{"id":"53694273-10ce-4430-8af1-ad9022e01892","arxiv_id":"2505.09084","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Using density functional theory and molecular dynamics, the authors show that in bilayer h-BN an out-of-plane field drives in-plane interlayer sliding through off-diagonal Born effective charges, so switching is avalanche-like and the coercive field is controlled by symmetry-breaking…","lead":"A computer study of bilayer boron nitride shows that when one layer slides past the other, a vertical electric field can push the layers sideways, and once sliding starts it becomes easier, like an avalanche. The work explains why this class of ferroelectric materials can switch at low voltages and high speeds, which matters for future memory devices.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Static force-balance coercive-field curve is not a demonstrated switching dynamics; the field-dependent wrinkle is not treated self-consistently, so the avalanche-like claim needs a direct finite-field dynamical test.","rationale":"The reader's weakest assumption—that the coercive-field and avalanche picture rests on a quasistatic force-balance estimate at frozen intermediate configurations with an arbitrary small initial sliding—identifies the same load-bearing concern. My reading confirms that the static E^r_z = f^r/Z*_xz construction is the foundation of the central claim, and it is not a substitute for a dynamical simulation of the switching trajectory. The paper's own admission that the domain-wall wriggling 'warrants further investigation' is a self-identified limitation that supports this concern. The AIMD result at 250 K is real evidence for the qualitative mechanism, but it does not establish the monotonic avalanche curve or the quantitative coercive-field reduction, especially because the wrinkle contribution is not self-consistent with the field that creates it. I would keep the verdict conditional: the mechanism is plausible and worth publishing with the stated caveats, but the quantitative claims should be verified by a finite-field dynamical check before being treated as established. This does not move the reader's verdict, so I recommend UNCHANGED.","tokens_in":8697,"tokens_out":7063,"duration_ms":78388,"concrete_test":"Run finite-field Born-Oppenheimer MD (or finite-field NEB) for the single-domain h-BN bilayer: start from the AB stack with a small random in-plane displacement (e.g., 0.01–0.05 Å), apply a constant out-of-plane field, and record s(t), P_z(t), and the field-relaxed wrinkle amplitude. Determine (i) the field at which P_z actually reverses and the time profile—does reversal accelerate once started?—and (ii) the self-consistent wrinkle amplitude at the reversal field. If the reversal field and the rate profile deviate substantially from the E^r_z(s) curve in Fig. 1(f), or if the self-consistent \\bar h at E_z = 0.61 V/Å is not about 0.01 Å, the avalanche and wrinkle-reduction claims need revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central quantitative claim—that E^r_z decreases monotonically with sliding and switching is avalanche-like—is based on the static force-balance estimate E^r_z = f^r/Z*_xz, evaluated at DFT-relaxed but zero-field frozen geometries along Path II, with an arbitrarily chosen initial sliding s = 0.045 Å to break the C3 symmetry. This estimate assumes that the field enters only through this balance and does not change the geometry or the Z* tensor. The paper partially relaxes this by adding a field-induced wrinkle, but not self-consistently: Fig. 2(c) shows the wrinkle amplitude \\bar h grows linearly with E_z, yet the E^r_z reduction is computed by inserting the \\bar h = 0.01 Å structure into the zero-field formula. If the E_z needed to produce \\bar h = 0.01 Å is not 0.61 V/Å, the claimed reduction to about one quarter is not a fixed point of the actual field-dependent problem. More generally, a monotonic E^r_z along a prescribed reaction coordinate does not by itself prove avalanche-like dynamics: inertia, damping, thermal nucleation, and field-induced barrier reshaping can all change the trajectory. The paper's own AIMD at 250 K provides qualitative support, but no time-dependent simulation of the single-domain avalanche or of the domain-wall 'wriggling' is reported; the latter is explicitly left as 'warranting further investigation' in the final section and in the Note. The mechanism is plausible and the static Z* analysis is informative, but the strongest form of the claim—that the coercive field is controlled by symmetry-breaking perturbations and that switching is avalanche-like—is not yet demonstrated dynamically.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This Letter uses DFPT and AIMD to study polarization switching in bilayer h-BN as a model sliding ferroelectric. The authors compute off-diagonal Born effective charges Z*_xz along the AB→SP→BA sliding path (Path II) and show that an out-of-plane electric field exerts an in-plane force proportional to Z*_xz. They find that the quasistatic balance field E^r_z = f^r/Z*_xz decreases monotonically with sliding, in contrast to conventional ferroelectrics such as BaTiO3 where it peaks mid-path. They further analyze how a field-induced layer wrinkle, the interlayer gap, temperature, and the presence of domain walls modify Z*_xz, f^r, and hence E^r_z, concluding that any perturbation breaking the in-plane C3 symmetry assists an avalanche-like switching process. The paper reports AIMD showing polarization switching at 250 K under E_z = 0.2 V/Å and static estimates predicting a much larger zero-temperature coercive field. Domain-wall calculations show enhanced off-diagonal Z* near the wall center, which the authors propose leads to a wriggling motion of domain walls.","tokens_in":8971,"tokens_out":6839,"duration_ms":62915,"significance":"If the avalanche-like switching picture is correct, it would explain why experimental coercive fields in sliding ferroelectrics are far below the naive barrier-derived estimates and why switching can be ultrafast. The identification of off-diagonal Born effective charges as the coupling mechanism is a valuable step and is supported by independent DFPT calculations. The paper's strengths include a clean benchmark against BaTiO3, explicit consideration of wrinkle and gap dependences, AIMD support for room-temperature switching at low fields, and an open disclosure of two overlapping preprints. However, the dynamical conclusions (avalanche and domain-wall wriggling) are inferred from static calculations and require either additional simulation or more careful wording.","major_comments":[{"comment":"The coercive field curve E^r_z = f^r/Z*_xz is a quasistatic force balance evaluated at zero-field relaxed geometries along Path II, starting from an arbitrarily chosen initial sliding s = 0.045 Å. This calculation demonstrates that the static balance field decreases with sliding, but it does not by itself demonstrate avalanche-like switching dynamics, which requires a time-dependent or finite-field instability. The paper contains no simulation of the switching trajectory under a sustained field; the AIMD in Fig. 3(b) shows switching at 250 K with E_z = 0.2 V/Å but is not analyzed in terms of the force-balance curve. Please either add a self-consistent finite-field calculation or a direct time-dependent simulation (e.g., AIMD with the field on from the start) to support the avalanche claim, or revise the wording throughout (abstract, Fig. 1(f)) to 'the static critical field decreases monotonically with sliding' rather than 'avalanche-like switching dynamics.'","section":"Eq. (2) and Fig. 1(f)"},{"comment":"The wrinkle reduction of E^r_z is not self-consistent. The wrinkle amplitude \\bar h is the equilibrium response to E_z (Fig. 2c, linear growth), but the reported E^r_z = 0.61 V/Å at \\bar h = 0.01 Å is obtained by inserting the \\bar h = 0.01 Å structure into the zero-field formula E^r_z = f^r/Z*_xz with s = 0.045 Å fixed. The paper does not verify that E_z = 0.61 V/Å actually produces \\bar h = 0.01 Å in the staggered geometry, nor that the finite field leaves Z*_xz and f^r unchanged. If the field needed for \\bar h = 0.01 Å differs from 0.61 V/Å, the claimed reduction to about one quarter is not a fixed point of the coupled problem. Please perform a self-consistent calculation (relax under a fixed E_z and recompute Z* and f^r at that geometry) or explicitly state the E_z value corresponding to \\bar h = 0.01 Å and check the consistency.","section":"Fig. 2(c)-(d)"},{"comment":"The 'wriggling motion' of domain walls is inferred from the static spatial profile of Z*_xz and Z*_yz in a relaxed domain-wall supercell. No dynamical simulation of the domain wall under E_z is reported, and the text itself states this 'warranting further investigation.' As written, the abstract and concluding sentence present the wriggling as a result ('results in a wriggling motion of domain walls'), which overstates the evidence. Please either add a minimal time-dependent simulation (e.g., MD of the DW under a field) or label the wriggling explicitly as a qualitative prediction, and adjust the abstract accordingly.","section":"Paragraph beginning 'Finally, it is worth to investigate...' and Fig. 4"}],"minor_comments":[{"comment":"The Supplemental Material reference [28] contains a corrupted citation string 'including Refs. [13, 24? ? ? ? ? ? ? ? ?]'; this must be repaired before publication.","section":"Reference [28]"},{"comment":"The choice of initial sliding s = 0.045 Å is arbitrary; please justify it from the thermal fluctuation amplitude at 250 K or provide a short sensitivity study over s to show that the monotonic decrease of E^r_z is robust.","section":"Fig. 1(f) and text near 'Beginning from a staggered stacking mode'"},{"comment":"Fig. 3(c) reports a temperature-dependent coercive field estimated from AIMD, but the main text does not describe the AIMD protocol (e.g., how E_z is ramped, what polarization threshold defines switching). Please specify the protocol in the main text or refer the reader to a detailed SM section.","section":"Fig. 3(c)"},{"comment":"The quantity \\bar h is used in Fig. 2(d) but defined only in the caption; please define it in the body text before the discussion of the wrinkle effect.","section":"Fig. 2(d) and text"}],"recommendation":"major_revision","confidential_remarks":"The authors openly acknowledge two overlapping preprints (refs [36,37]) that also address off-diagonal Born effective charges and domain-wall motion in sliding ferroelectrics. Given this overlap, the editor may wish to verify that the present manuscript's claims are sufficiently distinct and that the novelty of the avalanche and wriggling mechanisms is clearly positioned relative to those works. The main technical weakness is that the central dynamical claims are based on static force-balance arguments rather than time-dependent simulations, which should be addressed in revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my read.\n\nThe paper deserves a serious referee. The central mechanism—nonzero off-diagonal Born effective charges couple an out-of-plane field to in-plane sliding, and C3 symmetry at the AB/BA stacks forces Z*_xz to vanish there, so switching requires a symmetry-breaking perturbation and then proceeds avalanche-like—is clearly argued and supported by the DFPT data. The comparison with BaTiO3 is nicely done: it makes the distinction between climbing-like and avalanche-like coercive behavior concrete. The AIMD result that 0.2 V/Å switches the polarization at 250 K but not at 10 K gives real qualitative support for the thermal-fluctuation picture, and the wrinkle/gap trends are a sensible first look. The authors also honestly acknowledge the two concurrent preprints, and their added value—the symmetry argument, the finite-temperature trend, and the domain-wall local Z* analysis—is real.\n\nNow the soft spots, which are mostly about the quantitative claims. The coercive-field numbers come from f^r/Z*_xz evaluated on zero-field relaxed geometries with an arbitrarily chosen initial sliding s=0.045 Å. That is a static force balance, not a switching threshold. The monotonic decrease of E^r_z along the path is necessary for avalanche-like switching but not sufficient: inertia, damping, thermal nucleation, and field-induced changes to the path can all alter the outcome. The wrinkle reduction to about one quarter is also not self-consistent—the Z* and f^r are computed at a fixed \\bar h=0.01 Å structure, but the field that actually produces that wrinkle is not the same field at which the balance is evaluated. That is a real fixed-point gap, and the stress-test note has it right. The domain-wall 'wriggling' is plausible and the static local Z* profiles support it qualitatively, but no time-dependent domain-wall velocity is computed; the authors themselves say it warrants further investigation. So the physics is likely right, but the paper's strongest quantitative form—E_c values and avalanche claim—is inferred, not demonstrated dynamically.\n\nCitation pattern is clean: the comparison with experiment is honest (their static estimate is two orders above the measured coercive field, which is exactly the puzzle), and they don't fit parameters to experiment. Self-citations are to their own prior method papers, which is appropriate.\n\nWho benefits: people working on sliding ferroelectrics, 2D memory, and domain-wall dynamics. I'd send it to referee with a request to separate demonstrated static coupling from inferred dynamics, ideally with a direct finite-field MD or at least a self-consistent wrinkle treatment.","headline":"A solid ab initio case that off-diagonal Born effective charges and C3 symmetry make h-BN sliding switching avalanche-like, with quantitative coercive fields that are static estimates rather than dynamical predictions.","tokens_in":9581,"tokens_out":2530,"would_cite":true,"duration_ms":25163,"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":"Sliding ferroelectricity in bilayer h-BN switches by avalanche rather than by climbing a fixed intrinsic energy barrier, because off-diagonal Born effective charges couple a vertical electric field to horizontal ionic motion.","keywords":["sliding ferroelectricity","hexagonal boron nitride bilayer","Born effective charge","coercive field","avalanche switching","domain wall","density functional theory","ab initio molecular dynamics"],"falsifier":"Measure polarization reversal in an ultraclean, strain-free, atomically flat h-BN bilayer at low temperature under a perfectly homogeneous vertical electric field; the paper predicts zero or near-zero probability of switching unless a symmetry-breaking perturbation nucleates sliding, so observing low-field uniform switching would falsify the avalanche picture.","tokens_in":8438,"feed_emoji":"⚡","tokens_out":6455,"duration_ms":61331,"temperature":0.7,"pith_summary":"This paper seeks to explain how an out-of-plane electric field can switch the in-plane interlayer sliding that creates ferroelectric polarization in bilayer hexagonal boron nitride. Its central claim is that the switching is driven by off-diagonal Born effective charges—tensors that map a vertical field into horizontal forces on ions—so the required electric field is not an intrinsic constant but drops monotonically as sliding proceeds, giving avalanche-like reversal. The paper argues that any perturbation breaking the in-plane C3 symmetry, including thermal fluctuation, wrinkling, interlayer gap changes, or domain walls, sets the starting point and therefore controls the effective coercive field. If this holds, sliding ferroelectrics differ fundamentally from conventional ferroelectrics like BaTiO3, and their low switching fields are a consequence of symmetry, not just low energy barriers.","feed_headline":"Off-diagonal charge turns h-BN switching into an avalanche","feed_subtitle":"The coercive field is set by symmetry-breaking perturbations, not by a fixed intrinsic maximum.","key_machinery":"The central object is the Born effective charge tensor $Z^*_{\\kappa,ij}$, defined as the change of polarization of ion $\\kappa$ along direction $i$ due to a unit displacement along $j$ (equivalently the force on the ion from an electric field). In h-BN, its off-diagonal element $Z^*_{xz}$ vanishes in the AB/BA ferroelectric states by C3 symmetry, becomes finite and peaks sharply at the P=0 saddle-point state, and its layer sum governs the perpendicular field-to-sliding coupling. The paper uses this tensor together with the intrinsic resistance force $f^r$ from the energy barrier to estimate the required critical field $E^r_z = f^r/Z^*_{xz}$ along the switching path. That ratio is the mechanism that converts a symmetric barrier into an avalanche: the denominator grows at intermediate sliding, so the field needed to continue sliding falls even as the resistance force rises.","core_discovery":"For bilayer h-BN, the paper's key finding is that the off-diagonal Born effective charge tensor element $Z^*_{xz}$, which is exactly zero in the high-symmetry AB/BA ferroelectric states due to in-plane C3 symmetry, becomes nonzero as soon as the layers slide. Because the electrostatic force on an ion is $f_i = E_j Z^*_{ij}$, this nonzero $Z^*_{xz}$ provides the perpendicular coupling that lets an out-of-plane field drive in-plane motion. The resistance force from the energy barrier rises then falls, but the required critical field $E^r_z = f^r/Z^*_{xz}$ monotonically decreases with sliding, starting from an arbitrarily high value at the symmetric state and reaching a minimum near P=0, then rising as $Z^*_{xz}$ vanishes again. This avalanche-like behavior means the measured coercive field is set by the initial symmetry-breaking perturbation, not by an intrinsic maximum. The authors also show that the large $Z^*_{xz}$ near the P=0 intermediate state makes the center of a domain wall move faster than its flanks, producing a wriggling motion that may explain the high switching speeds reported experimentally.","pith_inferences":["If the avalanche picture is right, measurements of switching statistics on nominally identical h-BN devices should show a distribution of coercive fields, with the tail controlled by the largest thermal fluctuation or local wrinkle; this is a testable corollary the paper does not pursue.","The C3 symmetry-selection rule suggests that a perfectly homogeneous vertical field cannot switch a defect-free, unstrained, zero-temperature h-BN bilayer at any field strength, so the paper's switching mechanism implicitly relies on a fluctuating or inhomogeneous environment.","One could extend the calculation to the dynamics of a single domain wall under an electric-field pulse to obtain a predicted wall velocity, connecting the quasistatic estimates to the experimentally measured switching times."],"forward_implications":["The coercive field of sliding ferroelectrics is not a material constant; it can be engineered by strain, wrinkles, interlayer spacing, temperature, and domain-wall density.","Any C3-breaking perturbation, including thermal fluctuations, twist, edges, or defects, can nucleate switching, so nominally identical samples may show very different switching fields depending on preparation.","The avalanche mechanism should appear in other C3-symmetric sliding ferroelectrics, including 3R-MoS2 bilayers and MoS2/WS2 heterobilayers, not just h-BN.","The leading role of the domain-wall center means switching speed is governed by the large off-diagonal Born effective charge near P=0, which could account for the observed ultrafast polarization reversal."],"supporting_citations":[{"why":"Predicts sliding ferroelectricity in binary compound bilayers, the concept the paper builds on.","marker":"[13]"},{"why":"Reports experimental interfacial ferroelectricity in h-BN bilayer by sliding, providing the measured ~0.03 V/Å coercive field used for comparison.","marker":"[14]"},{"why":"Demonstrates stacking-engineered ferroelectricity in bilayer h-BN, establishing AB/BA stacking as the polar state.","marker":"[15]"},{"why":"Provides the stacking modes of bulk and bilayer h-BN used to define the AA', AA, AB, BA, and SP configurations.","marker":"[29]"},{"why":"Supplies the density-functional perturbation theory used to compute Born effective charges and resistance forces.","marker":"[31]"},{"why":"Gives the BaTiO3 Born effective charge tensor used as the conventional ferroelectric comparison (diagonal $Z^*$).","marker":"[33]"}],"fun_headline_variants":["Off-diagonal charge triggers avalanche switching in h-BN","Coercive field in h-BN set by symmetry-breaking","Wriggling domain walls speed up h-BN ferroelectric switching","h-BN switches like an avalanche, not a fixed field","Off-diagonal Born charge dictates h-BN switching dynamics"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the quasistatic force-balance estimate $E^r_z = f^r/Z^*_{xz}$, evaluated at frozen intermediate configurations starting from an arbitrary tiny sliding of 0.045 Å, represents the real switching trajectory; if inertia, dissipation, or field-dependent deformation changes that path, the monotonic decrease and coercive values would not hold.","fun_headline_variants_meta":{"raw":{"variants":["Off-diagonal charge triggers avalanche switching in h-BN","Coercive field in h-BN set by symmetry-breaking","Wriggling domain walls speed up h-BN ferroelectric switching","h-BN switches like an avalanche, not a fixed field","Off-diagonal Born charge dictates h-BN switching dynamics"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001334,"raw_usage":{"total_tokens":5431,"prompt_tokens":954,"completion_tokens":4477,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":570,"completion_tokens_details":{"reasoning_tokens":4391}},"tokens_in":570,"tokens_out":4477,"duration_ms":29469,"temperature":1.0,"reasoning_tokens":4391,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:41:06.402722+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure polarization reversal in an ultraclean, strain-free, atomically flat h-BN bilayer at low temperature under a perfectly homogeneous vertical electric field; the paper predicts zero or near-zero probability of switching unless a symmetry-breaking perturbation nucleates sliding, so observing low-field uniform switching would falsify the avalanche picture.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports experimental interfacial ferroelectricity in h-BN bilayer by sliding, providing the measured ~0.03 V/Å coercive field used for comparison."},{"cited_title":"Constantinescu, A","cited_arxiv_id":null,"evidence_quote":"Provides the stacking modes of bulk and bilayer h-BN used to define the AA', AA, AB, BA, and SP configurations."},{"cited_title":"Ghosez, J.-P","cited_arxiv_id":null,"evidence_quote":"Gives the BaTiO3 Born effective charge tensor used as the conventional ferroelectric comparison (diagonal $Z^*$)."}],"review_version":1}