switching","Polarization-sensitive transistor from a ferroelectric tin disulfide","Room-temperature ferroelectric transistor in layered SnS2","SnS2 transistor's current flips with light and ferroelectric memory"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the roughly 3 V threshold seen in the transistor's drain-current measurements on a 25-micrometer channel is the same ferroelectric coercive voltage measured by PFM on a roughly 50-nanometer flake, so the gate field of about 1.5 kV/cm at 15 V is strong enough to switch or modulate the polarization; if the true coercive field is much larger, the ferroelectric interpretation of the transport hysteresis collapses.","fun_headline_variants_meta":{"raw":{"variants":["Light-polarized control of a ferroelectric SnS2 transistor","4H-SnS2: first binary vdW crystal with room-T ferroelectric switching","Polarization-sensitive transistor from a ferroelectric tin disulfide","Room-temperature ferroelectric transistor in layered SnS2","SnS2 transistor's current flips with light and ferroelectric memory"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000601,"raw_usage":{"total_tokens":2871,"prompt_tokens":1071,"completion_tokens":1800,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":687,"completion_tokens_details":{"reasoning_tokens":1702}},"tokens_in":687,"tokens_out":1800,"duration_ms":13235,"temperature":1.0,"reasoning_tokens":1702,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:11:45.600974+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the polarization–electric-field (P–E) or positive-up–negative-down (PUND) hysteresis directly on the same 4H-SnS2 crystal used for the transistor, and repeat the transport measurements on a device where the channel thickness is known and the gate field is calculated from the actual crystal thickness; if the coercive field comes out much larger than about 1.2 kV/cm, or if the I_d–V_g hysteresis persists unchanged when the gate field is kept below any plausible coercive field, the ferroelectric explanation fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Theoretical prediction of spontaneous out-of-plane ferroelectric polarization in hexagonal 4H-SnS2, which this paper experimentally tests."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Raman mode assignment used to distinguish 2H and 4H polytypes, especially the Eg-mode doublet."},{"cited_title":"Huang, E","cited_arxiv_id":null,"evidence_quote":"Establishes SnS2 as a two-element indirect band gap semiconductor and provides reference Raman and FET context."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"CuInP2S6 van der Waals ferroelectric used as a comparative framework for PFM domain switching and ferroelectric behavior."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the coercive field of CuInP2S6 (~25-30 kV/cm) against which the lower 4H-SnS2 threshold is compared."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the fitting expression I(φ) = D + C sin(2φ) + L1 sin(4φ) + L2 cos(4φ) used to separate CPGE and LPGE contributions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Numerical simulation of transport anisotropy showing how cross-plane resistance suppresses gate leakage, supporting the all-SnS2 device geometry."},{"cited_title":"Zhen, H.-Y","cited_arxiv_id":null,"evidence_quote":"Provides the out-of-plane dielectric constant εr = 6.98 used to estimate gate capacitance and gate field."}],"review_version":1}