{"id":"0121af93-4280-4be5-9ffd-b41681cd5d65","arxiv_id":"2508.07351","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Monolayer Si2Te2 on HfTe2 shows a ~300 meV gap and step-edge states, reported as the first experimental realization of its predicted quantum spin Hall phase.","lead":"This paper grows a single layer of Si2Te2 on a HfTe2 crystal and uses scanning tunneling microscopy to find a large electronic gap and edge-localized states that match predictions for a quantum spin Hall insulator. If the interpretation holds, it offers a practical platform for room-temperature topological electronics.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central QSH claim hinges on identifying the zero-bias step-edge dI/dV peak as a topological edge state; the paper does not exclude trivial edge states, and the supporting DFT uses a monolayer HfTe2 substrate while the experiment is on bulk HfTe2.","rationale":"The reader's weakest-assumption analysis correctly identifies the step-edge dI/dV peak as the load-bearing piece of evidence. The paper's own reasoning relies on two proxies—spatial extent and robustness to geometry—that are not unique to topological edge states. This is the single most important vulnerability: if that peak is trivial, the central claim of experimental QSH realization collapses, no matter how good the DFT calculations are. The DFT results (Z2 invariant, edge LDOS, phonon stability) provide useful supporting evidence, but they are predictions, not experimental proof. I found no internal logical contradiction in the main derivation, but the bulk-vs-monolayer substrate mismatch and the lack of a trivial control are concrete correctness risks. The proposed test—checking whether the same edge peak appears in the previously reported trivial ML-Si2Te2/Sb2Te3 system—is a direct and feasible discriminator. Since the reader's verdict is already CONDITIONAL and this concern supports those conditions without moving the verdict, the appropriate label is UNCHANGED.","tokens_in":9914,"tokens_out":4950,"duration_ms":57430,"concrete_test":"Perform the same STS step-edge measurement on ML-Si2Te2 grown on Sb2Te3, which the authors previously identified as strained and topologically trivial [29], using identical tunneling parameters, tip conditions, and analysis. If a zero-bias edge peak with ~2 nm spatial extent and geometry-independent intensity also appears in that known trivial system, the current edge-state evidence cannot be considered topological. If the peak is absent there, the comparison would strengthen—but not by itself prove—the topological interpretation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central assertion is that ML-Si2Te2 on HfTe2 realizes a QSH phase, and the decisive experimental evidence is the zero-bias dI/dV peak observed at step edges in Fig. 5. The authors label this peak topological based on its ~2-3 nm spatial extent and robustness to step geometry. That logic is not sufficient: trivial edge states arising from step reconstruction, dangling bonds, or substrate-induced band bending can also be spatially extended and insensitive to step geometry. The paper does not provide spin-resolved STS, edge conductance measurements, or a direct comparison with a known trivial phase of the same material. In fact, the same authors previously found ML-Si2Te2 on Sb2Te3 to be strained and trivial [29]; no control experiment on that system is reported here. Additionally, the DFT heterostructure used for band topology and edge-state calculations is ML-Si2Te2/ML-HfTe2, whereas the experiment uses a bulk HfTe2 substrate. The calculated 319 meV gap and the position of the topological edge states could shift in the bulk-substrate geometry, potentially moving the edge states away from EF. Therefore, if the observed zero-bias peak is a trivial edge state, the headline claim of topological phase is unsupported. The paper also cites supplemental material via a placeholder URL with unrelated text, making the supporting data hard to verify.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports the epitaxial growth of monolayer Si2Te2 on HfTe2 and interprets the system as a quantum spin Hall (QSH) insulator. DFT is used to screen substrates, predict lattice matching, dynamical stability (phonons, AIMD), band topology (HSE06 gap of 319 meV, Z2=1), and edge-state LDOS. STM/STS measurements show a strain-free (1x1) Si2Te2 lattice with a ~300 meV gap and a zero-bias dI/dV peak localized within ~2-3 nm of island step edges, which the authors attribute to topological edge states. The central claim is that this constitutes the first experimental realization of the QSH phase in monolayer Si2Te2.","tokens_in":10302,"tokens_out":6698,"duration_ms":70782,"significance":"If the interpretation is correct, this is a significant milestone: it would demonstrate a predicted room-temperature QSH insulator in an artificial 2D material and validate a material-by-design approach that combines high-throughput substrate screening with MBE growth. The work has clear strengths: the DFT predictions are parameter-free and well matched to experiment (lattice constant 390 pm vs 389 pm; gap ~300 meV vs 319 meV); phonon and AIMD stability calculations support the synthesis; and the edge-state signal is reproducible and spatially extended. The main weakness is that the decisive experimental evidence does not uniquely establish the topological nature of the edge state, and key supporting data are relegated to an inaccessible placeholder reference.","major_comments":[{"comment":"The central conclusion rests on assigning the zero-bias dI/dV peak at step edges to topological edge states. The evidence offered is the ~2-3 nm spatial extent and the robustness to step geometry. These criteria do not exclude trivial mechanisms such as step-related dangling bonds, reconstruction, or substrate-induced band bending, which can also produce broad and robust edge-localized signals. The manuscript does not provide spin-resolved STS, quantized edge conductance, or a direct comparison with a known trivial phase (e.g., the strained Si2Te2/Sb2Te3 system from Ref. [29]). As written, the abstract's 'first experimental realization of the QSH phase' overstates what the data show. I recommend either adding a discriminating measurement or qualifying the claim to 'evidence for topological edge states' / 'candidate QSH phase.'","section":"Fig. 5 / 'Identification of the edge states'"},{"comment":"Reference [35] is cited more than a dozen times for critical data: the Z2 Wannier-center calculation, edge-state LDOS, exfoliation energies, AIMD details, dI/dV maps, and the statistics over 20 islands. However, the reference as printed is a placeholder URL (http://link.aps.org/supplemental/XXX) followed by text about 'coexistence of three stripe orientations,' which is unrelated to this manuscript. This makes the supporting data unverifiable and is a blocking issue for review. The supplemental material must be properly provided before the manuscript can be considered.","section":"Reference [35] / Supplemental Material"},{"comment":"The band-structure calculations, including the 319 meV gap, the Fermi-level position, and the edge-state dispersion, are performed for ML-Si2Te2 on a monolayer HfTe2 substrate, while the experiments use bulk HfTe2. The manuscript does not justify that the ML-HfTe2 model faithfully represents the bulk substrate. The measured gap agreement could be coincidental if the bulk substrate changes the band alignment or doping. Please provide a bulk-slab calculation or an explicit argument for why the monolayer substrate model is equivalent, especially since the paper already invokes a 0.25 eV post-hoc energy shift for substrate doping effects.","section":"Section 2 / Fig. 2"}],"minor_comments":[{"comment":"Typos and grammatical issues: 'V ASP' / 'Viennaab initio' spacing, 'valance' should be 'valence', 'remains challenge' should be 'remains a challenge', 'the exact energy positions ... appear smeared' is awkward. Please copyedit.","section":"Throughout"},{"comment":"In Fig. 4(b), the DFT projected DOS is compared to the total DOS of bulk HfTe2, while the band-structure calculations use ML-HfTe2. Clarify which HfTe2 DOS is used and whether bulk vs monolayer affects the comparison.","section":"Fig. 4"},{"comment":"The center-of-island spectrum shows a V-shaped dip with a small in-gap peak attributed to substrate-induced states. This complicates the extraction of the bulk gap; please quantify how the VBM and CBM were determined from the high-resolution spectra and how the substrate states are excluded.","section":"Fig. 5"},{"comment":"The Z2 invariant calculation via Wannier charge centers is cited to Giustino and Pasquarello (2006), which is primarily a phonon method. Please cite the standard references for the Z2 Wannier-center approach (e.g., Soluyanov and Vanderbilt, or Fu and Kane).","section":"Reference [39]"}],"recommendation":"major_revision","confidential_remarks":"The paper is plausible and the DFT/experimental agreement on the lattice constant and gap is compelling. However, the headline claim of QSH realization is not settled by the present edge-state evidence, and the placeholder supplemental reference is a serious issue that must be checked. If the authors can add a control experiment or spin-resolved data, the claim would be much stronger; otherwise, they should reframe the conclusion as evidence for a topological candidate. I would not reject outright because the core DFT predictions and the edge-state observation are valuable, but the manuscript needs substantial revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things up front. The growth is real: the authors get a strain-free 1x1 ML-Si2Te2 on HfTe2, and the measured ~300 meV gap sits close to the HSE06 value of 319 meV. That is solid experimental work and matches the paper's earlier strain screening logic. The second thing is that the topological edge-state claim is not nailed. The stress-test note is right: a zero-bias dI/dV peak at a step edge, with ~2 nm extent and robustness to geometry, does not by itself rule out trivial edge states from reconstruction or substrate coupling. No spin-resolved STS, no edge transport, no comparison with a trivial reference. So \"first experimental realization\" overstates what is shown; \"strong evidence for\" would be more accurate.\n\nWhat is genuinely new: previous work from the same group on Sb2Te3 gave a strained, trivial phase. Here the substrate choice is much better: HfTe2 has a lattice match, vdW interaction, and the DFT predicts the topological phase survives. The calculations are careful—phonons, AIMD, HSE06 gap, Z2 via Wannier charge centers. The experimental lattice constant 390 pm vs 389 pm DFT is exactly what you want. The gap measurement on 20 islands is a nice touch. So the paper is a credible step forward in the materials-by-design program.\n\nWhere it gets soft: (1) The edge-state identification is the load-bearing piece. The argument that spatial extent \"rules out trivial mechanisms\" is too strong—a step reconstruction can be extended and robust. They need either spin-polarized STS, a check against a trivial phase (the Sb2Te3-grown film would be a natural control), or at least an explicit acknowledgment that this is indirect. (2) The DFT uses ML-HfTe2 as the substrate while the experiment is on bulk HfTe2. That could shift the edge-state position relative to EF, so the comparison is not airtight. (3) The 0.25 eV rigid shift in Fig. 4 is a small free parameter; fine if stated, but it weakens the claim of \"well captured\" for the whole spectrum. (4) The supplementary URL is a placeholder with unrelated text—unacceptable in a submission and should be fixed.\n\nNone of these kill the paper; they are the difference between \"evidence for\" and \"realization of.\" The system is new, and the DFT support is strong enough that a serious referee should see it. I'd send it to peer review with the request that the authors either add the missing control/measurement or soften the language.\n\nFor whom: experimentalists working on 2D TIs and anyone interested in substrate screening for topological phases. Worth a reading group if you want to debate evidence standards in STM/STS.","headline":"The growth and bulk gap are solid and match DFT, but the topological edge-state claim is supported by indirect STM evidence only; still worth a serious referee.","tokens_in":10780,"tokens_out":3142,"would_cite":true,"duration_ms":31020,"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":"Monolayer Si2Te2 grown on HfTe2 realizes the quantum spin Hall phase, with a ~300 meV gap and topological edge states.","keywords":["monolayer Si2Te2","quantum spin Hall insulator","topological edge states","HfTe2 substrate","molecular beam epitaxy","scanning tunneling microscopy","band gap","Z2 invariant"],"falsifier":"Measure the two-terminal conductance of a lithographically patterned ML-Si2Te2 channel with the Fermi level inside the ~300 meV gap: a quantum spin Hall insulator must show a quantized $2e^2/h$ edge-conductance plateau that survives local non-magnetic disorder, whereas a trivial edge state would not quantize.","tokens_in":9855,"feed_emoji":"⚛️","tokens_out":10867,"duration_ms":100562,"temperature":0.7,"pith_summary":"This paper reports the first experimental realization of the predicted quantum spin Hall phase in a single layer of Si2Te2, a material with no three-dimensional parent. The quantum spin Hall phase is an insulating state whose edges carry conducting channels protected by time-reversal symmetry. By growing the monolayer on HfTe2, the authors obtain a strain-free (1x1) lattice held by van der Waals forces, preserving the spin-orbit band inversion that makes the free-standing layer topologically nontrivial. Scanning tunneling spectroscopy measures a bulk gap of about 300 meV, large enough to matter for room-temperature operation, and finds a sharp, spatially extended edge-state signal at step edges, the expected signature of helical topological channels. If the interpretation holds, the result would make ML-Si2Te2 a practical platform for studying dissipationless edge transport at high temperature.","feed_headline":"First quantum spin Hall phase seen in monolayer Si2Te2","feed_subtitle":"Strain-free growth on HfTe2 opens a ~300 meV gap with clear edge states at step edges.","key_machinery":"The carrying object is the ML-Si2Te2/HfTe2 heterostructure: a hexagonal Te-Si-Si-Te monolayer matched to HfTe2 at its natural lattice constant, so the SOC-induced band inversion at $\\Gamma$ survives. The argument combines three tools: (i) a substrate-screening criterion based on lattice mismatch and exfoliation energy; (ii) hybrid-functional DFT plus Wannier charge centers to establish a $Z_2=1$ topology; and (iii) STS measurements that locate an in-gap step-edge peak and map its ~2 nm spatial extent.","core_discovery":"The central claim is that monolayer Si2Te2 on HfTe2 is a quantum spin Hall insulator. The paper demonstrates this by first screening 4056 two-dimensional materials to select HfTe2 as a substrate whose in-plane lattice matches the free-standing monolayer, then showing with hybrid-functional DFT that the heterostructure retains the band inversion at $\\Gamma$ and a $Z_2=1$ invariant. Experimentally, MBE growth produces strain-free (1x1) islands with an in-plane constant of 390 pm; STS resolves a ~300 meV gap consistent with the calculated 319 meV; and dI/dV maps reveal a ~2 nm-wide peak at the Fermi level running continuously along island edges, independent of step geometry. The paper interpret","pith_inferences":["A direct test the authors leave implicit: measuring the edge conductance of a patterned ML-Si2Te2 channel; a quantized $2e^2/h$ plateau would confirm the helical nature that STS alone cannot show.","One might expect the topological edge state to be spin-momentum locked; spin-polarized STM or nonlocal transport measurements are natural next experiments to map the helical texture.","The same substrate-screening logic could be applied to other predicted large-gap 2D TIs, but interface doping or hybridization, seen here as a ~0.25 eV spectral shift, may alter the phase, so screening alone may not guarantee topology."],"forward_implications":["With a ~300 meV gap, the QSH phase in ML-Si2Te2 should persist at room temperature, in contrast to the meV-scale gaps of HgTe/CdTe and InAs/GaSb quantum wells.","The step-edge states, if helical, would provide dissipationless 1D conduction channels protected against backscattering by time-reversal symmetry.","Growth on HfTe2 makes an exfoliation-inaccessible artificial 2D topological insulator experimentally accessible, and the screening workflow can be reused for other predicted 2D TIs without bulk parents.","Because the phase is strain-sensitive, keeping the monolayer at its free-standing lattice constant is the enabling condition; the observed 390 pm lattice proves this condition is met.","The ~2 nm edge-state width suggests the topological channels can be patterned at the nanometer scale for spintronic or low-power interconnects."],"supporting_citations":[{"why":"Predicted the room-temperature quantum spin Hall phase in free-standing ML-Si2Te2, the target state this work realizes.","marker":"[28]"},{"why":"Earlier growth on Sb2Te3 showed that lattice strain drives ML-Si2Te2 into a trivial semiconducting phase, motivating the strain-free HfTe2 choice.","marker":"[29]"},{"why":"Supplies the 2D-materials database used to screen candidate substrates for lattice matching to ML-Si2Te2.","marker":"[31]"},{"why":"Supplies the hybrid-functional DFT method used to compute the band structure and the ~319 meV gap.","marker":"[38]"},{"why":"Provides the Wannier charge-center method used to determine the $Z_2=1$ invariant.","marker":"[39]"},{"why":"Benchmark STS study showing topological edge states appear as spatially extended dI/dV peaks, used to interpret the observed step-edge signal.","marker":"[22]"},{"why":"States that a bulk band gap is essential for the QSH phase, motivating the gap measurement.","marker":"[13]"}],"fun_headline_variants":["Quantum spin Hall phase realized in Si2Te2 on HfTe2","Edge states confirm topological Si2Te2 grown on HfTe2","Monolayer Si2Te2 on HfTe2 shows topological gap and edge states","First experimental topological phase in monolayer Si2Te2","Si2Te2 on HfTe2: a topological insulator with edge states"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The interpretation stands on the assumption that the sharp dI/dV peak at island step edges is a topological edge state rather than a trivial edge state from step reconstruction, substrate coupling, or tip effects; the paper does not directly measure spin helicity or quantized edge conductance.","fun_headline_variants_meta":{"raw":{"variants":["Quantum spin Hall phase realized in Si2Te2 on HfTe2","Edge states confirm topological Si2Te2 grown on HfTe2","Monolayer Si2Te2 on HfTe2 shows topological gap and edge states","First experimental topological phase in monolayer Si2Te2","Si2Te2 on HfTe2: a topological insulator with edge states"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000389,"raw_usage":{"total_tokens":1863,"prompt_tokens":700,"completion_tokens":1163,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":444,"completion_tokens_details":{"reasoning_tokens":1064}},"tokens_in":444,"tokens_out":1163,"duration_ms":10578,"temperature":1.0,"reasoning_tokens":1064,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T22:10:54.876168+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the two-terminal conductance of a lithographically patterned ML-Si2Te2 channel with the Fermi level inside the ~300 meV gap: a quantum spin Hall insulator must show a quantized $2e^2/h$ edge-conductance plateau that survives local non-magnetic disorder, whereas a trivial edge state would not quantize.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Predicted the room-temperature quantum spin Hall phase in free-standing ML-Si2Te2, the target state this work realizes."},{"cited_title":"Huang, R","cited_arxiv_id":null,"evidence_quote":"Earlier growth on Sb2Te3 showed that lattice strain drives ML-Si2Te2 into a trivial semiconducting phase, motivating the strain-free HfTe2 choice."},{"cited_title":"Haastrup, M","cited_arxiv_id":null,"evidence_quote":"Supplies the 2D-materials database used to screen candidate substrates for lattice matching to ML-Si2Te2."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the hybrid-functional DFT method used to compute the band structure and the ~319 meV gap."},{"cited_title":"Giustino and A","cited_arxiv_id":null,"evidence_quote":"Provides the Wannier charge-center method used to determine the $Z_2=1$ invariant."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Benchmark STS study showing topological edge states appear as spatially extended dI/dV peaks, used to interpret the observed step-edge signal."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"States that a bulk band gap is essential for the QSH phase, motivating the gap measurement."}],"review_version":1}