{"id":"4e9ad502-ce61-4ab2-aed9-317d72e2de71","arxiv_id":"2509.08320","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A volume-normalized feature of the spin-resolved density of energy is proposed as a descriptor correlating with the temperature and field needed to stabilize skyrmions, pointing to noncentrosymmetric semi-metals as ambient-condition candidates.","lead":"This perspective proposes that a feature in a material's computed spin density of energy, normalized by volume, predicts how hard it is to make magnetic skyrmions appear: higher values point to higher required fields or temperatures. It uses this to argue that noncentrosymmetric semi-metals with slightly more than half-filled d-orbitals are the best route to room-temperature skyrmions.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Table S2's 'EReq/Å3' values are not unit-cell-volume normalized as stated; the Figure 7b trend rests on an unstated per-magnetic-atom normalization and a non-algorithmic EReq extraction.","rationale":"The reader's weakest assumption identifies both the asserted physical mapping from a DOE feature to skyrmion-formation energy and the Table S2 normalization mismatch. My stress-test narrows this to the most concrete, checkable defect: the reported EReq/Å3 values are not EReq/V_cell as claimed, but EReq/(V_cell/N_mag), and the extraction of EReq is not specified enough to be reproducible. Because Figure 7b is the central quantitative support for the proposed descriptor, this is load-bearing. However, the paper is a perspective that also offers qualitative structure–property heuristics (d vs f electrons, half-filled vs slightly above half-filled orbitals, DM vs RKKY), which are not invalidated by the normalization issue. The reader's CONDITIONAL verdict—publish with clarifications and a corrected normalization—remains the right assessment, so no verdict change is needed. The proposed concrete test would settle whether the quantitative correlation survives; if it does not, the framework should be reframed as purely qualitative.","tokens_in":27316,"tokens_out":7190,"duration_ms":79588,"concrete_test":"Independently recompute EReq for all 14 hosts in Table S2 from the integrated DOEs using a fixed algorithmic rule (first local minimum of ∫DOEs dE in [EF, EF+Δ], with Δ set by the valence-band width) and check the reported values. Then re-plot Figure 7b using both normalizations, EReq/V_cell and EReq/(V_cell/N_mag), and compute rank correlations (e.g., Spearman ρ) of EReq/Å3 with T, H, and a joint T/H severity metric. If the 'larger EReq/Å3 ⇒ larger T and/or H' trend does not survive the correct unit-cell normalization or depends strongly on the extraction rule, the central quantitative claim is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative core of the paper—Figure 7b's claim that larger EReq/Å3 implies larger skyrmion-formation T and/or H—is not supported by the reported numbers under the normalization stated in the text. Section 2.5 and the Summary say EReq is normalized by unit-cell volume, but the Table S2 entries are EReq·N_mag/V_cell, i.e., EReq divided by volume per magnetic atom. For FeGe, EReq=3.88 eV and V=103.82 Å3 gives 0.037 eV/Å3, not the listed 0.149; for Cu2OSeO3, 2.42 eV/710.93 Å3 = 0.0034 eV/Å3, not 0.054. The correction factor varies across materials (×4 for B20 compounds, ×16 for Cu2OSeO3, ×8 for VOSe2O5), so the ranking in Figure 7b is not invariant to the normalization choice. Additionally, EReq is identified as 'a local minimum right above 4 eV where its first derivative is zero' with no algorithmic definition, window, or tolerance, and the SI concedes that extracting EReq from ground-state collinear DFT is 'likely to be an oversimplification' for hosts with helical ground states. Thus the apparent trend may be an artifact of the inconsistent normalization and subjective feature selection.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a semi-quantitative descriptor, EReq, extracted from the integrated spin density of energy (DOEs) above the Fermi level in collinear spin-polarized DFT, and claims that EReq normalized per unit-cell volume correlates with the temperature and magnetic-field conditions required for skyrmion formation across 14 established skyrmion hosts. The authors combine this with qualitative design rules (d- vs f-electron character, orbital filling, DM vs RKKY interactions) into a 'treasure map' and recommend noncentrosymmetric semi-metals with slightly more than half-filled d-orbitals as the most promising candidates for ambient-condition skyrmions. The central quantitative result is Figure 7b, which plots EReq/Å3 against TSmax and HSmax and is asserted to show that larger EReq/Å3 implies higher formation temperatures and/or fields.","tokens_in":27596,"tokens_out":7482,"duration_ms":72213,"significance":"If the EReq/Å3 correlation were robust, it would provide a computationally cheap, DFT-based screening descriptor for skyrmion-host discovery, connecting ground-state electronic structure to experimentally accessible skyrmion-phase conditions. The manuscript also offers a useful qualitative synthesis of the DM versus RKKY branches, the differing behavior of d- and f-electron systems, and the role of orbital filling. The authors are commendably explicit about several limitations, including the collinear-DFT approximation and the subjective nature of the EReq extraction. However, the quantitative core is not currently established: the stated unit-cell-volume normalization is not what is implemented in Table S2, the EReq feature is not defined algorithmically, one host (Cu2OSeO3) does not exhibit the claimed positive Espin signature, and the descriptor is validated only on the same hosts used to define the trend. The proposed candidate class is essentially a restatement of known room-temperature hosts, so the predictive value is limited as written.","major_comments":[{"comment":"The text states that EReq is normalized by unit-cell volume (§2.5 and Summary), but the Table S2 entries 'EReq per Å3' do not equal EReq/V. For FeGe, 3.88 eV/103.82 Å3 = 0.037 eV/Å3, not the listed 0.149; for Cu2OSeO3, 2.42/710.93 = 0.0034 eV/Å3, not 0.054. The multiplicative correction varies across materials (×4 for B20 compounds, ×16 for Cu2OSeO3, ×8 for VOSe2O5), so the ranking in Figure 7b is not invariant to the normalization choice. Because Figure 7b is the central quantitative claim, this inconsistency must be resolved: either recompute the trend with the stated EReq/V normalization, or explicitly redefine and justify the per-magnetic-atom normalization actually used.","section":"§2.5, Table S2, Figure 7b"},{"comment":"EReq is defined as 'a local minimum right above 4 eV where its first derivative is zero' with no algorithmic specification of the energy window, tolerance, or selection rule. For Cu2OSeO3, the positive Espin criterion is absent, and the authors instead appeal to 'changes in slope' (§2.5), effectively changing the descriptor from case to case. The SI itself concedes that extracting EReq 'solely from the ground-state band structure above EF is likely to be an oversimplification' for hosts with helical ground states. As written, the extraction is not reproducible and the chain from band-structure feature to skyrmion-formation energy is not established.","section":"§2.2, §2.5, SI 'Proposed approach'"},{"comment":"The claim that 'systems with larger EReq/Å3 require higher temperatures and/or magnetic fields to transition to a skyrmion state' is not supported as a monotonic statement by the reported data. Using Table S2, GaMo4S8 has EReq/Å3 = 0.082, TSmax = 25 K, HSmax = 0.1 T, while MnSi has 0.073, 28 K, 0.12 T; the larger-EReq material exhibits both lower T and lower H. Since the 'and/or' formulation is satisfied by almost any monotonic trend in a 3D plot, this counterexample shows that the claimed relationship is not a reliable guide. A quantitative measure (e.g., rank correlation) and explicit handling of outliers are needed.","section":"§2.6, Figure 7b"},{"comment":"The descriptor is validated on the same 14 hosts used to define the trend, with no out-of-sample prediction or leave-one-out cross-validation. The resulting recommendation—noncentrosymmetric semi-metals with slightly above half-filled d-orbitals—is a restatement of the known room-temperature hosts FeGe, Co8Zn9Mn3, Mn2Rh0.95Ir0.05Sn, and Mn1.4Pt0.9Pd0.1Sn listed in §4.3. To support the predictive value claimed in the Summary, the manuscript should identify new candidate materials with explicit EReq/Å3 thresholds and testable skyrmion-formation conditions, or at least perform a leave-one-out analysis to show the descriptor is not merely summarizing the input set.","section":"§5, Table S2, Figure 7"}],"minor_comments":[{"comment":"Typos: '1 Testa river' should be '1 Tesla river'; 'Dzyaloshinkii-Moriya' should be 'Dzyaloshinskii-Moriya'; 'Ruderman-Kittel-Kasuya-Yosdia' should be 'Ruderman-Kittel-Kasuya-Yosida'.","section":"§1, Figure 1 caption"},{"comment":"Typo: 'skrymion' should be 'skyrmion' in 'for skrymion hosts at ambient conditions'.","section":"§4.4"},{"comment":"The 3D scatter plot is difficult to read; material labels are not visible and the 'and/or' relationship is hard to assess. A 2D projection or labeled markers with a rank-order table would improve clarity.","section":"Figure 7"},{"comment":"Equation (1) appears after the text that refers to it, and Equations (3)-(5) are similarly placed after their discussion. Reorder for readability.","section":"Equations (1), (3)-(5)"}],"recommendation":"major_revision","confidential_remarks":"The central correlation relies in part on the authors' own arXiv preprints (refs 86 and 88) for the GdRu2X2 case studies; the editor may wish to consider the extent to which the framework depends on work that has not yet undergone peer review. The Table S2 normalization discrepancy (EReq/Å3 values not equal to EReq/V) suggests the numerical values were produced with a formula different from the one stated in the text, and this should be verified carefully before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper's real contribution is the spin-resolved DOE/DOEs framework: using the integrated spin density of energy to connect electronic instabilities to skyrmion formation conditions, and extending that instability picture to insulating hosts where Fermi-surface nesting does not apply. That is a genuinely fresh angle, and the paper includes new DFT-derived data for 14 known hosts. The isostructural comparisons (GdRu2Si2 vs GdRu2Ge2, MnSi/MnGe/FeGe, the Eu series) are internally consistent and give the framework some credibility. The conceptual narrative is clear and the DM/RKKY classification is competent review material.\n\nThe soft spots are real, and one is load-bearing. The paper repeatedly says Ereq is normalized by unit-cell volume, but Table S2 gives EReq·N_mag/V_cell—equivalently EReq divided by volume per magnetic atom. For FeGe, the stated 0.149 eV/Å3 comes from 3.88×4/103.82, not 3.88/103.82. For Cu2OSeO3, the factor is 16. So the rankings in Figure 7b are not invariant to the stated normalization, and the central claim that larger Ereq/Å3 means higher formation fields/temperatures is not supported by the numbers as reported. This needs to be fixed before the quantitative claim can be taken seriously.\n\nSecond, Ereq extraction is not algorithmic: \"a local minimum right above 4 eV where its first derivative is zero\" is a visual identification, with no window or tolerance. Combined with the collinear-DFT caveat the authors themselves state in the SI, the mapping from a ground-state band-structure feature to the thermodynamic cost of forming a skyrmion phase is asserted, not derived. Third, Cu2OSeO3 is retained despite lacking the claimed positive Espin signature, justified by \"changes in slope\"—that weakens the claim of a common signature. Fourth, validation is entirely in-sample; the final recommendation (noncentrosymmetric semimetals with slightly more than half-filled d-orbitals) is a reasonable summary of known room-temperature hosts, not a falsifiable new prediction.\n\nThe SI's theoretical validation model uses fitted parameters (θD and μ), which the authors acknowledge, and the model does capture the right energy scale for low-field, low-temperature systems. That is a minor concern relative to the normalization issue.\n\nOverall: this is a useful heuristic and a worthwhile perspective, but as it stands the quantitative claim overreaches. The framework deserves a serious referee—I would send it out with the expectation of major revision. The normalization must be recomputed or clearly redefined, the Ereq extraction should be made repeatable, and an out-of-sample test, even on one or two materials, would substantially strengthen it. For a reader in the skyrmion or quantum-materials screening space, this is worth the time; just don't cite it for the quantitative correlation until the numbers are fixed.","headline":"A plausible heuristic for skyrmion-host screening with new DFT data, but the central quantitative trend rests on a normalization inconsistency and a subjective feature extraction.","tokens_in":28159,"tokens_out":1881,"would_cite":false,"duration_ms":20445,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A local minimum in the spin density of energy just above the Fermi level measures the energy cost of entering a skyrmion phase, so a routine ground-state calculation can predict the field and temperature a candidate material will need.","keywords":["Skyrmion formation","Electronic instability","Density of energy","Spin density of energy","Dzyaloshinskii-Moriya interaction","RKKY interaction","d-orbital filling","ambient-condition skyrmions"],"falsifier":"Re-plot Figure 7b with Ereq normalized as the text states (per unit-cell volume) rather than as Table S2 implements it (per volume per magnetic atom), and check whether the correlation with temperature and field survives. Alternatively, run noncollinear supercell DFT on one host—for instance MnSi, with the skyrmion spin texture imposed—and compare the energy difference from the collinear ground state with the extracted Ereq; a disagreement beyond the paper's own ~30% scatter against experimental energy models would falsify the mapping.","tokens_in":27125,"feed_emoji":"🌀","tokens_out":11164,"duration_ms":114309,"temperature":0.7,"pith_summary":"This paper tries to give skyrmion research a practical signpost: a number computable from a routine ground-state calculation that tells you how hard a material has to be pushed—how low the temperature, how high the magnetic field—before it will form a skyrmion phase. The number, called Ereq, is read from the spin-resolved density of energy, a band-structure-derived quantity that captures both interatomic and atomic energy contributions separately for spin-up and spin-down electrons. Across fourteen established skyrmion hosts spanning insulating and metallic, d-electron and f-electron, and Dzyaloshinskii–Moriya and RKKY stabilized systems, the authors find that Ereq divided by unit-cell volume orders the experimentally observed temperature and field requirements. If true, the same routine calculation becomes a screening tool: compute Ereq/Å3 for an untested compound and you get a prediction of whether its skyrmion phase is reachable near room temperature and low field. The paper's headline prediction is that noncentrosymmetric semi-metals with slightly more than half-filled d-orbitals are the most promising place to look.","feed_headline":"One computed energy predicts how hard a skyrmion phase is to reach","feed_subtitle":"A density-of-energy minimum, scaled by cell volume, ranks fourteen skyrmion hosts and flags room-temperature candidates.","key_machinery":"The central object is the spin density of energy, DOEs = DOEdown − DOEup, computed from spin-polarized band structures and integrated up to the Fermi level to give Espin, a measure of spin instability. The load-bearing feature is the local minimum in that integral just above the Fermi level, named Ereq, which the paper treats as the excitation energy required to reach the skyrmion state; normalizing Ereq by unit-cell volume is what makes different crystal structures comparable. What this quantity does is convert a ground-state, zero-field band-structure calculation into a prediction of finite-field, finite-temperature phase behavior.","core_discovery":"The central claim is that the onset conditions of a skyrmion phase are encoded in the ground-state electronic structure, specifically in the spin density of energy. In every skyrmion host examined, the integrated spin density of energy shows a destabilizing feature near the Fermi level followed by a local minimum just above it; the authors identify that minimum, Ereq, as the energy cost the system must pay to reach the excited state in which skyrmions appear. Normalized by unit-cell volume, Ereq/Å3 correlates with the temperature and field at which each material shows its strongest magneto-entropy response: the larger the value, the harsher the conditions (Figure 7b). Two independent estimat","pith_inferences":["If Ereq/Å3 is a true transition cost, the strongest test is within a single substitution series (e.g., GdRu2Si2 → GdRu2Ge2 → GdRu2Sn2 or MnSi → MnGe → FeGe), where crystal-structure noise is minimal; the paper shows these series but does not perform a quantitative fit.","The volume normalization is doing significant work, and the supporting table implements it as energy per magnetic atom rather than per formula-unit volume as the text states; whether the Figure 7b ordering survives the normalization stated in the text is a direct check a reader can make.","One could invert the correlation: set a target ambient window (roughly 300 K and ≤0.1 T), read off the corresponding Ereq/Å3, and screen crystallographic databases for compounds whose computed DOEs shows a local minimum in that range—a screening strategy the paper implies but does not run.","The descriptor measures onset conditions, not stability: a material passing the Ereq screen might still host skyrmions only in a narrow field-temperature pocket, so pairing with micromagnetic or noncollinear calculations would be needed to bound the skyrmion phase itself."],"forward_implications":["A routine spin-polarized DFT calculation, post-processed into DOEs, can rank known and candidate skyrmion hosts by the temperature and field their skyrmion phase will demand, making Ereq/Å3 a screening descriptor.","Doping that shifts the Fermi level into a region of stronger electronic instability should convert non-skyrmion hosts into skyrmion hosts, as argued for Mn2RhSn → Mn2Rh0.95Ir0.05Sn and for hole-doped B20 compounds.","The framework predicts ambient-condition skyrmions are most likely in noncentrosymmetric semi-metals with d6–d7 configurations, which should focus experimental synthesis on that class.","Because the framework is grounded in ground-state collinear DFT, it is silent on skyrmion size and phase stability windows; the paper delegates those questions to noncollinear supercell calculations and micromagnetic modeling.","For known hosts, the DOEs curve itself suggests how to tune conditions: chemical adjustments that lower Ereq/Å3 should soften the field and temperature needed for skyrmion formation."],"supporting_citations":[{"why":"Defines the crystal-orbital Hamilton population (COHP) method and the bonding-analysis tools whose energy-resolved picture the density-of-energy analysis extends.","marker":"[76-78]"},{"why":"Introduces the density-of-energy function that the paper uses to capture both interatomic and on-site band-energy contributions.","marker":"[80]"},{"why":"Documents Fermi surface nesting in metallic skyrmion hosts (EuAl4, GdRu2Si2, Gd2PdSi3), the prior electronic-instability descriptor the framework generalizes to insulating hosts.","marker":"[72-75]"},{"why":"Supplies the magneto-entropy-based skyrmion phase coordinates (TSmax, HSmax) for FeGe that define the experimental axes of the Ereq/Å3 correlation.","marker":"[59]"},{"why":"Provides the near-room-temperature DM-driven hosts (FeGe, Co8Zn9Mn3, Mn2Rh0.95Ir0.05Sn, Mn1.4Pt0.9Pd0.1Sn) that anchor the ambient-conditions prediction.","marker":"[59-63]"},{"why":"Reports the multistep meron/skyrmion transition conditions in the GdRu2X2 family used as the first case study.","marker":"[35]"},{"why":"Establishes the antibonding and electronic-instability analysis of GdRu2X2 that motivates the density-of-energy framework.","marker":"[86-88]"},{"why":"Describes the plane-wave DFT code used to compute the spin-polarized ground states from which the spin density of energy is extracted.","marker":"[82], [85]"}],"fun_headline_variants":["A single energy value maps the path to room-temperature skyrmions","Skyrmion ease: density-of-energy minimum predicts field and temperature","One metric from ground-state energy flags skyrmion hosts for ambient use","Spin density-of-energy reveals cost to reach skyrmion phase","Ereq per volume ranks skyrmion materials for easier access"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The load-bearing premise is that the local minimum in the ground-state spin density of energy just above the Fermi level (Ereq) equals the energy a material must absorb to enter the skyrmion phase; the Figure 7b correlation rests entirely on that identification and on the chosen volume normalization, yet no derivation connects the band-structure feature to the thermodynamic cost of the transition.","fun_headline_variants_meta":{"raw":{"variants":["A single energy value maps the path to room-temperature skyrmions","Skyrmion ease: density-of-energy minimum predicts field and temperature","One metric from ground-state energy flags skyrmion hosts for ambient use","Spin density-of-energy reveals cost to reach skyrmion phase","Ereq per volume ranks skyrmion materials for easier access"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000204,"raw_usage":{"total_tokens":1185,"prompt_tokens":665,"completion_tokens":520,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":409,"completion_tokens_details":{"reasoning_tokens":430}},"tokens_in":409,"tokens_out":520,"duration_ms":5031,"temperature":1.0,"reasoning_tokens":430,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T20:47:55.551133+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-plot Figure 7b with Ereq normalized as the text states (per unit-cell volume) rather than as Table S2 implements it (per volume per magnetic atom), and check whether the correlation with temperature and field survives. Alternatively, run noncollinear supercell DFT on one host—for instance MnSi, with the skyrmion spin texture imposed—and compare the energy difference from the collinear ground state with the extracted Ereq; a disagreement beyond the paper's own ~30% scatter against experimental energy models would falsify the mapping.","supporting_citations":[],"review_version":1}