{"id":"c9d3510c-60f9-4ea7-94de-ba5a33b8756e","arxiv_id":"2608.07666","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"Vacancy defects in bilayer hBN link sliding ferroelectricity to localized magnetism, and an out-of-plane electric field can tune the amplitude of the resulting antiferromagnetic order.","lead":"This paper uses computer simulations to show that atomic vacancies in bilayer boron nitride can couple the material's sliding electric polarization to localized magnetic moments, creating a magnetoelectric response in a normally nonmagnetic material. The result suggests a route to electric-field-controlled magnetic memory and logic elements in simple stacked two-dimensional layers.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 10.9% registry contrast in the Néel amplitude rests on a half-cell spin-density partition; the main text calls it a defect-localized moment, but no atomic-basin or partition-independent check is provided.","rationale":"The reader's CONDITIONAL verdict already identifies the p_eff degeneracy and the absence of convergence tests, and I agree those are real issues. My stress-test pass sharpens this into a more specific, load-bearing concern: the headline 10.9% number is defined by a half-cell layer projection, while the text calls it a change in the defect-localized moment. Because the field-linear coupling Q_s E_z L_z^2 is symmetry-allowed, the symmetry framework cannot by itself tell us whether the computed L_z change is a true change in the local magnetic order or a redistribution of spin density between layers. A simple atomic-basin projection would settle the interpretation. This does not change the overall verdict: the paper is a plausible computational prediction with an explicit symmetry model, but its central quantitative claim should be verified with a partition-independent measure and, more broadly, with structural relaxation and charge-state stability checks. I therefore keep the reader's CONDITIONAL stance rather than moving to ACCEPT or REJECT.","tokens_in":16160,"tokens_out":9669,"duration_ms":110426,"concrete_test":"Repeat the AB and BA V_N-V_N calculations at E_z = 0, +/-0.05, and +/-0.10 V/A, and evaluate the spin moment with two independent partitions: (i) integrate the spin density only within atomic Voronoi or Bader cells centered on the three B atoms coordinated by each V_N; (ii) shift the planar dividing coordinate z0 by +/-0.5 A and recompute L_z. If the AB-BA contrast at 0.10 V/A drops below about 3% under method (i), or changes substantially with z0, then the reported 10.9% 'defect-localized moment' contrast is a projection artifact rather than a robust magnetic amplitude modulation. If the contrast persists in Bader-cell moments and is insensitive to z0, this specific concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central magnetoelectric result, a registry-dependent contrast of about 10.9% in |L_z| at |E_z| = 0.10 V/A, is obtained from L_z = (M_top,z - M_bottom,z)/2, where the layer moments are integrals of the out-of-plane spin density over half of the supercell divided at the interlayer midpoint (Eqs. S10-S12). The main text interprets this as a change in the 'defect-localized moment,' but a half-cell integral captures all spin density in each half of the cell, including delocalized tails and any field-driven spatial redistribution between the layers, rather than the spin moment on the three B atoms hosting each V_N state. The Q_s E_z L_z^2 coupling is symmetry-allowed and the fit to Table S4 is internally consistent, but the physically meaningful statement that the vacancy-localized spin amplitude is electrically modulated is only as strong as the partition used to define L_z. No convergence with respect to the dividing plane or an alternative projection (atomic spheres, Bader basins, or Wannier functions) is reported. Moreover, the total-energy fit only determines the combined coefficient p_eff = A P_pair + gamma L_0^2 (Eq. S19), so the energy data alone cannot separate the direct polar contribution from the magnetic renormalization; the claimed magnetoelectric mechanism relies on the L_z projection values. If the 10.9% contrast is an artifact of the half-cell partition, the central claim of electric-field amplitude control of a compensated Néel order is not established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a general mechanism that couples vacancy-localized magnetic moments to sliding ferroelectricity in nonmagnetic van der Waals bilayers, using bilayer hBN as a prototype. The authors perform first-principles DFT calculations for single nitrogen and boron vacancies and for interlayer vacancy pairs. They show that a single vacancy lifts the AB/BA registry degeneracy and creates a registry-even polarization offset, and that interlayer vacancy pairs exhibit sublattice-dependent exchange: V'_B-V'_B is nearly magnetically decoupled, V'_B-V_N forms a ferrimagnetic state, and V_N-V_N forms a compensated antiferromagnetic state. The central result is that for the V_N-V_N pair, an out-of-plane electric field selects the polar registry and changes the layer-resolved Néel order magnitude by about 10.9% at |E_z| = 0.10 V/Å. This behavior is rationalized with a Landau free-energy expansion in which Q_s E_z L_z^2 is the lowest-order invariant coupling the sliding coordinate, the field, and the squared Néel amplitude. The paper also reports Berry-phase polarizations and fully relativistic noncollinear calculations to support the symmetry arguments.","tokens_in":16493,"tokens_out":6097,"duration_ms":57326,"significance":"If the reported amplitude modulation is robust, this work provides a conceptually new route to magnetoelectric functionality in nominally nonmagnetic 2D systems, with potential for spintronic readout of sliding ferroelectricity. The symmetry-based Landau framework is elegant and well matched to the DFT results, and the authors are unusually transparent about the limitations of their model: they explicitly disclose the underdetermination of p_eff, the Fermi-level dependence of the vacancy charge states, and the rigid-sliding approximation. The paper also reports internal consistency checks, including time-reversal degeneracy of the Néel states and polarization invariance under L_z reversal. However, the quantitative centerpiece of the work, the 10.9% contrast in |L_z|, depends on a particular half-cell partition of the spin density and on calculations for which no convergence tests are shown. The significance is therefore conditional on additional validation of this order parameter and on the numerical precision of the small energy scales involved.","major_comments":[{"comment":"The central quantitative result of a ~10.9% registry-dependent contrast in |L_z| is obtained by partitioning the out-of-plane spin density at the interlayer midpoint (SM Eqs. S10-S12). This half-cell integral captures all magnetization in each half of the cell, including delocalized tails and field-induced charge redistribution, rather than isolating the vacancy-localized moment on the three B atoms neighboring each V_N. The main text refers to this quantity as the 'defect-localized moment' and interprets the field-induced change as amplitude modulation of the Néel order. Because the Q_s E_z L_z^2 symmetry invariant is compatible with any continuous registry-dependent spin-density redistribution, the physical interpretation is only as strong as the partition. Please provide a partition-independence test (e.g., varying the dividing plane, atomic-basin/Bader projection, or Wannier-based local moments) or revise the language to describe the change in the layer-projected magnetization density rather than the defect-localized moment.","section":"SM, Eqs. (S10)-(S12); main text Fig. 3(e)"},{"comment":"The total-energy fit determines only the combined coefficient p_eff = A P_pair + gamma L_0^2, not the magnetic coefficient gamma separately. Thus the registry preference under an electric field can be explained entirely by the direct polar term -A P_pair Q_s E_z, with no magnetic contribution. The claim that the electric field modulates the amplitude of the Néel order relies on the L_z values from the half-cell partition (Table S4) and on the fit of Eq. (S22). I ask the authors to either compute the pair polarization P_pair independently (e.g., from Berry-phase polarization of the pair system as a function of E_z) or perform constrained-L_z calculations that separate the two contributions. The manuscript already discloses this degeneracy, but the magnetoelectric attribution is a central claim and should be supported by data that do not presuppose the partition.","section":"SM Eq. (S19) and Eq. (S15); main text Eq. (3)"},{"comment":"No convergence tests with respect to plane-wave cutoff, k-point grid, or supercell size are reported. The decisive quantities are small: the AB-BA energy splittings for single vacancies are about 1-4 meV, the pair magnetic splittings range down to 0.07 meV, and the magnetic anisotropy energies in Table S2 are 0.1-4.9 µeV, close to total-energy noise. The 10.9% contrast and the linear slopes in Eq. (S22) are quantitative claims that require at least a subset of convergence checks (e.g., a 7x7x1 grid or an 8x8 supercell for the V_N-V_N pair at one or two fields). Without such checks, the numerical precision of the central result is not established.","section":"SM, Computational Methods; Tables S2 and S4"},{"comment":"The adopted Fermi-level window E_F - E_VBM ≈ 3.6 eV is motivated by N-rich, impurity-containing hBN. In clean or differently doped samples, the stable charge states in Table S1 change, so the neutral V_N species and the negatively charged V'_B species are not simultaneously present, and the specific vacancy pairs discussed here would not form. The authors disclose this assumption but the abstract and conclusion state the mechanism as 'robust' and 'general'. Please state more explicitly in the main text that the demonstrated AFM V_N-V_N pair requires this specific Fermi-level window, and discuss how the proposed mechanism would be modified if adjacent charge states become stable under gating.","section":"SM, 'Selection of Vacancy Charge States', Table S1"}],"minor_comments":[{"comment":"The figure labeled 'Fig. 4' appears after the references and is described only in an unnumbered paragraph; it should be placed in the main text or cited with a proper figure environment.","section":"Main text, End-matter Fig. 4"},{"comment":"The notation for τ_def is introduced only implicitly; please state explicitly that τ_def = +1, 0, -1 for V'_B, pristine, and V_N^×, respectively, in the same sentence as the definition of τ_def.","section":"Main text after Eq. (2)"},{"comment":"There is a formatting error: '2X i=1' should read the summation over i = 1 to 2; as printed it is not readable.","section":"SM, Eq. (S14)"},{"comment":"The phrase 'nonmagnetic van der Waals bilayers' is potentially confusing because the bilayers are nonmagnetic only in their pristine form; consider using 'nonmagnetic host' or 'nominally nonmagnetic'.","section":"Abstract and Introduction"},{"comment":"The superscript symbols for the vacancy charge states (e.g., V_N^•, V_N^', V'_B) are not defined in text; a one-sentence explanation of the charge-state notation would improve readability.","section":"SM, Table S1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is well-organized and the symmetry analysis is convincing. The principal concern is whether the 10.9% amplitude-modulation result is robust with respect to the spin-density partition; this is testable and fixable with additional calculations, so I do not recommend rejection at this stage. The charge-state and convergence caveats should be addressed in the revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a serious computational prediction with a clean symmetry story, and I think it deserves referee time. But the headline 10.9% effect is softer than the abstract suggests: it comes from a half-cell spin-density partition, and the total-energy fit only constrains a combination of direct-polar and magnetic coefficients. The authors know this and say so, which helps.\n\nWhat's new: a single vacancy breaks the AB-BA degeneracy in bilayer hBN, and a pair of nitrogen vacancies gives a compensated Néel order whose amplitude can be modulated by an out-of-plane field through the registry. The sublattice-selective exchange (BB weak, BN ferrimagnetic, NN antiferromagnetic) is a concrete pattern that could be tested in other vdW bilayers. The symmetry analysis rooted in the AA reference structure is the paper's own, and the Landau expansion is used consistently with it.\n\nWhat's good: raw DFT numbers are in tables, the charged-supercell treatment is at least careful, and the authors explicitly acknowledge that p_eff cannot be separated into polar and magnetic shares from energy data, and that the AFM state is a broken-symmetry solution. That disclosure makes me trust the numbers.\n\nSoft spots: first, the 10.9% registry contrast is defined via L_z obtained from a planar cut at the interlayer midpoint. That is a legitimate definition of a layer moment, but it is not a defect-localized moment, and no check is given for the cut position or an alternate projection (Bader, atomic spheres, Wannier). The spin density is likely well localized on the three boron neighbors, so I suspect the effect is real, but it is a missing sanity check. Second, no convergence tests (k-points, cutoff, supercell size) are shown for meV-scale splittings, and the sliding paths are rigid with fixed intralayer coordinates. Local relaxation around a vacancy could alter the registry coupling. Third, the simultaneous stability of V_N and V'_B rests on E_F - E_VBM = 3.6 eV from the literature; that is a physically motivated assumption but still an assumption.\n\nNet: this is a solid, internally consistent prediction, and the Q_s E_z L_z^2 symmetry argument is correct. The main risk is that the quantitative 10.9% effect is partition-dependent. I would send it to a referee, with the expectation that they ask for a spin-density partition check and convergence tests before publication.","headline":"A clean symmetry story and internally consistent DFT data that deserve referee time, but the 10.9% magnetic contrast needs a spin-density partition check before I'd bet on it.","tokens_in":17072,"tokens_out":4181,"would_cite":true,"duration_ms":38406,"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":"A vacancy pair turns nonmagnetic bilayer hBN into an electrically switchable antiferromagnet.","keywords":["hexagonal boron nitride","sliding ferroelectricity","vacancy defects","localized magnetism","Néel order","electric-field control","van der Waals bilayers","stacking registry"],"falsifier":"Prepare bilayer hBN with one nitrogen vacancy per layer, apply out-of-plane electric fields of $+0.10$ and $-0.10\\ \\mathrm{V/\\AA}$, and measure the layer-resolved magnetization: if the magnitude of the Néel order does not change by roughly 10.9% between the stacking registries selected by the two field directions, the central claim of electric-field amplitude control fails. A separate check is whether the Fermi level can actually be pinned in the $3.48$–$4.59\\ \\mathrm{eV}$ window in which both required vacancy charge states are stable.","tokens_in":15892,"feed_emoji":"🧲","tokens_out":9507,"duration_ms":80988,"temperature":0.7,"pith_summary":"The paper aims to establish that the sliding registry of a nonmagnetic van der Waals bilayer can serve as a local control parameter for magnetism once vacancy defects are introduced. In bilayer hexagonal boron nitride, a nitrogen vacancy carries a spin-1/2 moment and a boron vacancy a spin-1 moment; a single vacancy lifts the degeneracy between the two polar sliding states (AB and BA), and in interlayer pairs the hosting sublattice dictates whether the coupling is ferrimagnetic or antiferromagnetic. For the antiferromagnetic nitrogen-vacancy pair, an out-of-plane electric field selects the preferred registry and changes the magnitude of the layer-resolved Néel order by about 10.9% at $|E_z| = 0.10\\ \\mathrm{V/\\AA}$. The reason to care is that this is a purely electrostatic route to magnetoelectric control in a material with no magnetic atoms, avoiding the usual scarcity of single-phase two-dimensional multiferroics.","feed_headline":"Defects give nonmagnetic hBN bilayers electric control of magnetism","feed_subtitle":"Sliding registry shifts the Néel amplitude by about 11 percent under an out-of-plane field.","key_machinery":"The load-bearing object is a symmetry-adapted free energy for the bilayer built from the sliding coordinate $Q_s$ (BA = +1, AB = -1), the defect-layer variable $\\eta_{\\mathrm{def}}$, the defect-species variable $\\tau_{\\mathrm{def}}$ ($V'_B$, pristine, $V_N^\\times$), and the layer-resolved Néel order $L_z=(M_{\\mathrm{top},z}-M_{\\mathrm{bottom},z})/2$. Symmetry of the AA reference structure admits $Q_s E_z L_z^2$ as the lowest-order field-linear invariant that distinguishes AB from BA while leaving the two time-reversed Néel states degenerate; this term is what converts an out-of-plane electric field into a change in $|L_z|$ rather than a spin reorientation. The single-defect couplings $Q_s\\eta_{\\mathrm{def}}\\tau_{\\mathrm{def}}$ and $Q_s\\tau_{\\mathrm{def}}^2$ encode the vacancy-selected registry preference and the registry-even polarization offset, and all coefficients are fitted to first-principles total energies, Berry-phase polarizations, and field-dependent layer-resolved moments.","core_discovery":"On its own terms, the paper establishes that atomic vacancies in bilayer hBN are registry sensors, not passive impurities: a single neutral nitrogen vacancy $V_N^\\times$ or negatively charged boron vacancy $V'_B$ breaks the AB–BA equivalence of the sliding-ferroelectric bilayer and produces a registry-even polarization offset that reverses when the vacancy moves to the opposite layer. In interlayer pairs, the hosting sublattice fully dictates the exchange: $V'_B$–$V'_B$ moments remain nearly decoupled, $V'_B$–$V_N^\\times$ couples antiparallel into a ferrimagnet, and $V_N^\\times$–$V_N^\\times$ forms a compensated antiferromagnet stabilized by kinetic exchange. For the $V_N^\\times$–$V_N^\\times$ pair, a finite out-of-plane electric field lifts the AB–BA degeneracy approximately linearly, and the field-favored registry exhibits the larger defect-localized moment, with a registry-dependent contrast of about 10.9% in the layer-resolved Néel order $L_z=(M_{\\mathrm{top},z}-M_{\\mathrm{bottom},z})/2$ at $|E_z|=0.10\\ \\mathrm{V/\\AA}$. The field modulates the amplitude of the Néel vector, not its orientation, and the time-reversed states remain degenerate, so the result is amplitude control of a compensated antiferromagnetic order parameter without an external magnetic field or spin-orbit coupling.","pith_inferences":["If the mechanism generalizes as claimed, a moiré superlattice formed by twisting two defective hBN layers should show a spatial map of registry-dependent Néel amplitude, making the stacking texture electrically readable at the nanoscale.","Because the $V_N^\\times$–$V_N^\\times$ exchange is kinetic-exchange-like, interlayer pressure or twist angle should tune both the antiferromagnetic stabilization energy and the field response; this is a testable prediction beyond the paper's explicit calculations.","The invariant $Q_s E_z L_z^2$ is fixed by symmetry and does not depend on the specific defect chemistry, so the same amplitude modulation should appear in any polar sliding bilayer with localized vacancy moments, provided the two layers carry identical defects."],"forward_implications":["A single vacancy can act as a local, nondestructive readout of the stacking registry, since its formation energy and polarization offset distinguish AB from BA.","An out-of-plane electric field gives nonvolatile, purely electrostatic control over a compensated antiferromagnetic order parameter in a nonmagnetic parent material.","The hosting sublattice of the vacancy, not just the presence of a vacancy, determines the magnetic ground state of an interlayer pair, providing a selection rule for ferrimagnetic versus antiferromagnetic coupling.","The same registry–defect coupling should occur in other polar sliding bilayers, moiré domains, and transition-metal dichalcogenide heterostructures that support localized moments."],"supporting_citations":[{"why":"Demonstrates stacking-engineered ferroelectricity in bilayer hBN, establishing the host registry-switchable polarization.","marker":"[15]"},{"why":"Provides vacancy charge-state transition levels and the defect manifold used to select $V_N^\\times$ and $V'_B$.","marker":"[19]"},{"why":"Reports interfacial ferroelectricity from van der Waals sliding, supporting registry-dependent polarization.","marker":"[28]"},{"why":"Reviews sliding ferroelectricity in two-dimensional materials and its coupling to stacking order.","marker":"[29]"},{"why":"Supplies thermodynamic calculations of carbon point defects in hBN used to justify the Fermi-level window.","marker":"[32]"},{"why":"Identifies defect complexes that pin the Fermi level near 3.6 eV in N-rich hBN, stabilizing both vacancy charge states.","marker":"[33]"},{"why":"Provides ab initio theory of the negatively charged boron vacancy, supplying the $S=1$ defect manifold and spin state.","marker":"[34]"},{"why":"Computes vertical polarization of bilayer binary compounds, giving the AB/BA polarization reference for the pristine host.","marker":"[35]"},{"why":"Supplies the Berry-phase method used to compute out-of-plane polarizations.","marker":"[44]"}],"fun_headline_variants":["Vacancies turn nonmagnetic hBN bilayers into field-tunable magnets","Electric field tunes Neel order amplitude in vacancy-doped hBN bilayers","Sublattice choice sets magnetic coupling in defective hBN bilayers","Defect pairs dictate ferrimagnetic vs antiferromagnetic order in hBN","Stacking defects enable electric-field control of magnetism in hBN"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The scenario depends on the Fermi level in real hBN being pinned near $3.6\\ \\mathrm{eV}$ above the valence-band maximum so that neutral nitrogen vacancies and negatively charged boron vacancies are stable in the same sample, and on the layers sliding rigidly without relaxing around the vacancies.","fun_headline_variants_meta":{"raw":{"variants":["Vacancies turn nonmagnetic hBN bilayers into field-tunable magnets","Electric field tunes Neel order amplitude in vacancy-doped hBN bilayers","Sublattice choice sets magnetic coupling in defective hBN bilayers","Defect pairs dictate ferrimagnetic vs antiferromagnetic order in hBN","Stacking defects enable electric-field control of magnetism in hBN"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000843,"raw_usage":{"total_tokens":3706,"prompt_tokens":1014,"completion_tokens":2692,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":630,"completion_tokens_details":{"reasoning_tokens":2596}},"tokens_in":630,"tokens_out":2692,"duration_ms":21175,"temperature":1.0,"reasoning_tokens":2596,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T00:25:01.942253+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Prepare bilayer hBN with one nitrogen vacancy per layer, apply out-of-plane electric fields of $+0.10$ and $-0.10\\ \\mathrm{V/\\AA}$, and measure the layer-resolved magnetization: if the magnitude of the Néel order does not change by roughly 10.9% between the stacking registries selected by the two field directions, the central claim of electric-field amplitude control fails. A separate check is whether the Fermi level can actually be pinned in the $3.48$–$4.59\\ \\mathrm{eV}$ window in which both required vacancy charge states are stable.","supporting_citations":[{"cited_title":"Vizner Stern, Y","cited_arxiv_id":null,"evidence_quote":"Reports interfacial ferroelectricity from van der Waals sliding, supporting registry-dependent polarization."},{"cited_title":"Wu and J","cited_arxiv_id":null,"evidence_quote":"Reviews sliding ferroelectricity in two-dimensional materials and its coupling to stacking order."},{"cited_title":"Maciaszek and B","cited_arxiv_id":null,"evidence_quote":"Identifies defect complexes that pin the Fermi level near 3.6 eV in N-rich hBN, stabilizing both vacancy charge states."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Berry-phase method used to compute out-of-plane polarizations."}],"review_version":1}