{"id":"9e917237-939a-4951-9b43-3fc3580dd2db","arxiv_id":"2411.15818","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Adding electrons to a gated La1-xSrxVO3 channel increases its resistance, implying effective-mass enhancement and a charge gain of at least 100 times.","lead":"Researchers built a back-gated transistor with an ultrathin film of the correlated oxide La1-xSrxVO3 and found that adding electrons with the gate raises the channel resistance, opposite to an ordinary transistor. They interpret this as a Mott-system charge gain of at least 100 times the gate-induced charge, a step toward low-power correlated-oxide switches.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No control experiment excludes STO electromechanical strain or gate leakage as the source of the 1% resistance increase, leaving the ×100 charge gain unsupported.","rationale":"The reader's verdict already flags the same class of concern: the 1% ΔR/R at 100 K may include contributions from STO trap charging, gate leakage, piezoelectric strain, or gate-voltage-dependent artifacts. My stress-test agrees and sharpens it to the most load-bearing point: the ×100 charge-gain number is a quantitative amplification of the measured 1%, and the paper provides no control sample or gate-leakage measurement to separate the correlated response from electromechanical or parasitic effects. The paper's own statement in Section 2.2 that ΔQgate is an upper bound and its supplementary discussion of multiple contributions at low temperature (Section 5) further show that the assignment of the full effect to the gated layer is not independently anchored. The qualitative sign of the response—resistance increase upon electron addition—is consistent with mass enhancement, but a piezoresistive or surface-scattering mechanism could also produce the same sign. The proposed control experiment (non-correlated metal on STO) is a direct, decisive test. Since the reader already issued CONDITIONAL, my assessment does not change the verdict, but it reinforces the condition: the control and leakage measurements are prerequisites for accepting the central claim.","tokens_in":12395,"tokens_out":4702,"duration_ms":46473,"concrete_test":"Fabricate a Hall bar from a non-correlated metal (e.g., 40 nm Ti / 200 nm Au, identical to the contact stack, or Pt) on an identical 0.5 mm STO substrate with the same back-gate configuration, and measure ΔR/R vs Vg from -200 V to +200 V at 100 K using 4-point geometry. Also record gate leakage current IG vs VG on the actual LSVO device during a slow sweep. If the control metal shows a comparable same-sign ΔR/R (~0.1–1%) or if IG is non-negligible, the LSVO result cannot be unambiguously attributed to correlation. Repeating the LSVO sweep at different sweep rates would additionally test for STO trap-charging transients.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim of at least ×100 charge gain (Section 3.2) is built on a single measured quantity: ΔR/R ≈ 1% at 100 K and Vg = 200 V. From Eq. (1), ΔQgate ≈ 3×10^12 cm^-2 is computed assuming bulk STO capacitance; the paper itself calls this an upper bound in Section 2.2 because it 'neglects possible defects and traps in the substrate.' The entire effect is then assigned to a presumed 1–2 u.c. 'gated LSVO' layer, with the remaining 9–8 u.c. treated as a passive parallel resistor. Under that dilution, a 1% total change corresponds to roughly a 30–40% resistance change in the gated layer, so even a small artifact contribution to the measured 1% would destroy the inferred charge gain. The manuscript reports no gate-leakage current during the sweeps and no control sample. STO at 100 K with ε ≈ 1280 and a 200 V bias (4 kV/cm across 0.5 mm) can exhibit electrostrictive or piezoelectric strain; the coherently strained LSVO film is a piezoresistive oxide, and a strain-induced resistance change could have either sign and the same order of magnitude. A non-correlated metal Hall bar on identical STO would reveal such a response. The absent control is the load-bearing gap: without it, the observation that 'cannot be explained in a purely electrostatic framework' is not established, and the effective-mass enhancement and ×100 charge gain rest entirely on assigning the full 1% to correlated electron physics.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports back-gate field-effect measurements on 10 u.c. La1-xSrxVO3 (x = 0.15, 0.20, 0.25) Hall bars grown on SrTiO3 substrates. The authors observe that the channel resistance increases when a positive back-gate voltage nominally adds electrons, opposite to the expected electrostatic response for an electron-doped metal. They interpret this as gate-induced effective-mass enhancement as the Mott system approaches its integer-filling insulating state, and they estimate a 'charge gain' of at least ×100 relative to the electrostatically induced charge. The paper includes structural characterization, Hall measurements, and gate sweeps at 100 K and other temperatures.","tokens_in":12656,"tokens_out":4615,"duration_ms":36981,"significance":"If the interpretation is correct, the work would provide the first demonstration of electric-field control of electron correlation in a perovskite vanadate and would validate the Rozenberg prediction (ref 9) in a solid-state geometry. The qualitative sign of the resistance response and the absence of hysteresis are well presented, and the authors are transparent about the upper-bound nature of the gate charge estimate. However, the central quantitative claim of ×100 charge gain is not directly measured and depends on unverified assumptions about the gated-layer thickness and on the absence of non-correlated artifacts. With additional control experiments and device statistics, the result would be significant for oxide electronics and correlated-electron physics.","major_comments":[{"comment":"The load-bearing claim that the 1% resistance increase at 100 K and Vg = 200 V arises from capacitive electron doping is not supported by control measurements. The manuscript reports no gate-leakage current during sweeps and no control device (e.g., a metallic non-correlated Hall bar on the same STO substrate) to rule out electromechanical strain of the coherently strained LSVO film or gate-voltage-dependent contact artifacts. Since STO is piezoelectric/electrostrictive and the film is coherently strained, a strain-induced resistance change of the observed sign and magnitude cannot be excluded. Without such a control, the statement in the abstract that the response 'cannot be explained in a purely electrostatic framework' is premature.","section":"Section 2.2, Eq. (1), and Section 3.2"},{"comment":"The ×100 charge gain estimate rests on a model-dependent division of the film into a 1-2 u.c. 'gated LSVO' layer and a passive parallel resistor. The screening length is an assumed parameter (refs 43,44) rather than a measured quantity for this heterostructure, and Eq. (1) is itself labeled an upper bound in Section 2.2. Consequently, the numerical value of the charge gain is an upper-bound-based estimate, not a measured result. The authors should either measure the gated-layer response directly (e.g., by a thickness series or frequency-dependent gating) or present the ×100 figure as a model-dependent upper bound with explicit error propagation.","section":"Section 3.2"},{"comment":"The dataset comprises one device per composition, with no error bars or repeated-device statistics. The apparent optimum in Fig. 3c for the 20% device is based on a single point; this is insufficient to support the fine-tuning claim. The authors should report statistics from multiple devices, or explicitly state the single-device limitation in the main text.","section":"Figure 3 and Section 2.2"},{"comment":"The assignment of the resistance increase to effective-mass enhancement is inferred from a single transport measurement (ΔR/R) and from the absence of a simple mobility-degradation pattern at low temperatures (Supp. Fig. S6). This is indirect evidence; gate-dependent Hall measurements at 100 K, or a direct probe of the gated layer, would be needed to exclude other contributions such as field-dependent scattering or interface trap charging. The text should acknowledge this indirectness.","section":"Section 3.1 and Supplementary Section 5"}],"minor_comments":[{"comment":"The charge modulation is given as '3×10^12 cm2'; the unit should be cm^-2.","section":"Eq. (1) and Figure 3a"},{"comment":"The phrase '×1000 difference' is confusing; the text in Section 3.2 says the gate modulates 0.1% of the sheet density, which corresponds to a factor of 1000, not a ×1000 charge gain. Clarify the terminology.","section":"Figure 3d caption"},{"comment":"The '×100 charge gain' is presented before the assumptions are introduced; consider adding 'estimated' or 'suggested' to avoid overstatement.","section":"Abstract and Section 3.2"},{"comment":"The type of capping layer (TiO2-x or Al2O3-x) is not specified per device; since the cap may affect the gate response, please list which device received which cap.","section":"Section 5"},{"comment":"Several entries lack pagination or article identifiers (e.g., refs 1, 6, 7, 41); also fix the typo 'Univesity' in the Acknowledgments.","section":"References and Acknowledgments"},{"comment":"Low-temperature gate data are shown for three contributions; it would be helpful to include the gate leakage current or gate resistance in the figure or text.","section":"Figure S6"}],"recommendation":"major_revision","confidential_remarks":"The paper is a preliminary report with an interesting qualitative finding. The absence of control experiments and statistics is the main concern. I recommend major revision rather than rejection, because the central effect (resistance increase with positive gate) is clearly presented and the authors are honest about the upper-bound nature of ΔQgate. However, the ×100 charge gain claim should not appear in its present form without controls."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The key result is the first solid-state back-gate demonstration on La1-xSrxVO3: applying positive gate voltage increases channel resistance, opposite to what a rigid-band electron system would do. That qualitative observation is credible and worth knowing. The paper does good zero-bias homework: temperature-dependent resistivity, Hall carrier densities, thick/thin comparison, and a SrO monolayer to suppress interfacial 2DEG. The absence of hysteresis and the x-dependence (20% Sr shows the largest response, consistent with proximity to the MIT) all point to an intrinsic channel effect.\n\nThe soft spot is the quantitative charge-gain claim. The entire chain from Eq. (1) to ×100 uses a calculated ΔQgate, an assumed 1–2 u.c. screening length, and a parallel-resistor dilution argument. A 1% total resistance change becomes ~30–40% in the gated layer, so any artifact at the level of a fraction of a percent—STO electrostrictive strain, trap charging, gate leakage—can be magnified into the inferred mass enhancement. The paper does not report leakage currents or a control device (e.g., a metallic Hall bar on the same STO), and the stress-test is right that this is the missing control. I don't think that kills the qualitative story, because the resistance increase is systematic across compositions and temperatures, but it does mean the ×100 charge gain is not established.\n\nThe paper itself is honest about the upper-bound nature of ΔQgate and calls the results preliminary. The main overclaim is the abstract's 'cannot be explained in a purely electrostatic framework'—that's too strong given the unmeasured artifacts. A revision with gate-leakage data, a control sample, multiple devices, or a thickness series would turn this from an interesting preprint into a solid paper.\n\nThis is a paper for people working on Mott devices and oxide electronics. It deserves serious peer review—the experimental milestone is real, and the referees can push for the missing controls. I'd bring it to a reading group, and if I worked in the area I'd cite it as the first solid-state gating of LSVO, not for the charge-gain number.","headline":"Solid-state back-gating of LSVO shows a real but unquantified resistance increase; the ×100 charge gain is an estimate riding on assumptions that need control experiments.","tokens_in":13290,"tokens_out":2208,"would_cite":true,"duration_ms":20686,"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 reports that a back gate adding electrons to an oxide Mott film raises its resistance, a response the authors attribute to effective-mass enhancement near the Mott insulating state and translate into a charge gain of at least…","keywords":["Mott insulator","charge gain","effective mass enhancement","field-effect transistor","La1-xSrxVO3","perovskite vanadate","electron correlation","metal-insulator transition"],"falsifier":"Fabricate the same 10-unit-cell Hall bar and back-gate stack with a metallic SrVO$_3$ channel, which has no nearby Mott transition, and measure $\\Delta R/R$ at 100 K under the same gate voltages; if this control shows a resistance increase of the same order as the La$_{1-x}$Sr$_x$VO$_3$ devices, the correlated charge-gain interpretation is refuted.","tokens_in":12,"feed_emoji":"⚡","tokens_out":8417,"duration_ms":132416,"temperature":0.7,"pith_summary":"Applying a positive back-gate voltage to a thin film of the oxide Mott system La$_{1-x}$Sr$_x$VO$_3$ raises the film's resistance by up to about 1 percent, even though the gate is adding electrons that would make an ordinary metal more conductive. The paper argues that this 'wrong-sign' response cannot be accounted for by electrostatics alone and instead reflects the growth of the electron effective mass as the correlated metal is pushed toward its Mott insulating state. The inferred mass enhancement translates into a charge gain of at least $\\times100$ relative to the sheet charge actually injected by the gate, once the finite screening depth of the gated layer is taken into account. Demonstrating this effect in a solid-state geometry is a step toward Mott-based transistors that switch with far smaller gate voltages than silicon devices.","feed_headline":"A gate that adds electrons makes a Mott film more resistive","feed_subtitle":"The 1 percent resistance shift implies a correlated mass gain and at least ×100 charge gain.","key_machinery":"The load-bearing object is the correlated channel close to a filling-controlled Mott transition, described by the effective carrier density $n_{\\mathrm{eff}} = n/m^*$, where $n$ is the carrier density and $m^*$ the effective mass. The gate acts as a capacitor that injects a sheet charge $\\Delta Q_{\\mathrm{gate}} = V_g \\varepsilon_0 \\varepsilon_{\\mathrm{STO}} / t_{\\mathrm{STO}}$; the resistance response of the channel is then interpreted through the ratio $n/m^*$, and 'charge gain' is defined as the amplification between the gate-induced charge and the change in delocalized carriers. The chain of argument uses the assumed 1–2 unit-cell screening length to isolate the gated slice of the film from the unperturbed parallel resistance of the rest.","core_discovery":"The central claim is that gate-controlled electron correlation is observable in a perovskite vanadate in a solid-state device. In 10-unit-cell La$_{1-x}$Sr$_x$VO$_3$ films ($x = 0.15, 0.20, 0.25$) on SrTiO$_3$ at 100 K, a positive back-gate bias raises the four-point resistance, with the largest response ($\\sim 1\\%$) for the $x = 0.20$ film closest to the filling-control metal-insulator transition. Because the gate injects only $\\sim 3\\times10^{12}\\,\\mathrm{cm}^{-2}$ electrons into a sheet reservoir of $\\sim 10^{15}\\,\\mathrm{cm}^{-2}$, a simple rigid-band picture would predict a resistance decrease of roughly $0.1\\%$; the observed increase is instead read as evidence that $n/m^*$ falls as $m^*$ diverges near integer filling. The authors estimate the charge gain — the ratio of the change in delocalized carriers to the gate-induced charge — as at least $\\times100$ when the 1–2 unit-cell screening length of the gated region is accounted for against the rest of the film acting as a passive parallel resistor.","pith_inferences":["A control experiment on a metallic perovskite without a nearby Mott transition, such as SrVO$_3$ under identical gating, would isolate whether the 'wrong-sign' resistance increase is unique to the correlated channel or partly a substrate artifact.","The 1–2 unit-cell screening-length assumption could be tested by measuring the gate response as a function of film thickness: if the charge gain is truly $\\times100$, $\\Delta R/R$ should grow sharply as the channel approaches the screening length.","At larger gate charges one might expect the resistance response to become nonlinear as the system approaches the insulating state; a systematic study of $\\Delta R/R$ versus gate voltage beyond the small-bias limit would discriminate between a mass-enhancement mechanism and a gate-voltage-dependent artifact.","Because SrTiO$_3$'s dielectric constant and trap response both depend strongly on temperature, repeating the measurement at 300 K with a high-permittivity top gate would help separate the correlated response from substrate leakage and trapping."],"forward_implications":["A small gate-induced charge can produce a much larger change in conductive carriers, the functional advantage that Mott transistors would need for low-voltage switching.","The composition dependence — peak response near $x = 0.20$, closest to the reported $x = 0.18$ metal-insulator transition — shows the effect can be tuned by moving the channel toward the phase boundary.","The same correlated response is expected at higher temperatures with a top gate and thinner channels, pointing to a path toward room-temperature operation.","The result provides a direct experimental probe of effective-mass enhancement in a Mott system, testing theoretical predictions that $m^*$ diverges near integer filling.","If the charge gain holds, gate control of correlation could be combined with ferroelectric gates or epitaxial liftoff to integrate oxide Mott channels with semiconductor technology."],"supporting_citations":[{"why":"Defines the LSVO phase diagram and places the filling-control metal-insulator transition at $x = 0.18$, the reference for choosing the three compositions and interpreting their zero-bias transport.","marker":"[31]"},{"why":"Theoretical prediction that the effective mass diverges as a Mott system approaches integer filling; this is the mechanism the paper invokes to explain the sign of the gate response.","marker":"[9]"},{"why":"Source of the up-to-$\\times100$ charge-gain figure that the measured response is compared against.","marker":"[10]"},{"why":"Provides the SrTiO$_3$ dielectric constant (1280 at 100 K, essentially field-independent) used to compute the gate-induced sheet charge $\\Delta Q_{\\mathrm{gate}}$.","marker":"[42]"},{"why":"Basis for the 1–2 unit-cell screening length that separates the gated layer from the passive parallel resistance and raises the gain estimate to $\\times100$.","marker":"[43]"}],"fun_headline_variants":["Solid-state gating yields 100-fold charge gain in Mott film","Counterintuitive: gate adds electrons, Mott resistance rises","Oxide Mott channel: small gate, 100x charge gain","Mott film resists more when gate injects electrons","Electron injection strengthens correlation in vanadate FET"],"cache_read_input_tokens":15360,"weakest_assumption_plain":"The measured 1 percent resistance increase is assumed to be caused by the extra electrons entering the thin Mott film, with no significant contribution from trapping of charge in the strontium titanate substrate, leakage through the gate, or strain deforming the crystal.","fun_headline_variants_meta":{"raw":{"variants":["Solid-state gating yields 100-fold charge gain in Mott film","Counterintuitive: gate adds electrons, Mott resistance rises","Oxide Mott channel: small gate, 100x charge gain","Mott film resists more when gate injects electrons","Electron injection strengthens correlation in vanadate FET"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001415,"raw_usage":{"total_tokens":5746,"prompt_tokens":1006,"completion_tokens":4740,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":622,"completion_tokens_details":{"reasoning_tokens":4656}},"tokens_in":622,"tokens_out":4740,"duration_ms":33367,"temperature":1.0,"reasoning_tokens":4656,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T13:51:58.001015+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fabricate the same 10-unit-cell Hall bar and back-gate stack with a metallic SrVO$_3$ channel, which has no nearby Mott transition, and measure $\\Delta R/R$ at 100 K under the same gate voltages; if this control shows a resistance increase of the same order as the La$_{1-x}$Sr$_x$VO$_3$ devices, the correlated charge-gain interpretation is refuted.","supporting_citations":[],"review_version":1}