{"id":"191180b9-efae-455f-887b-4c965ffa305d","arxiv_id":"2412.09561","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"MX2 monolayers in space group P-3m1 are identified as orbital-obstructed atomic insulators whose finite hexagonal flakes show a bulk-predicted filling anomaly of two, four, or eight electrons.","lead":"This paper shows that certain two-atom-thick crystals called MX2 monolayers host a subtle form of topological insulation, where electron charge centers from the crystal symmetry do not sit where the atoms are. The authors predict that these materials trap electrons at corners when cut into flakes, producing a measurable mismatch in how many electrons can fill the system.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Anomaly magnitudes are not fixed by the bulk EBR solution: they require an assumed orbital-support count, and for SnS2 the flake deficit (8) matches the bulk count (4) only modulo 12.","rationale":"The paper is internally consistent, and the tight-binding model plus finite-flake spectra provide genuine evidence that an orbital obstruction and filling anomaly exist in these materials. The MgCl2 and ZrS2 cases show encouraging bulk-flake agreement, and the authors explicitly acknowledge the modulo-12 subtlety for SnS2, which is an honest limitation. However, the central quantitative claim — that the bulk EBR decomposition alone determines whether the anomaly is 2, 4, or 8 — has a soft spot precisely where the reader located it: the decomposition is non-unique (y2 free in Eq. 10), and the mapping from the pinned solution X_O to an electron count relies on assumed atomic valence configurations rather than on a computation of the actual support of the occupied Wannier functions. The SnS2 flake showing 8 while the bulk count is 4 makes the modulo-12 nature concrete and shows that the exact integer is not a unique bulk-determined number. Because the paper already conditions the claims (modulo 12, and with a candidate list not individually verified by flakes), the reader's CONDITIONAL verdict remains appropriate; the missing deposited inputs and Wannier-center check are exactly what a conditional acceptance should require.","tokens_in":20877,"tokens_out":12999,"duration_ms":116314,"concrete_test":"Perform an explicit Wannier-center calculation for the occupied valence bands of MgCl2, ZrS2, and SnS2 using the DFT wavefunctions (e.g., Wannier90 or the position operator projected on the occupied subspace), and count per formula unit the number of Wannier centers at the 1a and 2d Wyckoff positions (and whether any fall at 3e). If the number of centers associated with the 2d WP differs from the assumed one (MgCl2) or two (ZrS2, SnS2) electrons per chalcogen, the anomaly values in Table II change. This directly tests the orbital-support assumption that separates the MgCl2 and ZrS2 cases despite their identical B vectors.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Section IV and abstract) is that the bulk EBR data predict the exact number of electrons in in-gap states of a neutral symmetric flake. The step from the pinned solution X_O to an integer anomaly is not determined by the TQC algebra alone. Two additional inputs are imposed by hand. (i) In Eq. 10, all y2 ≠ 0 solutions are discarded solely because they would place charge at the unoccupied 3e WP; if a y2 ≠ 0 decomposition were physical, the predicted anomalies and Table II would change. (ii) The number of electrons carried by the pinned 2d EBR is assigned from atomic valence configurations: one electron per Cl in MgCl2 but two per S in ZrS2 and SnS2. This is why MgCl2 and ZrS2 have identical B vectors (Eqs. 5 and 16) and identical X_O (Eqs. 15 and 19), yet are assigned anomalies 2 and 4. The finite-flake spectra do not independently fix this support assignment; they are used to read off the anomaly. SnS2 exposes the residual freedom: the flake spectrum shows a deficit of eight electrons, whereas the bulk obstructed count is four, and the two are identified only by declaring the anomaly well defined modulo 12. Hence the bulk EBR decomposition fixes the anomaly only modulo the 12-electron MAI and modulo the assumed orbital support, so the exact values in Table II are not fully derived from bulk symmetry data.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies monolayer MX₂ compounds with space group P̄3m1 and argues that they realize orbital-obstructed atomic insulators (OOAIs). Using topological quantum chemistry (TQC), a minimal six-band tight-binding model, and first-principles DFT calculations, it predicts that neutral symmetric hexagonal flakes display a filling anomaly of two obstructed states for MgCl₂, four for ZrS₂, and four modulo twelve for SnS₂, with the predictions verified in finite-flake spectra in Fig. 4g–i. The work also discusses the relation between edge states and corner localization, and surveys a broader set of candidate materials in Table II.","tokens_in":21169,"tokens_out":10015,"duration_ms":91739,"significance":"If correct, the paper would establish a concrete bulk-boundary correspondence for 2D orbital-obstructed atomic insulators, showing that symmetry-data vectors and EBR decompositions can predict the number of electrons that must occupy in-gap states in finite flakes. The manuscript is in many ways exemplary: the EBR matrix and Smith decomposition are given explicitly, the tight-binding model is fully defined in the Appendix, and the flake spectra are computed with an independent code (SIESTA), providing a genuine nontrivial check of the bulk prediction. The identification of a filling anomaly for the fully filled 2d-EBR regime (MgCl₂ and ZrS₂) extends earlier work on half-filled cases such as SnS₂. However, as detailed in the major comments, the exact anomaly integers are not uniquely fixed by the bulk symmetry data alone; they require additional assumptions about the EBR decomposition and the number of electrons assigned to the obstructed EBR. These assumptions are validated for the three computed flakes, but not derived from first principles, which weakens the predictive claim for the rest of Table II.","major_comments":[{"comment":"The generic solution X in Eq. (10) contains a free integer parameter y₂; all y₂ ≠ 0 solutions are discarded solely because they would place charge centers at the unoccupied 3e WP. However, an obstructed atomic insulator (OAI) is precisely a phase whose Wannier charge centers are not located at the occupied Wyckoff positions (Ref. [16]), so the criterion 'charge centers at 3e are unphysical' is an additional physical assumption rather than a consequence of the symmetry data B. If a y₂ ≠ 0 decomposition were chosen, the pinned solution X_O and hence the predicted filling anomaly would change. The paper should either justify the choice y₂ = 0 from the orbital character of the DFT wavefunctions or explicitly state that X_O is one of several admissible decompositions and that the anomaly prediction depends on that choice, which currently is not made.","section":"Section IV, Eq. (10) and following text"},{"comment":"MgCl₂ and ZrS₂ have identical symmetry data vectors B (Eqs. 5 and 16) and identical pinned solutions X_O (Eqs. 15 and 19), yet are assigned filling anomalies of 2 and 4 solely because the authors assume one electron per Cl atom in the 2d EBR for MgCl₂ and two electrons per S atom for ZrS₂. This demonstrates that the exact anomaly magnitude is not fixed by the bulk EBR decomposition; it depends on an assumed atomic valence configuration that is not derived from the symmetry analysis. The finite-flake spectra in Fig. 4g,h independently confirm the assigned values for these two materials, but they do not validate the orbital-support assignment for the remaining compounds in Table II. A first-principles criterion (e.g., projected orbital occupations or Wannier-function spreads) should be provided for the number of electrons forming the obstructed EBR, or the claim should be restricted to a modulo-12 statement.","section":"Section IV, MgCl₂ and ZrS₂ paragraphs, Eqs. (15) and (19)"},{"comment":"For SnS₂ the finite flake shows a deficit of eight electrons at the neutrality point, whereas the bulk obstructed count is four; the two are reconciled by declaring the filling anomaly well defined only modulo 12 because the movable atomic insulator (MAI) carries 12 electrons. This means that, for this material, the bulk symmetry data alone determine the anomaly only modulo 12, and the exact integer is fixed by the flake calculation (or by an additional real-space assumption). The abstract and Section III C state that the filling anomaly is 'directly associated with the bulk configuration,' which overstates the precision of the bulk-only prediction. The modulo-12 qualification should appear prominently in the general derivation and in the abstract, not only in the footnote of Table II, and the paper should clarify whether the modulo-12 ambiguity also affects the exact values 2 and 4 assigned to MgCl₂ and ZrS₂.","section":"Section IV, SnS₂ paragraph and Table II footnote"}],"minor_comments":[{"comment":"The phrase 'orbital-mediated atomic obstruction requires the presence of orbitals that have no support in real space' is confusing and appears to contradict the body's definition in Section I, where OOAIs are defined as having all EBRs from occupied WPs but with a mismatch between the expected and actual orbital-induced representations. Please rephrase, e.g., 'EBRs whose induced charge centers are not supported at the expected atomic orbitals.'","section":"Abstract"},{"comment":"The sentence contains a typo: 'one o more solutions' should read 'one or more solutions.'","section":"Section IV, after Eq. (6)"},{"comment":"'which traduces in a four-state filling anomaly' should be 'which translates into a four-state filling anomaly.'","section":"Section IV, ZrS₂ paragraph"},{"comment":"The statement that the only symmetry-allowed real-space arrangement is '2 electrons at the 1a WP and one electron per site of the 2d WP' is specific to the minimal model with four electrons per unit cell; this qualification should be stated explicitly to avoid confusion with the real materials, which have more valence electrons.","section":"Section III A"},{"comment":"The caption of Fig. 2d does not list the tight-binding parameters used for the flake spectrum; since the same model is used for Fig. 3c, please either refer to the caption of Fig. 3c or provide the parameter values in the Fig. 2d caption.","section":"Figure 2d"},{"comment":"The footnote 'Recall that this value is well defined only modulo 12' is an important qualification that should be integrated into the main text where the anomaly values are introduced (Section IV, MgCl₂ paragraph), not only in the table footnote.","section":"Table II"}],"recommendation":"major_revision","confidential_remarks":"The paper presents a clean and useful demonstration of filling anomalies in specific 2D materials, with transparent computational methods and independent flake verification. The main weakness is that the exact anomaly integers are not determined by the bulk EBR data alone; they rely on an assumed orbital-support assignment and on a particular choice of EBR decomposition (y₂ = 0). This is not a fatal flaw, because the flake calculations confirm the assigned values for the three highlighted materials, but it does mean the abstract and the general claim about prediction from the bulk are oversold. The authors should be asked to explicitly separate what is derived from bulk symmetry (the modulo-12 obstruction) from what is assumed from atomic valence configurations, and to either justify the orbital-support assignment from first principles or soften the predictive claim for Table II. The overlap with Ref. [22] for the half-filled 2d EBR case should also be clearly delineated; the fully filled 2d EBR cases (MgCl₂, ZrS₂) are the genuinely new contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take on Arroyo-Gascón et al. The core claim—that monolayers in SG P-3m1 can be orbital-obstructed atomic insulators with a bulk-derived filling anomaly—is plausible and supported by consistent bulk and flake calculations. What's genuinely new is the filling anomaly as the bulk-boundary response of OOAI, including the completely-filled 2d-EBR cases (MgCl2, ZrS2) that go beyond Ref. 22's half-filled case. The tight-binding model gives a nice mechanistic picture, and the three first-principles examples with matching flake spectra are real evidence. The flake calculations are independent of the TB parameters, so there's no circularity.\n\nThe soft spot is the step from bulk EBR data to an exact integer anomaly. The EBR decomposition is non-unique; the authors discard y2≠0 solutions by appealing to unoccupied 3e WP, and they assign the electron count per EBR from nominal atomic valence configurations. That's physically reasonable, but it's an input, not a derivation. The consequences show up in MgCl2 vs ZrS2: identical B vectors and identical X_O, yet anomalies 2 vs 4 solely from the orbital-support assignment. SnS2 is the clearest example: the flake deficit is 8, the bulk obstructed count is 4, and they identify them modulo 12. They say so explicitly, which is honest, and it doesn't invalidate the mechanism, but it does mean Table II's exact values are conditional. Also, no input data are deposited, so reproducing Table II isn't possible without redoing the calculations.\n\nWho should read it: people working on TQC, higher-order topology, and 2D materials. It's a solid contribution, not revolutionary. The main requested revision would be a Wannier-center calculation to pin the 2d-EBR support and deposited inputs.\n\nSend it to review. It deserves referee time.","headline":"Worth refereeing: the paper establishes a filling-anomaly response for orbital-obstructed atomic insulators in common MX2 monolayers, though the integer anomaly values depend on assumptions the authors make explicit.","tokens_in":21740,"tokens_out":2304,"would_cite":true,"duration_ms":21144,"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":"Bulk symmetry data determines the number of in-gap states that neutral symmetric flakes of SG No. 164 $MX_2$ monolayers must host, via an orbital-obstructed atomic insulator mechanism.","keywords":["orbital-obstructed atomic insulator","filling anomaly","higher-order band topology","elementary band representations","space group P-3m1","MX2 monolayers","corner states","two-dimensional materials"],"falsifier":"Compute the symmetry data vector $B$ for $\\mathrm{ZrS_2}$ with the Zr $4s$ and $4p$ semicore states included in the valence manifold and re-solve the EBR equation: if the pinned remainder is no longer the $2d$-induced EBR with four electrons, while the DFT flake spectrum still shows four in-gap states, then the orbital-support choice, not the bulk symmetry data, is carrying the prediction. Conversely, a direct STM measurement of the corner charge on a neutral, symmetric hexagonal flake of $\\mathrm{MgCl_2}$ could confirm whether the neutrality point really sits inside the two-state in-gap manifold as claimed.","tokens_in":20653,"feed_emoji":"⚛️","tokens_out":11930,"duration_ms":97860,"temperature":0.7,"pith_summary":"This paper shows that a family of two-dimensional $MX_2$ compounds with space group $P\\bar{3}m1$—many of them already grown in the lab—are orbital-obstructed atomic insulators: their valence bands admit a Wannier description, but the electron charge centers demanded by the momentum-space band representation do not sit at the atomic positions where the electrons actually reside. Because of that mismatch, a neutral, symmetric hexagonal flake cannot be simultaneously gapped and charge neutral, so it must place a definite number of electrons in in-gap states, a filling anomaly. The paper demonstrates that this number is readable directly from the bulk decomposition into elementary band representations, and confirms the prediction with a minimal tight-binding model and first-principles spectra for three representative monolayers: two obstructed states for $\\mathrm{MgCl_2}$, four for $\\mathrm{ZrS_2}$, and four modulo twelve for $\\mathrm{SnS_2}$. If the claim holds, the bulk symmetry data of any such monolayer tells exactly how many corner electrons its finite flake must host.","feed_headline":"Bulk math predicts exact corner-state filling in MX2 monolayers","feed_subtitle":"For MgCl2, ZrS2, SnS2 the count is 2, 4, and 4 modulo 12, from symmetry data alone.","key_machinery":"The load-bearing tool is the elementary band representation (EBR) decomposition of the occupied bands, computed by Smith normal form from the matrix equation $\\mathrm{EBR}\\cdot X = B$, where $B$ is the vector of irreducible representation multiplicities at the high-symmetry points $\\Gamma$, $M$, $K$ and the EBR matrix encodes which atomic-orbital-induced representations exist in the space group. The argument works by identifying the movable band representations—charge centers that can slide adiabatically between the $1a$ and $2d$ Wyckoff positions (via the $6i$ site)—and subtracting them from each admissible insulator solution, leaving a pinned remainder. The electrons in the pinned remainder are the obstructed ones, and their count is the predicted filling anomaly.","core_discovery":"The central discovery is that orbital-mediated atomic obstruction occurs in SG $P\\bar{3}m1$ monolayers and produces a bulk-detectable filling anomaly. For a valence manifold described by elementary band representations, solving the equation $\\mathrm{EBR}\\cdot X = B$ for the symmetry-data vector $B$ gives several admissible integer decompositions; the physically retained ones have no charge density at the unoccupied $3e$ Wyckoff position, and after subtracting the movable atomic-insulator contribution, a pinned remainder survives consisting of a $2d$-induced EBR whose charge centers have no (or only partial) support at the atoms. The number of electrons carried by that pinned EBR—one per Cl atom in $\\mathrm{MgCl_2}$, two per S atom in $\\mathrm{ZrS_2}$, and two per S atom in $\\mathrm{SnS_2}$—matches the in-gap state count seen in finite flake spectra, establishing a bulk-boundary correspondence for the orbital obstruction. The paper also shows that the filling anomaly is robust to the presence or absence of edge states, and that it can appear for both fully filled and half-filled $2d$-induced EBRs.","pith_inferences":["One testable consequence the paper does not spell out: because the $\\mathrm{SnS_2}$ anomaly is defined modulo 12, adding or removing one electron per unit cell in a gated flake should cycle the in-gap population through the same twelve-state manifold, a signature that could be looked for in transport or charging experiments.","The predicted count is hostage to the assumed atomic valence configuration; recomputing the symmetry data with for example Zr semicore $4s$, $4p$ states in the valence manifold could shift the pinned remainder and would therefore change the expected anomaly for the same material.","The same 'pinned EBR without full real-space support' criterion could be screened over other layer groups that contain movable Wyckoff positions, suggesting that orbital-obstructed insulators are more common than the current list.","The paper links the presence of in-gap corner states to a non-protected edge-to-corner correspondence in the ribbon spectrum; if that correspondence is general, a simple ribbon band-structure calculation could serve as a diagnostic for corner-state formation in other materials."],"forward_implications":["If the bulk decomposition is right, corner-state or in-gap-state counts can be predicted from bulk symmetry data alone, without any finite-flake calculation.","The filling anomaly is robust even when edge states are absent or buried in the bulk bands, since it is tied to the pinned EBR, not to boundary state localization.","The mechanism extends beyond the half-filled $2d$-EBR case previously studied: compounds with fully filled $2d$-induced EBRs (several transition-metal dichalcogenides) also show the obstruction.","Standard symmetry-indicator invariants miss the anomaly for the halogen-based monolayers discussed here, so the full EBR decomposition is required to count obstructed states.","The same Smith-decomposition workflow can be applied to any other material in SG No. 164, and the paper's survey assigns a specific anomaly value (2 or 4, modulo 12) to a long list of candidate monolayers."],"supporting_citations":[{"why":"Defines elementary band representations and the atomic-limit classification that the paper uses to identify obstructed insulators.","marker":"[8]"},{"why":"Establishes that EBRs are induced from maximal Wyckoff positions, the basis for reading charge centers in real space.","marker":"[9]"},{"why":"Introduces the orbital-obstructed atomic insulator concept that the paper applies to SG No. 164.","marker":"[17]"},{"why":"Supplies the filling anomaly and fractional corner charge framework used to diagnose the obstruction in finite flakes.","marker":"[19]"},{"why":"Provides the spin-orbit coupled version of fractional corner charges that justifies the counting in the double-group setting.","marker":"[20]"},{"why":"Prior work on half-filled $2d$ EBR monolayers that this paper extends to fully filled and OOAI cases.","marker":"[22]"},{"why":"Smith decomposition algorithm for solving $\\mathrm{EBR}\\cdot X = B$, the tool that yields the admissible decompositions and pinned remainders.","marker":"[58]"}],"fun_headline_variants":["Orbital obstruction in 2D predicts corner states from bulk alone","Bulk symmetry alone fixes corner state counts in 2D monolayers","Orbital obstruction: bulk predicts corner states in 2D","Exact corner-state filling from bulk symmetry in MX2 monolayers","2D orbital obstruction: corner states predicted without edge info"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"All of the predicted numbers depend on which integer solution of the EBR equation is taken as the physical one and on the assumed valence-electron configuration: if the charge centers were allowed to sit at the unoccupied $3e$ Wyckoff position, or if the valence manifold includes different orbitals (for example semicore $d$ states), the predicted two, four, or four electrons could change, even though the finite-flake spectra themselves are unchanged.","fun_headline_variants_meta":{"raw":{"variants":["Orbital obstruction in 2D predicts corner states from bulk alone","Bulk symmetry alone fixes corner state counts in 2D monolayers","Orbital obstruction: bulk predicts corner states in 2D","Exact corner-state filling from bulk symmetry in MX2 monolayers","2D orbital obstruction: corner states predicted without edge info"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000811,"raw_usage":{"total_tokens":3597,"prompt_tokens":1028,"completion_tokens":2569,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":644,"completion_tokens_details":{"reasoning_tokens":2478}},"tokens_in":644,"tokens_out":2569,"duration_ms":19469,"temperature":1.0,"reasoning_tokens":2478,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T16:56:18.542160+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the symmetry data vector $B$ for $\\mathrm{ZrS_2}$ with the Zr $4s$ and $4p$ semicore states included in the valence manifold and re-solve the EBR equation: if the pinned remainder is no longer the $2d$-induced EBR with four electrons, while the DFT flake spectrum still shows four in-gap states, then the orbital-support choice, not the bulk symmetry data, is carrying the prediction. Conversely, a direct STM measurement of the corner charge on a neutral, symmetric hexagonal flake of $\\mathrm{MgCl_2}$ could confirm whether the neutrality point really sits inside the two-state in-gap manifold as claimed.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes that EBRs are induced from maximal Wyckoff positions, the basis for reading charge centers in real space."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the filling anomaly and fractional corner charge framework used to diagnose the obstruction in finite flakes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the spin-orbit coupled version of fractional corner charges that justifies the counting in the double-group setting."},{"cited_title":"Schindler, M","cited_arxiv_id":null,"evidence_quote":"Prior work on half-filled $2d$ EBR monolayers that this paper extends to fully filled and OOAI cases."},{"cited_title":"Manna, S","cited_arxiv_id":null,"evidence_quote":"Smith decomposition algorithm for solving $\\mathrm{EBR}\\cdot X = B$, the tool that yields the admissible decompositions and pinned remainders."}],"review_version":1}