{"id":"9509961b-f30a-4dcb-a357-b7beee51b430","arxiv_id":"2608.11832","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"Light-tunable threshold switching in planar Ag/MAPbI3 thin single-crystal devices is attributed to interface-mediated barrier modulation involving Ag+ migration, supported by back-to-back Schottky fits and control devices.","lead":"This paper reports planar devices made from thin methylammonium lead iodide (MAPbI3) single crystals with silver contacts that show a light-tunable, abrupt switch between low and high conductance states. The authors attribute the switching to the silver/perovskite interface, where light and voltage drive ionic rearrangement that modulates the contact barrier.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central Ag+ migration claim rests on a single weak EDS observation; without causal evidence linking that Ag signal to light/voltage-driven migration and to the switching event, the core mechanism is underdetermined.","rationale":"The reader's weakest-assumption analysis identified the same load-bearing concern: the causal link between the observed Ag signal and the switching mechanism is not established. I agree with that assessment. The paper is otherwise well constructed: the single-crystal platform is appropriate, the control experiments with Au and graphite contacts support a contact-mediated effect, and the back-to-back Schottky model captures the qualitative I-V asymmetry. However, the specific claim of Ag+ migration is supported only by a weak, unquantified EDS observation on one device, and the paper itself acknowledges limited reproducibility. These issues do not require rejection; the phenomenon and its interfacial nature remain plausible. The appropriate outcome is to keep the verdict conditional, requiring direct chemical or transport evidence that Ag+ migration is the causal agent. Since the reader already issued a conditional verdict, no adjustment is needed. The proposed ToF-SIMS/STEM-EDS experiment would settle whether the residual Ag signal is specific to the switched state and reversible, which is the key missing piece.","tokens_in":14111,"tokens_out":3177,"duration_ms":37189,"concrete_test":"Perform ToF-SIMS or cross-sectional STEM-EDS mapping on at least three independent devices in each of four conditions: (1) pristine, (2) illuminated but not biased, (3) biased in the dark, and (4) illuminated and biased into the low-resistance state. Define a quantitative Ag-excess metric as the integrated Ag signal in a 10 micrometer region adjacent to the contact minus the background level, expressed in units of measurement noise. If condition (4) does not show a statistically significant (≥3 sigma) Ag excess relative to conditions (1)-(3), or if conditions (2) or (3) show comparable excess, then the EDS-based link between switching and Ag+ migration fails. Additionally, repeating the measurement after a reverse sweep and showing that the Ag signal returns to background would directly test the reversibility invoked in the model.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that the threshold transition is caused by light-assisted Ag+ migration at one Ag/MAPbI3 interface, producing an abrupt, reversible change in that contact's effective barrier. The only direct evidence for Ag+ migration is the EDS analysis in SI S.4, which reports a 'weak residual Ag signal' in the perovskite adjacent to the contact after electrical stress. This observation is not sufficient to establish the claimed mechanism. The EDS data come from a single stressed device, with no quantitative threshold for the Ag signal relative to background, no error analysis, and no comparison to pristine or control devices subjected to the same bias/illumination history. The residual Ag could instead arise from smearing of the Ag paste during handling, from the rough contact edge, or from beam-induced contamination. Furthermore, the back-to-back Schottky fits and the Au/graphite control devices show that the behavior is contact-related, but they do not distinguish Ag+ migration from alternative interfacial mechanisms such as light-activated trap reconfiguration, photoelectrochemical modification of the contact, or migration of intrinsic iodine vacancies. The paper's own admission that only three of five devices showed the reported switching, with one stochastic and one non-switching, further weakens any implicit correlation between Ag presence and switching. Because the conclusion explicitly invokes Ag+ migration and interface-mediated barrier modulation, the EDS observation is the keystone of the mechanistic argument; without a causal test, the central claim is plausible but not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports the growth of thin MAPbI3 single crystals by a space-confined inverse-temperature method and their integration into planar two-terminal devices with directly deposited Ag contacts. In the dark, the devices show ultra-low currents and negligible hysteresis; under illumination, the I-V characteristics develop polarity-dependent hysteresis and a threshold-like transition between high- and low-resistance states, with the switching voltage shifting with optical power. Temperature-dependent dark measurements show thermally activated, contact-limited transport, and the authors fit a back-to-back Schottky diode model to the illuminated I-V curves. Control devices with Au and graphite contacts show no such switching, and EDS on one stressed device shows a weak residual Ag signal near the contact. The central claim is that switching is governed by interface-mediated barrier modulation, plausibly involving light-assisted Ag+ migration at one Ag/MAPbI3 interface, rather than by a spatially homogeneous bulk process.","tokens_in":14352,"tokens_out":5216,"duration_ms":56519,"significance":"If the central claim holds, the work provides a useful single-crystal platform for studying light-tunable resistive switching in halide perovskites with a contact-controlled mechanism, and it adds to the growing body of evidence that metal/perovskite interfaces, rather than bulk filaments, dominate memristive behavior in these systems. The paper's strengths include the high-quality single-crystal growth, the explicit control experiments with inert contacts, the polarity-inversion test, and the honest disclosure that only three of five devices showed the reported switching. However, the specific mechanistic attribution to Ag+ migration rests on limited evidence, and several quantitative claims lack statistical support.","major_comments":[{"comment":"The mechanistic conclusion that switching is caused by light-assisted Ag+ migration at one Ag/MAPbI3 interface rests primarily on the EDS analysis in SI S.4, but that analysis is not quantitative: it is performed on a single stressed device, reports only a 'weak residual Ag signal' with no threshold value relative to background, and lacks control measurements on an unstressed device or on a device subjected to the same voltage/illumination protocol without switching. As stated, the EDS evidence cannot exclude Ag paste smearing or beam-induced artefacts, and it does not causally link the residual Ag to the switching event. Please provide quantitative EDS with matched controls, or downshift the conclusion from Ag+ migration to a more general interfacial ionic/electrochemical mechanism.","section":"§2.2 and SI S.4"},{"comment":"Table S6 shows that the LRS fit is achieved by decreasing Rs by a factor of 3.5 and by reducing only one of the two Schottky barriers. Since the text defines Rs as the voltage drop along the MAPbI3 channel, this fit is equally consistent with a bulk channel conductance change as with an interfacial barrier change; it therefore does not by itself support the conclusion in §3 that switching is 'governed primarily by interface-mediated barrier modulation rather than by a spatially homogeneous bulk process.' The authors should either separate the contact and bulk contributions (e.g., four-point or variable channel-length measurements) or explicitly temper the claim.","section":"§2.2, Table S6"},{"comment":"The monotonic trends of integrated hysteresis Ahyst and switching voltage Vsw with incident power are reported for a single representative device without error bars or statistics. Given that only three of five devices exhibited switching (and one of those with poor light control, as stated in §2.2), the generality of these trends is not established. Please include device-to-device statistics (at least for the three switching devices) or clearly present the data as single-device observations with correspondingly qualified language.","section":"§2.2, Fig. 3e,f"},{"comment":"The abstract and conclusions refer to a 'normalized integrated-hysteresis metric' that varies weakly with temperature, but §2.3 defines and plots only the unnormalized hysteresis areas Ahyst+ and Ahyst−. The normalization procedure is never specified, so the claim cannot be checked. Please define the normalized metric and report it, or correct the abstract/conclusions to refer to the metric actually measured.","section":"Abstract and §2.3/§3"}],"minor_comments":[{"comment":"There are multiple typos and inconsistent spellings, e.g., 'cantered' for 'centered' (§2.1), 'MaPbI3' for 'MAPbI3' in the Figure 1a caption, 'detectectable' for 'detectable' in SI S.4, and 'behaviour behavior' inconsistencies.","section":"Throughout"},{"comment":"The figure numbering in the SI is inconsistent: section S.5 (Richardson analysis for negative-bias branch) refers to 'Figure S4', but the correct figure number is S5, as also referenced in the main text.","section":"SI S.5"},{"comment":"The sentence 'The substrates are then be separated using a blade' should read 'The substrates are then separated using a blade.'","section":"§2.1"},{"comment":"The notation '10-13-10-12 A' should be typeset with superscripts and an en dash: '10^−13–10^−12 A'.","section":"§2.2"},{"comment":"The tetragonal-to-cubic phase transition is mentioned as spanning the measurement window, but no explicit discussion is given of how the phase transition might affect the extracted Schottky barriers or the hysteresis; adding a sentence of clarification would strengthen the analysis.","section":"§2.3"}],"recommendation":"major_revision","confidential_remarks":"The paper is in scope for a condensed-matter/physical-chemistry journal and the experimental work appears carefully done in many respects. My main reservation is the gap between the strong mechanistic conclusion (Ag+ migration and interface-mediated barrier modulation) and the evidence actually presented: the EDS observation is a single, non-quantitative measurement, and the model fits are ambiguous about the bulk versus contact nature of the resistance change. The reproducibility caveat (3/5 devices) is disclosed, but the quantitative trends in Fig. 3e,f are not placed in that statistical context. These issues are fixable with additional measurements or by tempering the mechanistic language; they do not require rejecting the manuscript."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, this is a genuinely useful experimental paper: it shows light-tunable threshold switching in planar, thin MAPbI3 single crystals with Ag contacts, and it makes a good case that the switching is governed by the Ag/perovskite contacts rather than a bulk filament. Second, the more specific claim that Ag+ migration at one interface is the causal ionic process is plausible but not established; the only direct evidence is a weak EDS signal in one stressed device.\n\nWhat is new: the space-confined single-crystal growth gives clean, low-defect channels; dark currents are in the 10^-13 to 10^-12 A range; under illumination the I-V curves develop a pronounced, polarity-dependent hysteresis and a threshold transition that scales with optical power. The back-to-back Schottky model fits the curves reasonably. The controls are the right ones: Au and graphite contacts do not show the switching, and polarity inversion follows the contacts. The authors are also honest about the limits: only three of five devices switched, one was stochastic, the fit parameters are explicitly called effective, and Vsw is stated to be scan-rate dependent. That honesty is not just cosmetic; it makes the paper easier to trust.\n\nNow the soft spots. The Ag+ migration story rests almost entirely on the EDS line profile in SI S.4: a residual Ag signal near the contact after stress, in one device, with no background threshold, no error analysis, and no comparison with a stressed device that did not switch. That is not enough to distinguish Ag+ migration from light-activated trap reconfiguration, iodine vacancy migration, or a photoelectrochemical change at the contact. The controls rule out a purely bulk mechanism, but they do not identify the species. The paper's own language mostly stays cautious (suggesting, more likely), but the conclusion states that switching is governed primarily by interface-mediated barrier modulation — that part I can accept; the specific Ag+ identification is a step beyond the data.\n\nTwo smaller issues. The abstract and conclusions mention a normalized integrated-hysteresis metric that varies weakly with temperature, but the body only defines the unnormalized hysteresis area; the normalized version is never shown. And the main figures lack error bars or device-to-device statistics, which matters when reproducibility is three-out-of-five.\n\nFor whom: anyone working on perovskite memristors, light-tunable switching, or contact engineering in halide perovskites will want to read this. It is a solid, well-written addition to the literature, not a paradigm shift. I would send it to a serious referee; the right response is revision, not rejection. Ask for a cleaner causal test of Ag migration (for example, a switched vs non-switched device comparison, or a better probe), the normalized metric, and statistics across devices.","headline":"A careful, honest experimental study of light-tunable threshold switching in planar Ag/MAPbI3 single crystals; the interface-mediated conclusion holds, but the Ag+ migration mechanism rests on a slimmer EDS pillar than the prose suggests.","tokens_in":15029,"tokens_out":3086,"would_cite":true,"duration_ms":31612,"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-tunable threshold switching in planar Ag/MAPbI3 single-crystal devices is governed by the Ag/perovskite interfaces, not by bulk or filamentary processes.","keywords":["MAPbI3","single crystal","threshold switching","hysteresis","ionic migration","Schottky barrier","memristor","light-tunable"],"falsifier":"Measure a switched device with a technique sensitive to buried silver, such as depth-resolved ToF-SIMS or cross-sectional elemental mapping: if the low-resistance state is reached without any detectable silver accumulation at the switched interface, the Ag+ migration claim fails. Equivalently, if an ion-blocking barrier layer between Ag and MAPbI3 still allows the photoactivated threshold transition, the mechanism is not silver migration.","tokens_in":13866,"feed_emoji":"⚡","tokens_out":8817,"duration_ms":81462,"temperature":0.7,"pith_summary":"This paper reports that thin MAPbI3 single-crystal devices with silver contacts can be switched between a high- and a low-resistance state by light, and argues that the switch happens at the Ag/perovskite interface rather than through the crystal bulk. In the dark, the devices conduct almost nothing and show no hysteresis; under illumination they develop a sharp, polarity-dependent threshold transition whose size and switching voltage move with optical power. The authors fit the current–voltage curves with a back-to-back Schottky diode model in which one of the two Ag contacts abruptly becomes more conductive while the other stays fixed. If correct, this makes Ag contact engineering a route to optically tunable, ion-based memory devices in halide perovskites.","feed_headline":"Silver contacts, not the bulk, control switching in perovskite devices","feed_subtitle":"Light lowers one contact barrier to flip the device between two conductance states, a step toward optical memristors.","key_machinery":"The load-bearing object is the back-to-back Schottky diode model: the device is represented as two rectifying Ag/MAPbI3 junctions in series with a resistive channel. The forward and reverse current branches are fit with the double-Schottky I–V expression and a series resistance $R_s$. Fits of the high-resistance and low-resistance states show that switching is captured mainly by a drop in one barrier height and in $R_s$, while the second barrier remains nearly constant. The paper's mechanistic link is light-assisted Ag$^+$ migration at that one contact, which alters the barrier and makes the junction quasi-ohmic; the weakly varying normalized hysteresis metric in temperature-dependent dark measurements places the ionic/interfacial dynamics as a secondary, thermally activated effect.","core_discovery":"The central discovery is a light-assisted, contact-controlled threshold switching mechanism in planar two-terminal MAPbI3 single-crystal devices with directly deposited silver electrodes. The paper shows that illumination converts an almost perfectly insulating, hysteresis-free dark device into one with pronounced polarity-dependent hysteresis and a sharp transition between two conductance states. Control devices with Au and graphite contacts show photoconductivity but no threshold transition, and EDS maps after electrical stress show weak residual silver near the switched contact. Combining these observations with fits to a back-to-back Schottky model, the authors conclude that the transition is not a bulk filament but an abrupt, voltage- and light-assisted lowering of the effective Schottky barrier at one Ag/MAPbI3 interface, driven by interfacial charging and silver-ion migration. The same framework accounts for the temperature-dependent dark transport, which is thermionically activated with effective barriers between about 0.44 and 0.6 eV, decreasing as the bias increases.","pith_inferences":["A testable extension: if Ag+ migration is the cause, endurance should be finite and limited by the silver reservoir at the switched contact.","The same planar geometry could be used with contacts made of other easily ionized metals, such as copper, to test whether the light-tunable switching generalizes beyond silver.","A three-terminal or impedance-spectroscopy measurement could resolve the hidden reverse transition that the second reverse-biased diode currently masks in two-terminal I–V curves.","The 3-of-5 device yield hints that switching depends on the microscopic quality of the as-deposited silver–perovskite interface; correlating yield with interface morphology would show whether this variability is fundamental or process-limited."],"forward_implications":["Increasing optical power widens the hysteresis loop and shifts the switching voltage toward more negative bias, so light can set both the size and the threshold of the memory window.","The back-to-back Schottky analysis identifies the high-to-low resistance transition with a drop in one effective barrier height and in series resistance, so device state can be monitored through those parameters.","Control devices with Au and graphite contacts show no photoactivated threshold transition, so the switching is specific to the silver contact chemistry and not a bulk single-crystal property.","Temperature-dependent dark transport follows a Richardson-type law with effective barriers from roughly 0.44 to 0.6 eV, indicating contact-limited thermionic emission rather than bulk-limited conduction.","Because the channel is hundreds of micrometres long and currents stay low, the transition is not a metallic filament through the perovskite."],"supporting_citations":[{"why":"Provides the two-diode I–V model used to fit both high- and low-resistance branches.","marker":"[31]"},{"why":"Supports treating Ag/MAPbI3 contacts as Schottky diodes with barrier-dependent transport.","marker":"[30]"},{"why":"Recent demonstrations that Ag-based electrode engineering controls contacts in halide-perovskite devices.","marker":"[14,15]"},{"why":"Source for light-assisted formation and dynamics of mobile Ag+ and iodine species at perovskite interfaces.","marker":"[26]"},{"why":"Supports the claim that surfaces and interfaces dominate hysteresis in hybrid perovskites.","marker":"[37]"},{"why":"General memristive framework in which diffusion-controlled internal dynamics enhance hysteresis, used to interpret temperature trends.","marker":"[11]"},{"why":"Identifies Au and graphite as relatively inert contact materials, motivating the control-device comparison.","marker":"[27,28]"}],"fun_headline_variants":["Light flips perovskite switch via silver interface","Silver contacts, not filaments, drive light-triggered switching","Optical barrier lowering enables perovskite threshold switching","Light-assisted contact barrier controls perovskite conductance states"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that the weak silver signal seen near the contact after switching is the cause of the lowered barrier, not a byproduct of the switched state; if Ag+ migration is not the actual trigger, the interface-modulation story loses its microscopic mechanism.","fun_headline_variants_meta":{"raw":{"variants":["Light flips perovskite switch via silver interface","Silver contacts, not filaments, drive light-triggered switching","Optical barrier lowering enables perovskite threshold switching","Light-assisted contact barrier controls perovskite conductance states"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000231,"raw_usage":{"total_tokens":1511,"prompt_tokens":998,"completion_tokens":513,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":614,"completion_tokens_details":{"reasoning_tokens":453}},"tokens_in":614,"tokens_out":513,"duration_ms":5644,"temperature":1.0,"reasoning_tokens":453,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:25:10.429276+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure a switched device with a technique sensitive to buried silver, such as depth-resolved ToF-SIMS or cross-sectional elemental mapping: if the low-resistance state is reached without any detectable silver accumulation at the switched interface, the Ag+ migration claim fails. Equivalently, if an ion-blocking barrier layer between Ag and MAPbI3 still allows the photoactivated threshold transition, the mechanism is not silver migration.","supporting_citations":[],"review_version":1}