{"id":"114f13f4-a8b9-4655-a7d6-9c51565c96a6","arxiv_id":"1909.02154","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"In ZrSiTe, topological drumhead surface states appear only in the region between the projections of two nodal lines where the Z2 Berry phase is pi, as confirmed by ARPES.","lead":"This paper shows where topological drumhead surface states appear on the (001) surface of the nodal line semimetal ZrSiTe, and confirms the prediction with photoemission experiments. The result gives a simple Z2 rule for when surface states survive when two nodal lines project onto the same region.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Z2 drumhead classification is only established without SOC; Section III.C concedes SOC gaps the nodal lines, and no SOC Wilson-loop check is presented.","rationale":"The paper is a coherent, well-supported study: the Wilson-loop machinery is standard, the WCC mapping in Appendix A is explicit, the slab DFT reproduces the drumhead and floating-band features, and the ARPES dispersion cuts match the no-SOC prediction in the region between the nodal-line projections. The reader identified SOC as the weakest assumption, and I agree. The point is load-bearing because the headline claim is specifically that the drumhead states are topologically required by a Z2 Berry phase; that invariant is only computed without SOC, while the real material has SOC and the nodal lines are conceded to gap. The provided SOC spectral function is reassuring but does not establish that the bulk WCC parity survives, since trivial surface states can appear in the same region as demonstrated by the floating band. Other possible concerns—quantitative ARPES calibration, lack of shared code, novelty overstatement—are less damaging because the qualitative experimental-theoretical agreement is strong and the core machinery is standard. A single additional calculation, the SOC Wilson-loop spectrum, would settle the issue; hence the verdict should be CONDITIONAL rather than outright ACCEPT, although there is no reason to suspect the claim is wrong.","tokens_in":12634,"tokens_out":15172,"duration_ms":157817,"concrete_test":"Recompute the kz Wilson loop and WCC spectrum for the 28 occupied spinor bands of ZrSiTe with SOC, along the same (kx, ky = 0.04 Å^-1) path used in Table I and Fig.2(d), using the symmetry-eigenvalue mapping in Appendix A generalized to the spinful Mz eigenvalues. Check whether exactly one WCC remains at the π-invariant position between the projected gapped NL1 and NL2, and zero WCCs at π in the overlap region. Any change in the parity of π WCCs relative to the no-SOC count would invalidate the claim that SOC is a perturbative correction to the Z2 invariant.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim that the observed ribbon states are topologically required rests on the Z2 Berry-phase analysis of Section III.B, which is performed for the spinless 14-band problem. Section III.C explicitly states that the nodal lines in ZrSiTe are not stable with respect to SOC: they gap once SOC is included. The only SOC evidence offered is the spinful slab spectral function in Fig.4, which shows that drumhead-like surface features persist and split by a few meV. But a surface spectral function cannot certify that the bulk WCC structure survives SOC: trivial surface states, including the floating band discussed in Section III.B, appear even where the Berry phase is 0. With SOC the occupied space doubles to 28 spinor bands, the Mz eigenvalue branches in Appendix A change, and the exact WCC mapping used to prove one WCC at Pos.1b is no longer directly applicable. If SOC moves WCCs away from the π-invariant position or changes their parity, the Z2 modular-arithmetic explanation—and the claim that the drumhead ribbon is topologically protected in the real material—fails, even though ARPES would still show surface resonances in the same momentum region.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents a combined DFT, Wilson-loop, and ARPES study of the nodal-line semimetal ZrSiTe. It identifies two mirror-protected nodal lines, NL1 in the kz=0 plane and NL2 in the kz=pi plane, and computes the abelian Berry phase along kz as a function of (kx,ky), obtaining a region of pi Berry phase between the surface projections of the two nodal lines. Using Wannier charge centers constrained by the nonsymmorphic mirror Mz, the authors show that in this region one WCC is pinned at the pi-invariant position, giving rise to topological drumhead surface states on the (001) surface; where the two nodal lines overlap in projection, the total Berry phase is 2pi = 0 modulo 2pi, so no topological state is expected, illustrating the 'Z2 modular arithmetic.' Slab calculations and ARPES measurements at BESSY II reveal a crescent-shaped drumhead state confined between the surface projections of NL1 and NL2, split into two branches by SOC, together with a trivial floating band near X. The paper concludes that this is the first complete characterization of topological surface states in the square-net nodal-line family of materials.","tokens_in":12840,"tokens_out":6410,"duration_ms":68263,"significance":"If the claims hold, the paper provides a clear example in which surface states are demonstrably confined to the region of nontrivial Berry phase in the surface Brillouin zone, and it explicitly demonstrates how two pi contributions cancel when nodal lines overlap in projection. The analysis in terms of Wannier charge centers and Mz eigenvalue data is a parameter-free, symmetry-based explanation of the surface-state pattern, and the ARPES data directly support the predicted spatial extent of the drumhead ribbon. The careful separation of topological drumhead states from the trivial floating band is a valuable control. The principal weakness is that the topological classification is established only without spin-orbit coupling, while the material has SOC that gaps the nodal lines; the robustness of the classification to SOC is argued but not certified by a spinful Wilson-loop computation. This weakness is localized and addressable, so the work is a strong candidate for publication after revision.","major_comments":[{"comment":"The central topological claim concerns the real material, yet the Wilson-loop and Wannier-charge-center analysis in Section III.B is performed for a spinless 14-band Hamiltonian. The authors state that the nodal lines in ZrSiTe gap once SOC is included, and they argue that SOC can be treated as a small perturbation. However, no spinful Wilson-loop spectrum or spinful Mz-eigenvalue mapping is presented. The spinful slab spectral function in Fig. 4 shows that surface spectral weight remains, but a surface spectral function cannot certify the bulk topological classification; trivial surface states such as the floating band also appear in regions with zero Berry phase. Because the conclusion that the ARPES-observed drumhead ribbon is topologically required in ZrSiTe depends on the spinless classification surviving SOC, this is a load-bearing gap. I recommend either (i) computing the Wilson-loop spectrum with SOC along the same kx cuts as in Fig. 3(c) and tabulating the resulting N_+1, N_-1, and N_{alpha,alpha*} counts for the 28 occupied spinor bands, or (ii) explicitly rephrasing the central claim as a prediction of the spinless model that is consistent with, but not fully certified by, the SOC-included calculations.","section":"Section III.C"}],"minor_comments":[{"comment":"The sentence 'Therefore, two WCCs are quantized to 0, pi respectively, i.e. both Pos.1a and 1b are occupied' is ambiguous because in the immediately preceding region Table I gives N_+1=1 and N_-1=1; please clarify the correspondence between the N_+1/N_-1 counts and the WCCs at Pos.1a and Pos.1b.","section":"Section III.B"},{"comment":"The term 'Z2 modular arithmetic' is evocative but is never explicitly defined; please provide a precise definition, e.g., that the Berry phase takes values in {0, pi} and combines additively modulo 2pi, so that pi + pi = 0.","section":"Abstract and Section I"},{"comment":"The claim of 'excellent agreement' between ARPES and calculation is based on visual overlap of the traced drumhead region with the calculated projection; a quantitative comparison, such as the energy dispersion of the drumhead branches along cuts 1-4 or the k-space width of the ribbon as a function of energy, would strengthen the experimental confirmation.","section":"Section IV"},{"comment":"The red line superimposed on the calculated Fermi-surface projection is described only as 'the bulk nodal line'; please specify which nodal line or combined projection is shown and whether it is obtained with or without SOC.","section":"Fig. 5(c)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a strong combined theory-experiment study, and the ARPES data are compelling. My main concern, consistent with the stress-test note, is that the spinless WCC argument does not automatically survive the SOC-induced gapping of the nodal lines; the authors should either supply a spinful Wilson-loop check or soften the claim about topological protection in the real material. This is fixable within the scope of the manuscript, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid and readable study of ZrSiTe, and it earns its main claim: drumhead surface states appear exactly where the spinless Wilson-loop Berry phase is π, and the ARPES data show them in that region. The Z2 modular arithmetic framing is a nice conceptual addition, not just a reformulation. The paper does something genuinely new: it gives the first complete surface-state characterization for the square-net nodal-line family, separating topological drumhead states from the trivial floating band. The Wilson-loop analysis is rigorous in the spinless limit, and the experimental data support the predicted spatial extent of the drumhead ribbon.\n\nThe soft spots are real, but they are manageable. Section III.C is the weakest. The authors concede that SOC gaps the nodal lines, which means the spinless Z2 classification does not directly apply to the real material. They argue perturbatively and show a spinful spectral function with a small splitting. The stress-test note is fair: a spectral function cannot certify that the bulk WCC structure survives. A SOC Wilson loop, or at least a symmetry-based argument, would close the gap. The ARPES comparison is also visual rather than quantitative; the ribbon boundaries and dispersions are not fitted or extracted with error bars. That is a minor point for a study like this. Finally, the absence of shared code is a reproducibility limitation, not a correctness issue.\n\nOverall, the paper deserves a serious referee. The central argument holds up in its spinless limit, and the experimental confirmation gives independent support. I would recommend acceptance after the authors either provide a SOC Wilson-loop check or soften the topological-protection claim for the real material, and ideally add a quantitative overlay of the predicted and measured ribbon boundaries.","headline":"Solid, well-executed study of ZrSiTe drumhead states with a useful Z2 framing; the SOC treatment is the one soft spot that should be addressed in revision.","tokens_in":13371,"tokens_out":2813,"would_cite":true,"duration_ms":27558,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"In the nodal-line semimetal ZrSiTe, a confined Berry phase of π makes topological drumhead surface states appear exactly between the surface projections of two nodal lines, and ARPES confirms the prediction.","keywords":["nodal line semimetal","drumhead surface states","Berry phase","Wilson loop","Wannier charge centers","ZrSiTe","ARPES","Z2 topology"],"falsifier":"Repeat the Wilson-loop computation with spin-orbit coupling included in the DFT Hamiltonian: if the number of Wannier charge centers at the unit-cell boundary (the $\\phi=\\pi$ positions) changes parity relative to the no-SOC calculation anywhere inside the predicted drumhead region, the $\\mathbb{Z}_2$ classification fails and the drumhead ribbon would not appear where claimed. Experimentally, a spin-resolved ARPES map that resolves the two SOC-split drumhead branches could check whether they remain confined between the NL1 and NL2 projections at all binding energies.","tokens_in":12460,"feed_emoji":"🥁","tokens_out":9424,"duration_ms":83335,"temperature":0.7,"pith_summary":"ZrSiTe is a layered material whose band structure contains two nearly flat nodal lines, one in the $k_z=0$ plane and one in the $k_z=\\pi$ plane. The paper shows that these nodal lines leave a precise fingerprint on the (001) surface: a Berry phase of $\\pi$, computed by Wilson loops along the stacking direction, exists only in the surface region where exactly one of the two nodal lines projects. In the region where both project, their Berry phases add to $2\\pi$, which is the same as zero, so no topological state is allowed. This rule, a kind of $\\mathbb{Z}_2$ modular arithmetic, predicts a ribbon of topological drumhead surface states bounded by the nodal-line projections, and angle-resolved photoemission measurements place the observed surface states exactly in that ribbon. The paper also separates these topological states from a coexisting trivial floating band surface state that arises from surface symmetry breaking, giving the first complete characterization of topological surface states in the square-net nodal-line semimetal family.","feed_headline":"Drumhead states in ZrSiTe pinned by a π Berry phase","feed_subtitle":"ARPES confirms the topological surface states live only between the two nodal-line projections, as the Z2 rule predicts.","key_machinery":"The central object is the Wilson loop along $\\mathbf{k}_z$: $W(\\ell)=\\mathcal{P}\\exp[-\\oint_\\ell d\\mathbf{k}\\cdot\\mathbf{A}(\\mathbf{k})]$, whose determinant defines the Berry phase $\\gamma$ via $e^{i\\gamma}=\\det W$. The glide mirror $\\bar{M}_z$ reverses the loop, forcing $\\gamma$ to be quantized to $0$ or $\\pi$. The eigenvalues of the Wilson loop are the Wannier charge centers (WCCs) $\\phi_i(k_x,k_y)$; under $\\bar{M}_z$ they sit either at $\\phi=0$, at $\\phi=\\pi$, or in complex-conjugate pairs $(\\lambda,-\\lambda)$. An odd number of WCCs at the unit-cell boundary ($\\phi=\\pi$) corresponds to a Berry phase $\\pi$ and, by slab geometry, to topological drumhead surface states. An exact algorithm (Appendix A) derives the number of WCCs at these positions from the $\\bar{M}_z$ eigenvalues of occupied bands at $k_z=0$ and $k_z=\\pi$, which is how the paper obtains the modular arithmetic: the Berry phases of the two nodal lines add modulo $2\\pi$, so overlapping projections sum to $2\\pi\\equiv 0$ and kill the drumhead states.","core_discovery":"On the (001) surface of ZrSiTe, the paper establishes that topological drumhead surface states are controlled by a $\\mathbb{Z}_2$-quantized Berry phase $\\gamma(k_x,k_y)$ obtained from Wilson loops along $k_z$. The glide mirror $\\bar{M}_z$ forces $\\gamma$ to be $0$ or $\\pi$. Where the surface projections of nodal lines NL1 ($k_z=0$) and NL2 ($k_z=\\pi$) overlap, each line contributes $\\pi$, so $\\gamma=2\\pi\\equiv 0$ and no drumhead states form; where exactly one line projects, $\\gamma=\\pi$ and drumhead states exist; elsewhere $\\gamma=0$. The same $\\mathbb{Z}_2$ structure acts as modular arithmetic on the surface states, so states derived from different nodal lines hybridize and gap in the overlap region. The paper confirms this with slab calculations, with Wilson-loop spectra derived from the $\\bar{M}_z$ eigenvalues of occupied bands, and with ARPES data that locate the drumhead states precisely between NL1 and NL2. It additionally identifies a topologically trivial floating band state whose steep dispersion distinguishes it from the drumhead states.","pith_inferences":["The same $\\mathbb{Z}_2$ modular arithmetic should apply to any material with two symmetry-related nodal lines of the same class projecting onto a surface: the drumhead ribbon is the region of the surface Brillouin zone where an odd number of nodal lines project, i.e., the XOR of the projected disks, not their union.","If a material could be tuned so that the two nodal-line projections exactly coincide, the $\\mathbb{Z}_2$ sum would annihilate the drumhead states across the whole ribbon, providing a topological transition that could be driven by pressure or strain.","The bulk-only Wilson-loop mapping in the paper could be converted into a symmetry-indicator screening tool to search square-net materials for drumhead surface states without expensive slab calculations.","The coexistence of a topological and a trivial surface state with very different dispersions might let scanning tunneling spectroscopy isolate the topological contribution by comparing regions inside and outside the predicted Berry-phase ribbon."],"forward_implications":["The drumhead states appear as a ribbon restricted to the region where exactly one nodal line projects; ARPES confirms their location in ZrSiTe.","Where the two nodal-line projections overlap, the drumhead states hybridize and gap, a direct consequence of the $\\mathbb{Z}_2$ modular arithmetic that can be checked in other materials.","The Wilson-loop spectrum, obtained from symmetry eigenvalues, predicts surface-state topology without computing surfaces, so it can be used as a bulk-only diagnostic.","The topological drumhead states and the trivial floating band states coexist but have clearly different dispersions, so they can be separated in ARPES and in transport experiments.","Spin-orbit coupling, though it gaps the nodal lines, splits the drumhead states into two branches while leaving them clearly visible, so the topological surface states should survive perturbatively in real ZrSiTe."],"supporting_citations":[{"why":"Provides the Wilson-loop formalism relating the loop spectrum to topological invariants and surface states.","marker":"[50]"},{"why":"Establishes the mapping between Wilson-loop eigenvalues and the symmetry eigenvalues of occupied bands.","marker":"[51]"},{"why":"Extends Wilson-loop analysis to nonsymmorphic symmetries, underpinning the glide-mirror constraint used here.","marker":"[52]"},{"why":"Identified the square-net family of nodal-line semimetals and the nodal-line structure of the parent compound ZrSiS.","marker":"[43]"},{"why":"Reported the synthesis and crystal characterization of ZrSiTe used in the ARPES experiment.","marker":"[44]"},{"why":"Documented the floating band surface state in ZrSiS and its origin in surface symmetry breaking, the comparison for the trivial state.","marker":"[57]"},{"why":"Supplies the WannierTools package used to compute nodal lines and Wilson loops.","marker":"[48]"}],"fun_headline_variants":["Z2 Berry phase dictates drumhead states in ZrSiTe","Modular arithmetic of surface states in ZrSiTe","ARPES confirms Z2-governed drumhead states in ZrSiTe","Drumhead states arise only between nodal-line projections","Z2 modular arithmetic pins drumhead states in ZrSiTe"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The classification is carried out without spin-orbit coupling, and the paper assumes SOC can be treated as a small perturbation that leaves the $\\mathbb{Z}_2$ Berry phase and the drumhead region essentially unchanged.","fun_headline_variants_meta":{"raw":{"variants":["Z2 Berry phase dictates drumhead states in ZrSiTe","Modular arithmetic of surface states in ZrSiTe","ARPES confirms Z2-governed drumhead states in ZrSiTe","Drumhead states arise only between nodal-line projections","Z2 modular arithmetic pins drumhead states in ZrSiTe"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000633,"raw_usage":{"total_tokens":2939,"prompt_tokens":978,"completion_tokens":1961,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":594,"completion_tokens_details":{"reasoning_tokens":1886}},"tokens_in":594,"tokens_out":1961,"duration_ms":13293,"temperature":1.0,"reasoning_tokens":1886,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:57:55.218654+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the Wilson-loop computation with spin-orbit coupling included in the DFT Hamiltonian: if the number of Wannier charge centers at the unit-cell boundary (the $\\phi=\\pi$ positions) changes parity relative to the no-SOC calculation anywhere inside the predicted drumhead region, the $\\mathbb{Z}_2$ classification fails and the drumhead ribbon would not appear where claimed. Experimentally, a spin-resolved ARPES map that resolves the two SOC-split drumhead branches could check whether they remain confined between the NL1 and NL2 projections at all binding energies.","supporting_citations":[{"cited_title":"Alexandradinata , author Z","cited_arxiv_id":null,"evidence_quote":"Provides the Wilson-loop formalism relating the loop spectrum to topological invariants and surface states."},{"cited_title":"Alexandradinata , author X","cited_arxiv_id":null,"evidence_quote":"Establishes the mapping between Wilson-loop eigenvalues and the symmetry eigenvalues of occupied bands."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Identified the square-net family of nodal-line semimetals and the nodal-line structure of the parent compound ZrSiS."},{"cited_title":"Topp , author J","cited_arxiv_id":null,"evidence_quote":"Reported the synthesis and crystal characterization of ZrSiTe used in the ARPES experiment."},{"cited_title":"u neis , author L. M \\","cited_arxiv_id":null,"evidence_quote":"Documented the floating band surface state in ZrSiS and its origin in surface symmetry breaking, the comparison for the trivial state."},{"cited_title":"Wu , author S","cited_arxiv_id":null,"evidence_quote":"Supplies the WannierTools package used to compute nodal lines and Wilson loops."}],"review_version":1}