{"id":"96684185-e97f-41b9-a926-4927ac8fcc4e","arxiv_id":"2504.20393","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A ZrTe2/FeTe van der Waals heterostructure becomes superconducting below about 10 K and exhibits a 29% superconducting diode effect.","lead":"By stacking thin crystals of ZrTe2, a Dirac semimetal, on the antiferromagnet FeTe, the authors observed superconductivity below about 10 K. The interface showed strong one-way current effects, including a 29% efficient superconducting diode, suggesting a new platform for superconducting electronics.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Superconductivity attribution to the interface is plausible but unproven; hidden confound is that bare FeTe reference samples were not measured, so the claim that ZrTe2 overgrowth is causally required for the 10 K transition is not established by the data shown.","rationale":"I agree with the reader's identification of the interface localization as the weakest assumption. The paper self-identifies this as a hypothesis, states the ZrTe2 Dirac state in the superconducting regime is open, and the SM URL is a placeholder. The key additional load-bearing concern I emphasize is the absence of a bare FeTe control sample in Fig. 2a, since recent work cited as ref [37] claims stoichiometric FeTe is intrinsically superconducting; without such a control, the causal role of ZrTe2 overgrowth cannot be isolated. I also note the dsc=12.9 nm estimate exceeds the ZrTe2 thickness for the 6 UC sample (about 4 nm) while being far below the FeTe thickness, so a homogeneous parallel-field fit does not localize the condensate at the interface; a two-layer or finite-thickness analysis would be needed. The DFT charge transfer calculation is supportive but not decisive, as it addresses doping and magnetic order rather than the superconducting condensate location. Given the paper's own caveats and missing data, CONDITIONAL is appropriate: accept only if the authors provide the bare FeTe control and interface-sensitive evidence (e.g., STM), and the placeholder SM URL is filled with the relevant data.","tokens_in":10822,"tokens_out":1583,"duration_ms":14687,"concrete_test":"Grow an identical 35 UC FeTe film on SrTiO3 in the same MBE chamber, without ZrTe2 overgrowth, and measure R(T), IVCs, Hc2 anisotropy, and BKT exponent on the same Hall bar geometry. If the bare FeTe film also shows Tc ~10 K with 2D signatures and a 29% diode effect, then the claimed emergent interface superconductivity is not established. As a complementary check, perform cross-sectional STM/EDX on the ZrTe2/FeTe interface to directly identify Fe interstitials in the FeTe bulk and their removal at the interface.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is that superconductivity is emergent at the ZrTe2/FeTe interface, with the abstract saying 'two non-superconducting materials' and 2D superconductivity arising 'at the heterointerface.' However, FeTe itself is now reported as intrinsically superconducting when stoichiometric (arXiv:2603.16115, ref [37]), and the authors' own hypothesis is that ZrTe2 overgrowth removes Fe interstitials from the FeTe surface, restoring superconductivity 'only near the interface.' This means the observed Tc ~10 K could be bulk/surface FeTe that has been locally healed, not a new superconducting state induced in ZrTe2. The data shown (Fig. 2a) includes no bare FeTe control sample from the same growth run, so one cannot distinguish 'interface superconductivity' from 'stoichiometric FeTe superconductivity.' Also, the extracted dsc = 12.9 nm is much larger than the ZrTe2 thickness for the 6 UC sample (~4 nm), and much smaller than the 35 UC FeTe thickness (~35 nm), so the 2D/interface assignment with a homogeneous model is not uniquely supported. The conclusion concedes 'the state of ZrTe2 Dirac fermions in the superconducting state remains an open question.' Therefore, the strongest claim about the platform coupling DSM Dirac fermions with superconductivity is weakened if the superconducting condensate is simply in the FeTe layer.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports molecular beam epitaxy growth and characterization of ZrTe2/FeTe heterostructures on SrTiO3, showing a superconducting transition around 10 K in transport measurements. The authors analyze the upper critical field anisotropy, BKT behavior, and non-reciprocal transport (magneto-chiral anisotropy and superconducting diode effect with 29% efficiency). They claim that 2D superconductivity emerges at the interface of two non-superconducting van der Waals materials, ZrTe2 and FeTe, and propose this as a platform for coupling Dirac fermions with superconductivity and magnetism.","tokens_in":11273,"tokens_out":5969,"duration_ms":54973,"significance":"If the central claim were established, the work would provide an epitaxial vdW platform combining a Dirac semimetal, an antiferromagnet, and a ferromagnetic cap, with interesting non-reciprocal transport and a high diode efficiency. The transport data are analyzed with standard tools (BCS Hc2 fits, BKT scaling, second-harmonic measurements), and the structural characterization (RHEED, STEM, EDX, XRD, ARPES) is thorough. However, the manuscript's own text labels the interface location of the superconductivity as a hypothesis, and the absence of a bare FeTe control sample means the central claim is not yet supported beyond reasonable doubt.","major_comments":[{"comment":"The abstract states that '2D superconductivity arises at the heterointerface' and characterizes ZrTe2 and FeTe as 'two non-superconducting van der Waals materials,' but the text later hypothesizes that as-grown FeTe films have Fe interstitials and that superconductivity occurs only near the interface where these are removed during ZrTe2 overgrowth (text after Fig. 2(a): 'we hypothesize...'). The manuscript further cites ref [37] showing that defect-free FeTe is intrinsically superconducting. No bare FeTe control sample is shown, so the data cannot distinguish interface superconductivity from superconductivity in a FeTe layer whose defects are healed by overgrowth. This is load-bearing for the paper's central novelty claim; please revise the abstract and conclusion to match the level of support, or provide a control experiment.","section":"Abstract and text after Fig. 2(a)"},{"comment":"The extracted superconducting thickness d_sc = 12.9 nm is larger than the ZrTe2 thickness of the 6 UC sample (~4 nm) and smaller than the 35 UC FeTe layer (~35 nm). The homogeneous 2D model used in Eqs. (1)-(2) assumes a single superconducting layer; the fitted d_sc therefore does not uniquely place the condensate at the ZrTe2/FeTe interface. Please discuss how the model applies to a bilayer and whether d_sc can be interpreted only as an effective parameter.","section":"Upper critical field analysis, Eqs. (1)-(2), Fig. 3(a)"},{"comment":"The BKT analysis (Halperin-Nelson fit, power-law IVC with alpha=3 at 9.5 K) supports two-dimensional superconductivity, but two-dimensional behavior would also result from a thin superconducting FeTe layer at the interface or in the FeTe film. The BKT evidence therefore does not resolve which layer hosts the condensate. This is a logical gap in the argument for interface superconductivity.","section":"BKT analysis, Fig. 3(b)-(d)"},{"comment":"The manuscript interprets the magneto-chiral anisotropy as arising from the Dirac semimetal ZrTe2 and uses this to argue that helical spin-momentum locking is not essential. However, if the superconductivity resides in FeTe (as the proposed healing mechanism suggests), the non-reciprocal transport may originate at the FeTe interface rather than from the ZrTe2 Dirac bands. The conclusion that the DSM band structure influences the MChA requires either direct evidence from ZrTe2 or a more careful disentangling of the two layers.","section":"Non-reciprocal transport, Fig. 4(c)-(d)"}],"minor_comments":[{"comment":"The reported critical temperature is inconsistent: the abstract states Tc ~ 10 K, the BKT analysis gives T_BKT = 9.6 K, and the conclusion states 'a critical temperature of 12 K.' Please unify the definition and values.","section":"Abstract, BKT section, Conclusion"},{"comment":"The supplemental material reference [42] is a placeholder ('INSERT_URL_HERE'); please update before publication.","section":"References"},{"comment":"The resistance is normalized by the 300 K value, but the text defines Tc relative to the 30 K value; please clarify the normalization convention.","section":"Fig. 2(a)"},{"comment":"The phrase 'N´ eel temperature' contains a stray accent; please check for similar encoding issues throughout.","section":"Text near Fig. 2(a)"},{"comment":"The description of the second harmonic measurement would benefit from a precise statement of the sign convention and the relative orientation of current and magnetic field, as the expression for R_2ω depends on the geometry.","section":"Eq. (3)"},{"comment":"The term 'superconducting length' might be better rendered as 'effective superconducting thickness' to avoid confusion with a coherence length.","section":"Text after Eq. (2)"}],"recommendation":"major_revision","confidential_remarks":"The paper relies on the overlapping-author preprint [37] (arXiv:2603.16115) for the claim that stoichiometric FeTe is intrinsically superconducting. Since this is not yet peer-reviewed and the authors appear to share authorship with the present manuscript, the editors may wish to ensure that the paper does not overstate the external validation of this premise. The key technical concern, however, is the absence of a bare FeTe control sample, which is the basis of the main major comment."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nYou should know about this paper if you follow FeTe-based hybrid superconductivity. It is a clean MBE study of ZrTe2/FeTe heterostructures that shows a 2D superconducting transition around 10 K, BKT behavior, magnetochiral anisotropy, and a 29% diode efficiency with a CrTe2 cap. The materials work is solid: RHEED, STEM, EDX, XRD, and ARPES showing Dirac bands in ZrTe2. The transport analysis is standard and internally consistent: Hc2 fits, Halperin-Nelson BKT fits, IV exponent alpha = 3 at 9.5 K, and a gamma divergence. The data support that the superconductivity is 2D in nature. That is new for this material combination. Prior work used topological insulators or CrTe2 directly interfaced with FeTe; this uses the Dirac semimetal ZrTe2 and shows comparable non-reciprocal response despite spin-degenerate bulk Dirac bands, with a threefold enhancement from a ferromagnetic cap a few nm away.\n\nWhere it is softer: the abstract claims \"two non-superconducting van der Waals materials\" and \"2D superconductivity arises at the heterointerface.\" The text later says the interface location is a hypothesis: they suppose as-grown FeTe has Fe interstitials that suppress superconductivity, and ZrTe2 overgrowth removes them near the interface. The obvious control—a bare FeTe film from the same growth run—is not shown. Given recent work suggesting stoichiometric FeTe is intrinsically superconducting, and that reference [37] is an overlapping-author preprint, the causal role of ZrTe2 is not established by the data. The extracted d_sc = 12.9 nm also sits awkwardly: larger than the ZrTe2 thickness (~4 nm), smaller than the FeTe thickness (~35 nm), so the homogeneous 2D-layer fit does not uniquely locate the condensate. The paper itself concedes the state of the ZrTe2 Dirac fermions in the superconducting state is open. Minor items: the supplemental placeholder URL, and error bars largely absent from the Hc2 and diode numbers.\n\nI don't read this as fatal. The platform is new, the transport measurements are direct, and the authors are honest about the hypothesis. But the headline should be softened: emergent superconductivity likely resides in FeTe locally healed at the interface, with proximity coupling to ZrTe2 still unknown. That is still worth publishing after revision, with a bare FeTe control and a more careful abstract. Who gets value: experimentalists working on vdW superconducting heterostructures, superconducting diodes, and FeTe interface superconductivity. I would send it to a serious referee; it deserves referee time. Whether I would cite it in my own work depends on whether the control appears.","headline":"Clean MBE study of a new ZrTe2/FeTe platform with genuine 2D superconductivity and strong diode response, but the interface attribution is a hypothesis that needs a bare-FeTe control before the abstract's claims are warranted.","tokens_in":11695,"tokens_out":2050,"would_cite":false,"duration_ms":21060,"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":"Interface of two nonsuperconductors superconducts at 10 K.","keywords":["van der Waals heterostructures","Dirac semimetal","interface superconductivity","FeTe","ZrTe2","magneto-chiral anisotropy","superconducting diode effect","Berezinskii-Kosterlitz-Thouless transition"],"falsifier":"A cross-sectional scanning tunneling microscopy study across the ZrTe2/FeTe interface would settle the mechanism: if the superconducting gap appears in FeTe regions far from the interface, or if the interface region shows no local removal of iron interstitials, then the proposed interface-induced superconductivity is wrong. Measuring the upper critical field of bare FeTe films grown under the same conditions would also test whether the superconductivity is intrinsic to FeTe.","tokens_in":10611,"feed_emoji":"⚡","tokens_out":14964,"duration_ms":133951,"temperature":0.7,"pith_summary":"This paper reports that growing the Dirac semimetal ZrTe2 directly on the antiferromagnet FeTe produces a superconducting state at their shared interface, even though neither material superconducts on its own in this form. Electrical transport shows two-dimensional superconductivity below $T_c\\sim10$ K, with a Berezinskii-Kosterlitz-Thouless transition at 9.6 K and a coherence length near 2 nm. In the transition region the device conducts differently for opposite current directions, giving a magneto-chiral anisotropy comparable to topological insulator hybrids, and a one-monolayer ferromagnetic CrTe2 cap triples that effect. Below $T_c$, the critical current depends on current direction strongly enough that the structure acts as a superconducting diode with 29% efficiency. The authors propose this as an epitaxial van der Waals platform for coupling Dirac fermion topology with superconductivity and magnetism.","feed_headline":"Stacking ZrTe2 and FeTe makes a 10 K interface superconductor","feed_subtitle":"It also acts as a 29% efficient superconducting diode, pointing toward non-reciprocal devices for superconducting electronics.","key_machinery":"The central object is the epitaxial ZrTe2/FeTe van der Waals interface, where the superconducting condensate is claimed to reside. The argument is carried by three quantitative tools: the Halperin-Nelson and Berezinskii-Kosterlitz-Thouless analysis of resistance and current-voltage characteristics, which fixes the two-dimensional nature and transition temperature; the upper critical field fits of Eqs. (1) and (2), which yield $\\xi_0=1.9$ nm and $d_{sc}=12.9$ nm; and second-harmonic transport, which extracts the magneto-chiral coefficient $\\gamma$ from $R_{2\\omega}/R_{\\omega}$ versus magnetic field. Mechanistically, the paper proposes that ZrTe2 overgrowth removes Fe interstitials from the FeTe surface and transfers holes into FeTe, restoring superconductivity at the interface while leaving the bicollinear antiferromagnetic order intact.","core_discovery":"The central claim is that emergent two-dimensional superconductivity appears at the ZrTe2/FeTe interface below $T_c\\sim10$ K and coexists with the bicollinear antiferromagnetic order of FeTe. The authors infer the two-dimensional character from a Berezinskii-Kosterlitz-Thouless fit to the resistance and from the current-voltage exponent reaching $\\alpha=3$ at $T_{BKT}=9.6$ K, with upper critical field fits giving $\\xi_0=1.9$ nm and a superconducting thickness $d_{sc}=12.9$ nm. They further report that the superconducting state supports non-reciprocal transport: a magneto-chiral anisotropy comparable to that seen in topological insulator/FeTe hybrids, enhanced threefold by a CrTe2 ferromagnetic cap, and a superconducting diode efficiency of 29%. The proposed mechanism is that ZrTe2 overgrowth removes interstitial Fe from FeTe near the interface, restoring superconductivity locally while first-principles calculations show hole doping that stabilizes the bicollinear antiferromagnetic phase.","pith_inferences":["Editorial inference: if interfacial charge transfer is the control knob, tuning the thickness or work function of the telluride cap should systematically shift $T_c$ and the diode efficiency; a scan across several cap materials would test this without changing the FeTe itself.","Editorial inference: the threefold enhancement from a ferromagnet placed about 4 nm away suggests the non-reciprocal response is not a simple exchange proximity effect; interposing nonmagnetic spacers of increasing thickness would locate the coupling range.","Editorial inference: the authors leave open whether ZrTe2 Dirac fermions remain gapless in the superconducting state; if they do, tunneling spectroscopy into the ZrTe2 surface could search for the bulk point nodes and surface Majorana modes predicted for superconducting Dirac semimetals."],"forward_implications":["Below $T_c\\sim10$ K the heterostructure carries supercurrent in a two-dimensional sheet with a BKT transition at 9.6 K, so the superconducting state is confined near the interface.","The upper critical field exceeds 14 T in both orientations and extrapolates to roughly twice the Pauli limit, a result the paper links to strong spin-orbit interaction, spin-triplet pairing, or a Fulde-Ferrell-Larkin-Ovchinnikov state.","Magneto-chiral anisotropy appears in a Dirac semimetal with spin-degenerate bulk Dirac bands, showing that helical surface states are not required for non-reciprocal transport.","A monolayer ferromagnetic cap triples the magneto-chiral anisotropy even though it sits about 4 nm from the superconducting interface, a result the paper attributes tentatively to broken time-reversal symmetry.","The 29% diode efficiency places this van der Waals stack among the most efficient superconducting diodes reported."],"supporting_citations":[{"why":"Provides the FeTe-based interface superconductivity precedent and the 2D upper critical field analysis used to extract $\\xi_0$ and $d_{sc}$.","marker":"[26]"},{"why":"Reports that stoichiometric, defect-free FeTe is intrinsically superconducting, the basis for the interstitial-removal mechanism proposed here.","marker":"[37]"},{"why":"Establishes the molecular beam epitaxy growth and Dirac semimetal band structure of ZrTe2 used for the DSM layer.","marker":"[38]"},{"why":"Provides the ZrTe2/CrTe2 hybrid platform and band structure reference for the ferromagnet-capped stack.","marker":"[39]"},{"why":"Reports magneto-chiral anisotropy in a topological insulator/FeTe hybrid and supplies the comparison value for $\\gamma$.","marker":"[27]"},{"why":"Reports non-reciprocal transport enhancement when CrTe2 is directly interfaced with FeTe, the comparison for the threefold enhancement seen here.","marker":"[36]"},{"why":"Supplies the Halperin-Nelson resistance form used to extract the Berezinskii-Kosterlitz-Thouless transition temperature.","marker":"[62]"},{"why":"Reports the superconducting diode effect and provides the efficiency measure $\\eta$ used in the paper.","marker":"[69]"}],"fun_headline_variants":["Two non-superconductors stack into 10 K interface superconductor","ZrTe2/FeTe interface superconducts at 10 K, diode at 29%","Emergent 2D superconductor from Dirac semimetal/antiferromagnet stack","Interface of ZrTe2 and FeTe: 10 K superconductor with 29% diode effect"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the superconducting state is created at the ZrTe2/FeTe interface when the ZrTe2 layer strips away excess iron atoms from the FeTe surface; the paper's own conclusion notes that whether ZrTe2's Dirac electrons actually enter the superconducting state is still an open question, so if the superconductivity turns out to be a bulk FeTe property the central platform claim is weakened.","fun_headline_variants_meta":{"raw":{"variants":["Two non-superconductors stack into 10 K interface superconductor","ZrTe2/FeTe interface superconducts at 10 K, diode at 29%","Emergent 2D superconductor from Dirac semimetal/antiferromagnet stack","Interface of ZrTe2 and FeTe: 10 K superconductor with 29% diode effect"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000916,"raw_usage":{"total_tokens":3958,"prompt_tokens":995,"completion_tokens":2963,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":611,"completion_tokens_details":{"reasoning_tokens":2866}},"tokens_in":611,"tokens_out":2963,"duration_ms":21089,"temperature":1.0,"reasoning_tokens":2866,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T05:30:35.166659+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A cross-sectional scanning tunneling microscopy study across the ZrTe2/FeTe interface would settle the mechanism: if the superconducting gap appears in FeTe regions far from the interface, or if the interface region shows no local removal of iron interstitials, then the proposed interface-induced superconductivity is wrong. Measuring the upper critical field of bare FeTe films grown under the same conditions would also test whether the superconductivity is intrinsic to FeTe.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the FeTe-based interface superconductivity precedent and the 2D upper critical field analysis used to extract $\\xi_0$ and $d_{sc}$."},{"cited_title":"Stoichiometric FeTe is a Superconductor","cited_arxiv_id":"2603.16115","evidence_quote":"Reports that stoichiometric, defect-free FeTe is intrinsically superconducting, the basis for the interstitial-removal mechanism proposed here."},{"cited_title":"Tsipas, D","cited_arxiv_id":null,"evidence_quote":"Establishes the molecular beam epitaxy growth and Dirac semimetal band structure of ZrTe2 used for the DSM layer."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the ZrTe2/CrTe2 hybrid platform and band structure reference for the ferromagnet-capped stack."},{"cited_title":"Meissner Effect and Nonreciprocal Charge Transport in Non-Topological 1T-CrTe2/FeTe Heterostructures","cited_arxiv_id":"2412.09354","evidence_quote":"Reports non-reciprocal transport enhancement when CrTe2 is directly interfaced with FeTe, the comparison for the threefold enhancement seen here."},{"cited_title":"Halperin and D","cited_arxiv_id":null,"evidence_quote":"Supplies the Halperin-Nelson resistance form used to extract the Berezinskii-Kosterlitz-Thouless transition temperature."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports the superconducting diode effect and provides the efficiency measure $\\eta$ used in the paper."}],"review_version":1}