{"id":"585624b0-f629-4011-8d07-8f9fcd364bd1","arxiv_id":"1908.08295","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Gate voltage and light together tune the resistance of an LSMO film by about ±7.5% at 300 K, and an interfacial LTO layer reveals separate mechanisms: photoinduced spin disorder, interfacial charge injection, and oxygen vacancy diffusion.","lead":"This paper shows that shining blue or ultraviolet light while applying a gate voltage changes the resistance of a thin manganite film by about 7.5 percent, with the sign of the change depending on the voltage direction. The authors interpret the effect as a combination of photoexcited spin flips, charge injection across the interface, and light-driven oxygen vacancy movement that alters the film's orbital occupancy.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proposed oxygen-vacancy strain/orbital mechanism for blue-light photoconductivity in sample d10 is not directly measured; without in-situ orbital or lattice probes, the central 'orbital occupancy control' claim is underdetermined.","rationale":"The reader's verdict identifies the strain-orbital mechanism as the weakest load-bearing assumption; my stress-test agrees. The experimental core is a clean set of transport measurements showing gate-polarity-dependent sign inversion of photoresistance in LSMO/STO (sample d0) and suppression of the UV gate-response by an LTO spacer (sample d10). The LTO/STO control (Fig. 4) adds real evidence that the UV effect involves interfacial electron injection from STO, and the opposite sign of the blue-light response in d10 vs LTO/STO argues against a trivial 2DEG contribution. However, the specific explanation for the d10 blue-light photoconductivity, namely oxygen-vacancy diffusion through LTO, oxygen-ion accumulation, octahedral compression, and d_x2-y2 stabilization, rests entirely on extrapolation from other materials systems. The transient current experiment (Fig. 5) was performed on LTO/STO without the LSMO cap, so it does not demonstrate vacancy motion under the actual top-contact conditions. No orbital-resolved spectroscopy or lattice-parameter measurement is reported. This is not an objection to the existence of the effect; it is a gap between the data and the mechanism. A single in-situ XLD/XRD experiment under the operating conditions would either validate or falsify the proposed orbital-occupancy control. Since the paper is transparent about the indirect nature and the reader's verdict is already CONDITIONAL, I recommend no change to the verdict.","tokens_in":13644,"tokens_out":13198,"duration_ms":140441,"concrete_test":"Perform in-situ X-ray absorption linear dichroism (XLD) at the Mn L2,3 edge on sample d10 while illuminating with 441 nm light at V_g = -60 V, and compare the extracted d_x2-y2 vs d_3z2-r2 orbital weight with the dark, V_g = 0 baseline. The strain-orbital mechanism predicts a measurable enhancement of d_x2-y2 under these conditions; if no change in orbital polarization is seen, the central mechanism is falsified and the blue-light photoconductivity must be re-explained.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing premise is the strain mechanism used to explain the blue-light photoconductivity of sample d10 (LSMO/LTO/STO) at V_g = -60 V. The paper infers that blue light doubly ionizes oxygen vacancies in STO, that the resulting V_O diffuse toward the bottom electrode under negative gate bias, and that the opposite-moving oxygen ions accumulate at the LSMO/LTO interface, increase the c parameter of LTO [refs 41,42], compress MnO6 octahedra in the LSMO overlayer along z, stabilize d_x2-y2, and thereby increase in-plane hopping and conductivity. No direct observation of any link in this chain is made in this heterostructure: there is no measurement of orbital occupancy, octahedral distortion, or even of V_O mobility in the LSMO-capped stack (the transient current data in Fig. 5 are for LTO/STO only). The extrapolation is non-trivial: a metallic LSMO cap can change the boundary conditions for ion transport; and the mechanical sign is not obvious, since an increased c spacing of the underlying LTO layer would, if coherently coupled, tend to stretch the overlayer along z rather than compress it. If this strain-orbital pathway does not operate, the explanation for the d10 blue-light photoconductivity reduces to speculation, and the abstract's claim of 'manipulation of orbital occupancy' is unsupported, even though the core observation of gate-tunable photoresistivity/photoconductivity may remain valid.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports transport measurements on La0.7Sr0.3MnO3 (LSMO) thin films grown on SrTiO3 (STO), with and without a 10 unit-cell LaTiO3 (LTO) interfacial layer, under simultaneous illumination (UV 325 nm and blue 441 nm) and back-gate voltage at 300 K. The authors observe photoresistivity at zero gate voltage, which they attribute to photoexcited down-spin eg electrons disrupting double exchange. For nonzero gate voltage, they report a sign inversion (photoconductivity) at negative Vg in the LSMO/STO sample and wavelength-dependent behavior in the LSMO/LTO/STO sample. They propose that UV-driven electron injection from STO into LSMO is blocked by the LTO layer, while blue light excites oxygen-vacancy in-gap states and makes vacancies mobile; under negative Vg, oxygen ions accumulate at the interface, strain the MnO6 octahedra, stabilize the dx2-y2 orbital, and enhance in-plane hopping. Transient current measurements on LTO/STO are used to estimate vacancy mobilities.","tokens_in":13922,"tokens_out":4879,"duration_ms":49214,"significance":"If the mechanism is correct, the work demonstrates simultaneous electrical and optical control of orbital occupancy and charge transport in a manganite at room temperature, which is of potential interest for oxide electronics. The experimental design is logical: the d0/d10 comparison and LTO/STO control measurements are informative, and the transient current data provide a quantitative estimate of vacancy mobility. The paper would be strengthened by direct structural or spectroscopic evidence for the orbital/octahedral changes and by statistical reproducibility of the small resistance changes; nevertheless, the core observations appear internally consistent.","major_comments":[{"comment":"The transient current experiments that provide the vacancy mobility values are performed on a bare LTO(10 u.c.)/STO sample, not on the LSMO/LTO/STO heterostructure (d10) for which the blue-light photoconductivity is observed. The metallic LSMO cap changes the electric-field distribution and the oxygen-exchange boundary conditions, so the extracted mobilities (5.0 x 10^-7 cm2/Vs for blue, 1.5 x 10^-8 for UV, 1.7 x 10^-8 for dark) cannot be directly transferred to the d10 mechanism. The authors should either perform the transient measurement on d10 or explicitly justify why the cap does not alter the ionic transport.","section":"Transient current measurements, Fig. 5"},{"comment":"The strain-orbital chain is the load-bearing explanation for blue-light photoconductivity in d10, but no direct measurement of the orbital occupancy, MnO6 distortion, or even the c-axis change of LTO in the actual stack is provided in this manuscript. Moreover, the sign of the strain transfer is not self-evident: an increase in the LTO c parameter would, if coherently coupled, tend to stretch the LSMO overlayer along z rather than compress it, and the argument that the overlayer is compressed because the in-plane lattice is fixed by STO requires a more careful mechanical model. Direct evidence, such as X-ray diffraction of the stack during illumination, X-ray absorption spectroscopy of the Mn L-edge, or a lattice-constraint calculation, is needed to support the orbital-occupancy claim.","section":"Orbital-occupancy mechanism, discussion of refs. 41 and 42"},{"comment":"The central experimental results are single-trace measurements without error bars or replicate samples. Given that the reported resistance changes are only about +/-7.5% and that gate electric fields can induce leakage or contact effects, the absence of reproducibility data makes it difficult to assess whether the sign inversion and the Vg-independence of the UV response in d10 are robust. The authors should report standard deviations over several cycles or samples, and ideally also show the raw resistance-versus-time traces for the key conditions.","section":"Figs. 2 and 3"}],"minor_comments":[{"comment":"The abstract and text contain grammatical errors: 'it's' should be 'its' in the abstract; 'Keithly' should be 'Keithley' in the experimental section; 'valance' should be 'valence' in the LTO/STO discussion; 'a a polar phase' should be 'a polar phase' in the strain discussion.","section":"Abstract and main text"},{"comment":"The text defines Delta R = R(L) - R(D), but the figures show positive values for photoresistivity; please confirm the sign convention is consistent throughout the paper and in the figure captions.","section":"Definition of Delta R"},{"comment":"The penetration depth argument relies on an absorption coefficient taken from ref. 27, but the specific values used are not stated; a brief note on the extracted alpha values would improve reproducibility.","section":"Penetration depth estimate"},{"comment":"The circuit diagram in the inset of Fig. 5 is not described in the text; please specify explicitly whether the transient current is measured between the gate electrode and a sample contact or through the film.","section":"Fig. 5, inset"},{"comment":"The statement that persistent photoresistivity rules out thermal effects is reasonable, but a quantitative estimate of the laser-induced temperature rise under the stated power densities (680-690 mW/cm2) would make the argument stronger.","section":"Thermal effects"}],"recommendation":"major_revision","confidential_remarks":"The transport observations are potentially interesting, but the mechanism claims are far broader than the direct evidence in the manuscript. The authors should be encouraged either to provide direct orbital/strain measurements or to substantially soften the abstract's 'manipulation of orbital occupancy' statement. The self-citation to related work from the same group (refs. 26, 30, 31, 32) is not problematic, but the novel contribution should be positioned more clearly relative to those papers. The paper might be better suited to a condensed-matter experiment journal that values device-oriented transport demonstrations, provided the mechanistic claims are scaled to the evidence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper reports a real, reproducible-looking set of resistance measurements: at 300 K, light alone gives photoresistivity in LSMO/STO, and adding a back gate changes the sign—negative gate gives photoconductivity, positive gate enhances photoresistivity, reaching about ±7.5%. The addition of an ultra-thin LTO interfacial layer is a smart experiment: it blocks the UV-driven gate injection, leaving only the spin-disorder photoresistivity, while under blue light the gate-dependent effect persists, pointing to a separate oxygen-vacancy pathway. The transient current data on LTO/STO showing a tenfold mobility increase under blue light is also a concrete, independent piece of evidence for photo-activated vacancy motion. These are clean observations and the paper deserves credit for the experimental design.\n\nThe soft spot is the mechanism section. The authors attribute the blue-light photoconductivity in the LSMO/LTO/STO stack to a chain: blue light converts singly to doubly ionized oxygen vacancies in STO, vacancies migrate toward the bottom electrode under negative gate, oxygen ions accumulate at the LSMO/LTO interface, expand the LTO c-axis, compress the LSMO MnO6 octahedra along z, stabilize d_x2-y2, and enhance in-plane hopping. Not one link in this chain is directly measured in this heterostructure. There is no orbital occupancy probe, no lattice parameter measurement on the gated stack, and the transient current (their own data) is on LTO/STO, not on the full LSMO/LTO/STO stack. The stress-test point about the strain sign is also fair: if the LTO c-axis expands, the LSMO overlayer would more naturally be stretched along z, not compressed, unless there is partial relaxation—so the sign of the orbital effect is at least non-obvious. That leaves the \"manipulation of orbital occupancy\" claim in the abstract unsupported by the present evidence, even though the core observation of gate-tunable photoresistivity stands.\n\nMinor issues: no error bars or replicate traces are shown; the key percentages are reported as single values. The mobility estimate from SCL theory is reasonable but depends on assumptions about sample thickness and the nature of the injected species; it would be stronger with more details. Self-citation in the vacancy literature is acceptable and not excessive.\n\nWho is this for? Researchers working on oxide electronics, manganite thin films, or gate-controlled photoinduced effects. It is a useful contribution but not a definitive mechanistic demonstration. A serious referee could help by requesting direct structural/spectroscopic evidence or, failing that, a clear presentation of the mechanism as a hypothesis rather than a conclusion. I would send it to review rather than desk reject.\n\nRecommendation: accept conditional—the experimental observation is worth publishing, but the orbital-occupancy claim needs to be either supported by direct measurements or explicitly framed as a tentative interpretation.","headline":"A solid experimental observation of gate-tunable photoresistivity in LSMO, with a speculative strain/orbital mechanism that needs direct probing before the 'orbital control' claim is accepted.","tokens_in":14434,"tokens_out":2244,"would_cite":false,"duration_ms":21538,"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":"Light and gate voltage together tune the resistance of an LSMO manganite film at 300 K by about ±7.5%, with the sign set by gate polarity.","keywords":["manganite thin films","photoresistivity","photoconductivity","oxygen vacancy migration","orbital occupancy","electric field gating","LSMO","oxide heterostructure"],"falsifier":"Measure the MnO6 octahedral distortion and the e_g orbital occupation of the LSMO overlayer while blue light and a negative gate voltage are applied; if the resistance drops without any detectable out-of-plane octahedral compression or shift toward $d_{x^2-y^2}$ occupation (for example, by X-ray absorption linear dichroism or scanning transmission electron microscopy), the proposed vacancy-strain-orbital chain is not what is producing the photoconductivity.","tokens_in":13412,"feed_emoji":"💡","tokens_out":8983,"duration_ms":76822,"temperature":0.7,"pith_summary":"This paper tries to establish that light and an electric field applied through the substrate can jointly, reversibly tune the resistance of a La0.7Sr0.3MnO3 film at room temperature by roughly ±7.5%, and that the sign of the change is set by the gate-voltage polarity. At zero gate voltage, UV and blue light both raise the resistance; this is attributed to photoexcited $e_g^\\downarrow$ electrons disrupting the magnetic order that enables metallic double-exchange hopping. With a negative gate voltage the response inverts to photoconductivity, explained by injection of photogenerated electrons from the SrTiO3 substrate into the film. Inserting an ultrathin LaTiO3 barrier blocks that injection for UV light, but blue-light photoconductivity survives, which the paper traces to oxygen vacancies in SrTiO3 becoming doubly ionized, migrating, and squeezing the MnO6 octahedra so the $d_{x^2-y^2}$ orbital is stabilized and in-plane hopping is enhanced. The broader payoff claimed is simultaneous control of charge and orbital degrees of freedom in a manganite heterostructure at room temperature.","feed_headline":"Light plus gate voltage flips manganite resistance by 7.5%","feed_subtitle":"The sign of the photoresponse follows gate polarity, pointing to oxygen-vacancy migration and orbital control.","key_machinery":"The load-bearing object is the oxygen vacancy in the SrTiO3 substrate, acting as a photoactivatable, gate-movable source of interfacial strain. Under blue light, singly ionized vacancies give up trapped electrons and become doubly ionized, with a lower diffusion barrier (activation energy 0.6 eV versus 1 eV). The gate field then migrates these vacancies; oxygen ions move in the opposite direction and pile up at the interface, expanding the out-of-plane lattice constant and compressing the MnO6 octahedra in the LSMO overlayer along the film normal. That compression is equivalent to tensile strain and lowers the $d_{x^2-y^2}$ orbital relative to $d_{3z^2-r^2}$, increasing planar hopping and conductivity. A second, injection-based mechanism, gate-controlled transfer of photoexcited electrons from SrTiO3 into LSMO, is identified by its blockade in the presence of an insulating LaTiO3 layer. Transient current measurements under space-charge-limited conditions supply mobility values ($5.0 \\times 10^{-7}$ cm$^2$/Vs for blue light versus $1.5 \\times 10^{-8}$ cm$^2$/Vs for UV) that tie the blue-light effect to mobile vacancies.","core_discovery":"The central discovery is that the photoinduced resistance change in an LSMO heterostructure at 300 K can be switched between positive photoresistivity and negative photoconductivity by the sign of a back gate voltage, giving about ±7.5% resistance modulation. The paper identifies two wavelength-dependent routes behind this. UV light (325 nm, 3.8 eV) generates electron–hole pairs in the SrTiO3 substrate; a negative gate voltage injects those electrons into the half-metallic LSMO film, lowering resistance, while a positive gate voltage blocks injection and traps interfacial conduction electrons, raising it. Blue light (441 nm, 2.8 eV) is below the SrTiO3 band gap and instead excites in-gap oxygen-vacancy states; the doubly ionized vacancies migrate under the gate field, oxygen ions accumulate at the interface, compress MnO6 octahedra along the film normal, stabilize the $d_{x^2-y^2}$ orbital, and thereby increase in-plane hopping. The LaTiO3 interlayer experiments are the key evidence: a 10-unit-cell LTO layer suppresses the gate-dependent UV response, showing that injection is interfacial and blocked, while blue-light photoconductivity persists, pointing to vacancy diffusion rather than injection.","pith_inferences":["If the vacancy-strain-orbital chain holds, the same heterostructure design should work in other strongly correlated oxide overlayers on SrTiO3, allowing orbital occupancy to be written optically at room temperature without changing chemical doping.","The two wavelength-separated mechanisms behave like an optical AND condition: a resistance change appears only when light of the right wavelength and a gate bias are present together, which could be exploited in low-power oxide logic or memory.","A direct extension would be to vary laser power density and film thickness to see whether the ±7.5% ceiling is set by the supply of mobile vacancies or by the orbital susceptibility of the manganite.","Patterning the back gate could make vacancy migration lateral rather than vertical, creating locally strained metallic and insulating regions in a single LSMO film."],"forward_implications":["At room temperature, an LSMO film's resistance can be modulated by about ±7.5% using light plus a back gate, with the sign of the response set by gate polarity.","A 10-unit-cell LaTiO3 interlayer blocks the gate-dependent UV photoresponse, confirming that the UV gating channel is interfacial carrier injection from SrTiO3.","Blue-light photoconductivity that survives the LaTiO3 barrier indicates a second, non-injection channel: light-activated oxygen-vacancy diffusion that alters orbital occupancy in the manganite.","Transient current data show oxygen-vacancy mobility increases by about an order of magnitude under blue illumination, consistent with doubly ionized vacancies being the mobile species.","The combination of gate voltage and wavelength selects between spin-disorder photoresistivity, interfacial electron injection, and orbital-occupancy strain, giving multiple handles on the same resistance state."],"supporting_citations":[{"why":"supplies the pump-probe evidence that $e_g^\\downarrow$ photoexcitation creates long-lived spin disorder, the basis for photoresistivity at zero gate voltage.","marker":"[22]"},{"why":"provides the absorption coefficients used to argue that UV and blue light penetrate through the LSMO layer into the SrTiO3 substrate.","marker":"[27]"},{"why":"reports comparable photoinduced electron injection at an LSMO/STO interface, used as precedent for the negative-gate injection mechanism.","marker":"[28]"},{"why":"establishes that oxygen vacancies donate electrons to the interface and become mobile under light and gate bias in titanate heterostructures.","marker":"[35]"},{"why":"gives the activation energies for singly versus doubly ionized oxygen vacancies and the light-induced conversion that underpins the blue-light diffusion mechanism.","marker":"[36]"},{"why":"shows in gated SrTiO3 that vacancy migration accumulates oxygen ions and elongates TiO6 octahedra, the strain source for the orbital mechanism.","marker":"[41]"},{"why":"links out-of-plane octahedral compression in manganite overlayers to $d_{x^2-y^2}$ stabilization and enhanced in-plane hopping.","marker":"[42]"},{"why":"provides the space-charge-limited transient current formula used to extract oxygen-vacancy mobility from the measured current peaks.","marker":"[47]"}],"fun_headline_variants":["Gate polarity flips light-driven manganite resistance","UV and blue light switch manganite resistance with gate","Light plus gate voltage: ±7.5% resistance swings in manganite","Interfacial layer reveals two photo-gate effects in manganite","Photoinduced resistance in manganite controlled by gate sign"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The fragile link is the claim that oxygen vacancies migrating in the SrTiO3 substrate squeeze the MnO6 octahedra in the LSMO overlayer and change its orbital occupancy; this is inferred from other oxide systems and is not directly measured in this heterostructure.","fun_headline_variants_meta":{"raw":{"variants":["Gate polarity flips light-driven manganite resistance","UV and blue light switch manganite resistance with gate","Light plus gate voltage: ±7.5% resistance swings in manganite","Interfacial layer reveals two photo-gate effects in manganite","Photoinduced resistance in manganite controlled by gate sign"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000228,"raw_usage":{"total_tokens":1485,"prompt_tokens":968,"completion_tokens":517,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":584,"completion_tokens_details":{"reasoning_tokens":429}},"tokens_in":584,"tokens_out":517,"duration_ms":5183,"temperature":1.0,"reasoning_tokens":429,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:42:48.585420+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the MnO6 octahedral distortion and the e_g orbital occupation of the LSMO overlayer while blue light and a negative gate voltage are applied; if the resistance drops without any detectable out-of-plane octahedral compression or shift toward $d_{x^2-y^2}$ occupation (for example, by X-ray absorption linear dichroism or scanning transmission electron microscopy), the proposed vacancy-strain-orbital chain is not what is producing the photoconductivity.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the pump-probe evidence that $e_g^\\downarrow$ photoexcitation creates long-lived spin disorder, the basis for photoresistivity at zero gate voltage."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the absorption coefficients used to argue that UV and blue light penetrate through the LSMO layer into the SrTiO3 substrate."},{"cited_title":"Rastogi, J","cited_arxiv_id":null,"evidence_quote":"reports comparable photoinduced electron injection at an LSMO/STO interface, used as precedent for the negative-gate injection mechanism."},{"cited_title":"Kalabukhov, R","cited_arxiv_id":null,"evidence_quote":"establishes that oxygen vacancies donate electrons to the interface and become mobile under light and gate bias in titanate heterostructures."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"gives the activation energies for singly versus doubly ionized oxygen vacancies and the light-induced conversion that underpins the blue-light diffusion mechanism."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"shows in gated SrTiO3 that vacancy migration accumulates oxygen ions and elongates TiO6 octahedra, the strain source for the orbital mechanism."},{"cited_title":"Lin and A","cited_arxiv_id":null,"evidence_quote":"links out-of-plane octahedral compression in manganite overlayers to $d_{x^2-y^2}$ stabilization and enhanced in-plane hopping."},{"cited_title":"Liu, T.-C","cited_arxiv_id":null,"evidence_quote":"provides the space-charge-limited transient current formula used to extract oxygen-vacancy mobility from the measured current peaks."}],"review_version":1}