{"id":"ae633d44-fe83-4d1d-a11b-c43cb01f7925","arxiv_id":"2411.14670","paper_version":1,"verdict":"UNVERDICTED","confidence":"MODERATE","novelty_score":2.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A perspective review arguing that hybrid ferroelectric tunnel junctions and ferroelectric resonant tunneling diodes offer polarization-controlled tunneling effects useful for future nonvolatile electronics.","lead":"This preprint surveys hybrid ferroelectric tunnel junctions, in which a ferroelectric layer works together with dielectric, multiferroic, superconducting, or 2D layers, and reviews ferroelectric resonant tunneling diodes. It is a perspective rather than a new experiment, aimed at researchers choosing directions in nonvolatile memory and oxide nanoelectronics.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"FeRTD outlook hinges on the BTO/SRO/BTO demonstration [188]; because the text itself concedes redox/defect mechanisms in adjacent systems, the ferroelectric origin of that room-temperature NDR must be established before the central perspective can carry its weight.","rationale":"The concern is load-bearing because the abstract's promise is specifically about room-temperature ferroelectric control of the resonant peak, the peak tunnel-current ratio, and NDR, and the perspective provides no primary data to establish that control; it relies on cited demonstrations. The paper deserves credit for honestly flagging non-ferroelectric mechanisms in adjacent systems and for correctly presenting much of the FeRTD work as theoretical, but that honesty makes the missing control check for [188] more salient, not less. The reader's weakest_assumption already identified the same point; my attack sharpens it to the single key device [188] and to the specific absence of an STO control stack. Because the manuscript is a perspective/review rather than a primary claim, I would not change the verdict from UNVERDICTED; the concern should be recorded as a condition on how much weight the FeRTD outlook can carry.","tokens_in":33276,"tokens_out":6802,"duration_ms":73972,"concrete_test":"Fabricate two nominally identical stacks under the same deposition conditions: BTO(barrier)/SRO(well)/BTO(barrier) as in [188], and a control with non-ferroelectric STO barriers of the same thickness replacing BTO. Run the same poling and I-V protocol on both. If the control shows comparable NDR and OFF/ON ratios, or if the BTO device's NDR peak voltage does not track the polarization state verified by piezoelectric force microscopy on the same device, then the ferroelectric-control interpretation is not supported; if the control is flat and the peak shifts reproducibly with switched polarization, the central concern is resolved. An additional oxygen-annealing variant would separate oxygen-vacancy contributions.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that ferroelectric polarization can control resonant tunneling and negative differential resistance at room temperature, making FeRTDs a radically new device class. For that claim to hold, the cited room-temperature FeRTD demonstrations must genuinely be ferroelectric-controlled resonant tunneling. The paper's own text weakens this premise. Section 2.4 concedes that in MoSi/YBCO and MoSi/BFO/YBCO junctions ferroelectricity 'did not play a central role' and that redox reactions, oxygen-vacancy motion, and electrochemical interface changes dominate the resistance switching in superconducting-electrode junctions, including YBCO/Al and YBCO/ITO. Section 2.5a concedes that the early FeRTD proposals based on Li-doped ZnO are questionable because the observed hysteresis may be defect-mediated. Once those cases are set aside, the room-temperature experimental base for the FeRTD proposition is effectively the authors' own BTO/SRO/BTO device [188] and the BTO-well device [180]. For [188], the text states that NDR was 'attributed to the growth of orthorhombic SRO and the electron-electron correlations between its Ru-t2g and Ru-eg states,' and this perspective reports no control experiments that exclude non-ferroelectric mechanisms such as oxygen-vacancy migration, interface redox, or trap-assisted conduction. If those mechanisms, rather than BTO polarization, set the NDR peak and the OFF/ON ratio, the central proposition is not established by the cited evidence.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript is a perspective/review of hybrid ferroelectric tunnel junctions (FTJs), defined as junctions in which electrons tunnel through a functional layer in series with an ultrathin ferroelectric barrier. It surveys ferroelectric/dielectric composite-barrier junctions, ferroelectric/multiferroic junctions, ferroelectric/2D-material junctions, ferroelectric/superconductor junctions, and ferroelectric resonant tunneling diodes (FeRTDs). The central claim is that ferroelectric polarization can control resonant tunneling and negative differential resistance (NDR) at room temperature, so that FeRTDs constitute a new device class for nonvolatile memory and neuromorphic computing. The review explicitly acknowledges some caveats, including redox-dominated switching in YBCO-based superconductor junctions and doubts about Li-doped ZnO ferroelectricity, but the FeRTD outlook is built primarily on the authors' own BTO/SRO/BTO demonstration [188].","tokens_in":33527,"tokens_out":8265,"duration_ms":77210,"significance":"As a synthesis, the paper is useful: it brings together a wide literature, is mostly traceable to peer-reviewed sources, and is candid about electrochemical complications in superconducting and oxide systems. The conceptual framing of hybrid FTJs and FeRTDs is clear, and the figures are generally informative. The significance of the central claim is high if it holds: polarization-controlled room-temperature resonant tunneling would indeed add a distinct functionality to oxide electronics. However, the central perspective is only as strong as the experimental evidence for ferroelectric control of NDR, and that evidence is concentrated in the authors' own previous work; the review does not present independent corroboration or new mechanistic analysis. For this reason the manuscript needs revision before the claim can carry the weight the authors place on it.","major_comments":[{"comment":"The paper's core proposition—that ferroelectric polarization controls resonant tunneling and NDR at room temperature—is not established by the cited evidence. The only room-temperature FeRTD demonstration is the BTO/SRO/BTO device reported by the authors in [188], and the text states that NDR was 'attributed to the growth of orthorhombic SRO and the electron-electron correlations between its Ru-t2g and Ru-eg states.' No control experiments or independent measurements are described that exclude non-ferroelectric mechanisms such as oxygen-vacancy migration, interface redox, or trap-assisted conduction. Given that Section 2.4 concedes that redox reactions and oxygen stoichiometry dominate resistance switching in YBCO-based junctions, and Section 2.5a concedes that the Li-doped ZnO FeRTDs are questionable because their hysteresis may be defect-mediated, the room-temperature experimental base for the FeRTD proposition is too thin. Please either add a critical discussion of the mechanistic evidence in [188] (e.g., polarization-dependence of the NDR peak, retention/endurance data, comparison with non-ferroelectric control samples) or soften the central claim to a perspective-level proposal rather than an established phenomenon.","section":"Section 2.5b and Section 4"},{"comment":"The sentence 'In 2022, we resolved this longstanding problem by fabricating a FeRTD with a QW heterostructure comprised of BTO/SRO/BTO' is a first-person, self-referential claim that is not appropriate in a review without the same critical scrutiny applied to other cited works. The phrase 'we resolved' also overstates what a single demonstration establishes, especially when the accompanying assertion of 'the largest OFF/ON resistance ratios values in tunneling devices that utilize BTO as the barriers' is itself drawn from [188]. Please rephrase neutrally and state explicitly what [188] does and does not prove.","section":"Section 2.5b"},{"comment":"The paper defines FeRTDs as not strictly differentiated from FTJs that exhibit resonant tunneling. This ambiguity makes the central claim difficult to falsify: if any FTJ showing a current peak is labelled a FeRTD, then the 'radically new approach' becomes a matter of terminology rather than a new physical regime. Please clarify the operational distinction, for example by specifying the required double-barrier/quantum-well geometry and the criterion for ferroelectric control of the resonance.","section":"FeRTD section (unnumbered, after Section 2.4)"}],"minor_comments":[{"comment":"The Introduction promises that FeRTDs are reviewed in Section 3, but the FeRTD material appears in an unnumbered section with subsections 2.5–2.7, and the following sections are numbered 3 and 4; please renumber consistently.","section":"Introduction and section numbering"},{"comment":"The caption says 'BTO/STO/BTO FeRTD' while the text and Section 2.5b describe BTO/SRO/BTO; correct the mismatch.","section":"Figure 11b caption"},{"comment":"The abstract contains a grammatically incomplete phrase: 'enabling further modalities and functional capabilities than in addition to tunneling electroresistance.'","section":"Abstract"},{"comment":"The text lists 'MoS2, WeS' among 2D materials; this should be 'MoS2, WSe2' (or the intended material written correctly).","section":"Section 2.3 (introductory paragraph)"},{"comment":"The phrase 'Si or Ge substate' should be 'Si or Ge substrate'.","section":"Section 2.1c"},{"comment":"The caption describes panels a–c and then e) for the I–V curve, but no panel d) is identified; the panel labels should be corrected.","section":"Figure 9 caption"},{"comment":"Reference [190] lacks a year in the reference list; the citation should be completed.","section":"Reference list"}],"recommendation":"major_revision","confidential_remarks":"The paper's perspective is built heavily on the authors' own prior result [188]; this is not improper, but the review should be explicit about the evidentiary status of that result and should not present it as fully resolved. I recommend major revision rather than rejection because the issue is fixable by reframing the FeRTD claims, adding mechanistic caveats, and clarifying the definition. The manuscript fits the journal's scope as a perspective."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a competent, readable review of hybrid FTJs and ferroelectric resonant tunneling diodes, and the \"hybrid junction\" definition plus the FeRTD taxonomy (ferroelectric well, ferroelectric barrier, domain-wall well) are genuinely useful organizing frames. The literature coverage is broad and mostly traceable. But the paper's headline claim—that resonant tunneling offers a radically new approach and that FeRTDs hold immense potential—rests on a very thin experimental base, and the paper's own text concedes most of that thinness.\n\nWhat the paper does well: it systematically walks through ferroelectric/dielectric, ferroelectric/multiferroic, ferroelectric/2D and ferroelectric/superconductor junctions, and it does not hide the messy physics. Section 2.4 is candid that MoSi/YBCO and MoSi/BFO/YBCO switching is dominated by redox/oxygen-vacancy effects rather than ferroelectric polarization, and Section 2.5a flags the Li-doped ZnO controversy. The discussion of 2D vdW ferroelectrics and sliding ferroelectricity is up to date. Citation practice looks honest; self-citations are not excessive for a perspective, though the phrase \"we resolved this longstanding problem\" in Section 2.5b gives the authors' own BTO/SRO/BTO device a prominence that would be fine in a research proposal but is slightly pushy in a review.\n\nWhere it is soft: the central FeRTD outlook depends on exactly two room-temperature experimental demonstrations—the BTO-well device [180] and the authors' BTO/SRO/BTO device [188]—and for [188] the NDR is attributed partly to the orthorhombic SRO layer and electron-electron correlations, not purely to ferroelectric switching. That's not a fatal flaw in a perspective, but it means the \"radically new approach\" language in the abstract and conclusion oversells what the evidence supports. The paper would be stronger if it explicitly said that the ferroelectric origin of the room-temperature NDR in [188] has not yet been isolated from possible ionic/trap contributions. Also, the section numbering is inconsistent (FeRTDs are introduced as both Section 2 and Section 3; the introduction promises Section 3 but the text jumps from 2.7 to \"3 Challenges\"). That's cosmetic but it makes the manuscript look less finished than it is.\n\nBottom line: this deserves a serious referee. A moderately critical reviewer can push for tempered claims and a fixed structure; the underlying synthesis is useful and the caveats are already mostly in the text. I'd be happy to read a revised version, and I'd bring it to a reading group focused on oxide tunneling devices. I would not, however, use it as evidence for polarization-controlled NDR without going back to the primary data.\n\nRecommendation: send out for review, with a note to the authors that the abstract/conclusion should match the hedged body.","headline":"Useful, honest review of hybrid FTJs and FeRTDs, but the 'radically new' FeRTD claim rests on a thin experimental base that the text itself concedes.","tokens_in":34108,"tokens_out":2988,"would_cite":true,"duration_ms":29640,"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":"A review argues that ferroelectric polarization can control resonant tunneling and negative differential resistance at room temperature, opening a route to low-power nonvolatile memory and neuromorphic devices.","keywords":["ferroelectric tunnel junction","resonant tunneling","negative differential resistance","tunneling electroresistance","quantum well","2D van der Waals ferroelectrics","multiferroic tunnel junction","hafnia"],"falsifier":"A decisive test is to fabricate the same junction stack with the ferroelectric layer replaced by a non-polar insulator of similar thickness and dielectric constant, such as SrTiO3 in place of BaTiO3, and to measure the NDR peak voltage and OFF/ON ratio: if the NDR features persist with comparable magnitude and shift only with bias polarity rather than with poling history, the ferroelectric-switching mechanism is falsified. A complementary observation is to correlate the ON/OFF state and the NDR peak shift with a direct piezoresponse force microscopy hysteresis measurement of the active device area: the absence of a switchable out-of-plane polarization in the device region would falsify the central claim.","tokens_in":33037,"feed_emoji":"⚡","tokens_out":4976,"duration_ms":47888,"temperature":0.7,"pith_summary":"This perspective argues that the next step beyond simple ferroelectric tunnel junctions is the hybrid junction, where electrons tunnel through a functional layer in series with an ultrathin ferroelectric film, and beyond that, the ferroelectric resonant tunneling diode, where a quantum well replaces the single ferroelectric barrier. The paper's central proposition is that resonant tunneling offers a radically new way to exploit tunneling in ferroelectric heterostructures, because ferroelectric polarization can shift the confined energy levels in the well and thereby control the resonant peak, the current ratio at the peak, and the negative differential resistance at room temperature. If this is right, hybrid junctions and FeRTDs become credible building blocks for low-energy, high-speed, nonvolatile memory and for neuromorphic and multi-state logic that go beyond what simple FTJs offer.","feed_headline":"Ferroelectric switching tunes resonant tunneling at room temperature","feed_subtitle":"A review of hybrid ferroelectric junctions points to a nonvolatile-memory path with huge ON/OFF ratios and negative differential resistance.","key_machinery":"The central object is the FeRTD: a double-barrier quantum-well structure, typically a ferroelectric layer acting as a barrier or as the well, in which electrons tunnel resonantly when their injection energy aligns with a discrete quantized level of the well. Reversing the ferroelectric polarization changes the electrostatic potential profile across the structure, shifting the quantized levels and hence the resonant bias, the peak current, and the onset of negative differential resistance; the same polarization effect, when applied to a hybrid FTJ, acts by modulating the barrier height or width asymmetry at the ferroelectric/functional-layer interface. A key variant is the head-to-head domain wall inside a single ferroelectric film, which creates a V-shaped potential profile that acts as a barrier/well/barrier structure equivalent to a quantum well. These mechanisms convert ferroelectric polarization into a shift of the tunneling resonance, which is what the paper argues constitutes a radically new way to exploit ferroelectric heterostructures.","core_discovery":"The paper claims that ferroelectric polarization can be used to actively control resonant tunneling and the associated negative differential resistance at room temperature, and it assembles the experimental and theoretical evidence for this claim across four hybrid junction families: ferroelectric/dielectric, ferroelectric/multiferroic, ferroelectric/superconducting, and ferroelectric/2D van der Waals junctions, culminating in FeRTDs. The flagship demonstration is the authors' own BTO/SRO/BTO FeRTD, where a metallic SRO quantum well is confined between ferroelectric BaTiO3 barriers; that device shows the hallmarks of resonant tunneling followed by NDR at room temperature, with a large OFF/ON resistance ratio of about 2×$10^{4}$ that is modulated by polarization reversal. The paper further asserts that 2D van der Waals ferroelectrics such as CIPS and α-In2Se3 can deliver giant tunneling electroresistance (up to $10^{7}$–$10^{9}$) and solve scalability and stability problems, and that ferroelectric/dielectric composite barriers can amplify TER through polarization-switchable 2DEGs or through quasi-resonant states in the dielectric. In short, the review seeks to establish the hybrid FTJ and FeRTD as a flexible platform in which ferroelectric polarization does not merely modulate a direct tunnel barrier, but reshapes the entire quantum-transport spectrum.","pith_inferences":["Editorial inference: the paper's own evidence suggests that the ferroelectric mechanism is not the only plausible explanation in several flagship demonstrations; if a non-ferroelectric interpretation such as oxygen-vacancy or redox-mediated switching gains ground, the FeRTD roadmap would shift toward materials and interfaces that suppress ionic motion, not just toward better epitaxy.","Editorial inference: the triangular-well mechanism shown for BTO/STO composite barriers implies that quasi-resonant peaks may appear in many asymmetric composite-barrier FTJs even without a deliberately engineered quantum well, which would mean FeRTD-like behavior is more widespread than the label suggests.","Editorial inference: a direct testable extension of the paper's outlook would be to replace the metallic SRO well with a strongly correlated oxide or a 2D superconducting layer, coupling the resonant-tunneling resonance to a metal–insulator transition or to superconductivity; such a device would either confirm the universality of the polarization-controlled resonance or reveal that the BTO/SRO/BTO"],"forward_implications":["If the central claim is correct, resonant tunneling in FeRTDs can replace direct tunneling in ferroelectric memories, yielding OFF/ON resistance ratios that are orders of magnitude larger than those of conventional FTJs.","Room-temperature ferroelectric control of NDR would enable nonvolatile multi-valued logic and oscillator or amplifier functions in a single device, moving beyond simple two-state memory.","Integration of 2D van der Waals ferroelectrics or HfO2-based ferroelectrics with composite barriers would improve scalability, reduce the read voltage, and address the high resistance-area product that has limited perovskite-based FTJs.","The multiferroic variant, combining ferroelectric control of tunneling with ferromagnetic electrodes, should give four distinct resistance states (two ferroelectric times two magnetic), enabling denser storage and neuromorphic weight encoding.","Superconducting and 2D superconducting hybrid junctions could link nonvolatile ferroelectric switching with superconducting transport, opening a path to electrically controlled superconducting memories."],"supporting_citations":[{"why":"Reports the room-temperature BTO/SRO/BTO FeRTD with resonant tunneling, NDR, and large OFF/ON ratio; the flagship experiment of the perspective.","marker":"[188]"},{"why":"Introduced the ferroelectric/dielectric composite barrier concept where the non-polar layer acts as a polarization switch, predicting giant TER.","marker":"[36]"},{"why":"First-principles prediction of a switchable 2DEG at the BTO/STO interface that bridges tunneling, supporting the 2DEG-based hybrid junction mechanism.","marker":"[46]"},{"why":"Experimental demonstration that ferroelectric polarization of a BTO potential well modulates resonant tunneling and NDR, a direct precursor to FeRTDs.","marker":"[180]"},{"why":"Shows a 2D van der Waals Cr/CIPS/graphene FTJ with giant TER above 10^7 attributed to polarization-induced barrier-height modulation.","marker":"[139]"},{"why":"Provides the cautionary example of MoSi/BFO/YBCO where resistance switching is dominated by electrochemical redox reactions rather than ferroelectricity, defining what FeRTD claims must exclude.","marker":"[33]"},{"why":"Demonstrates quantum oscillations of conductance in LSMO/BTO/LSMO arising from resonant tunneling through a head-to-head domain-wall quantum well.","marker":"[192]"},{"why":"Physical model of a HZO/Ta2O5/HZO quantum well predicting NDR and TER enhanced by resonant tunneling through ferroelectric barriers.","marker":"[184]"}],"fun_headline_variants":["Ferroelectric polarization tunes resonant tunneling at room temperature","Room-temperature resonant tunneling via ferroelectric switching","Hybrid ferroelectric junctions enable room-temperature resonant tunneling","Ferroelectric gates control resonant tunneling at room temperature","Polarization-switchable resonant tunneling at room temperature"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the resistance switching and negative differential resistance reported in the cited hybrid junctions genuinely originate from ferroelectric polarization reversal, not from oxygen-vacancy motion, redox reactions, or other ionic or defect dynamics; the paper itself concedes this is doubtful for the superconducting junctions and for Li-doped ZnO.","fun_headline_variants_meta":{"raw":{"variants":["Ferroelectric polarization tunes resonant tunneling at room temperature","Room-temperature resonant tunneling via ferroelectric switching","Hybrid ferroelectric junctions enable room-temperature resonant tunneling","Ferroelectric gates control resonant tunneling at room temperature","Polarization-switchable resonant tunneling at room temperature"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000851,"raw_usage":{"total_tokens":3781,"prompt_tokens":1107,"completion_tokens":2674,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":723,"completion_tokens_details":{"reasoning_tokens":2597}},"tokens_in":723,"tokens_out":2674,"duration_ms":25110,"temperature":1.0,"reasoning_tokens":2597,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:01:32.678923+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test is to fabricate the same junction stack with the ferroelectric layer replaced by a non-polar insulator of similar thickness and dielectric constant, such as SrTiO3 in place of BaTiO3, and to measure the NDR peak voltage and OFF/ON ratio: if the NDR features persist with comparable magnitude and shift only with bias polarity rather than with poling history, the ferroelectric-switching mechanism is falsified. A complementary observation is to correlate the ON/OFF state and the NDR peak shift with a direct piezoresponse force microscopy hysteresis measurement of the active device area: the absence of a switchable out-of-plane polarization in the device region would falsify the central claim.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports the room-temperature BTO/SRO/BTO FeRTD with resonant tunneling, NDR, and large OFF/ON ratio; the flagship experiment of the perspective."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Experimental demonstration that ferroelectric polarization of a BTO potential well modulates resonant tunneling and NDR, a direct precursor to FeRTDs."},{"cited_title":"Sanchez-Santolino, J","cited_arxiv_id":null,"evidence_quote":"Demonstrates quantum oscillations of conductance in LSMO/BTO/LSMO arising from resonant tunneling through a head-to-head domain-wall quantum well."},{"cited_title":"Chang, Y","cited_arxiv_id":null,"evidence_quote":"Physical model of a HZO/Ta2O5/HZO quantum well predicting NDR and TER enhanced by resonant tunneling through ferroelectric barriers."}],"review_version":1}