{"id":"4db1e9f4-3248-4c72-8022-679e844176a6","arxiv_id":"1908.00959","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Magnetic exchange force microscopy resolves atomic-site-dependent magnetic interactions in a Mn/W(110) spin spiral, with DFT attributing the contrast to Mn-terminated tips and competing direct and Zener-type exchange.","lead":"This paper measures the tiny magnetic force between a magnetized metal tip and a single layer of manganese atoms arranged in a twisting magnetic pattern called a spin spiral. It shows that force-based sensing reveals atomic-scale magnetic details that ordinary current-based imaging misses, and computer simulations explain why.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The exchange-force extraction assumes identical chemical forces at t_i and t'_i; if magnetoelastic or electronic symmetry breaking makes chemical contrast site-dependent, the central site-resolved exchange-force claim is contaminated.","rationale":"The reader's weakest-assumption analysis identifies exactly the same load-bearing point: the subtraction that isolates the exchange force presumes identical chemical forces at the two crystallographically equivalent sites. I agree with the CONDITIONAL verdict. The experimental work is careful, the DFT modeling is substantive, and the qualitative site-dependent differences between top and hollow sites are unlikely to be pure artifacts; however, the quantitative exchange-force field and the comparison of sensitivity to spin-polarized current are conditional on the chemical-equivalence premise. The excluded half of the spin-spiral data and the selection of a Mn-terminated tip in the calculations are additional limitations, but they are secondary: even if those were resolved, the central claim would still hinge on the Δfex subtraction. A single DFT check can settle whether the premise holds, so the appropriate disposition stays CONDITIONAL rather than ACCEPT or REJECT.","tokens_in":12643,"tokens_out":2921,"duration_ms":35039,"concrete_test":"Perform spin-orbit-coupled DFT on the Mn/W(110) spin spiral (or on representative local-moment configurations corresponding to φ = 0°, 90°, and 180° within a c(2x2) cell) with full structural relaxation, and compute (i) the relative vertical displacement between t_i and t'_i, and (ii) the difference in total tip-sample force for a non-magnetic tip (e.g., W) placed over t_i versus t'_i at the experimental height z1. If the non-magnetic force difference exceeds the experimental threshold (roughly the equivalent of 0.2 Hz in Δf or ~1 pN in force), the chemical-equivalence assumption fails and the extracted Fex values and the sensitivity comparison must be corrected.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The entire quantitative exchange-force analysis rests on the subtraction Δfex,i = Δfi - Δf'i, which is interpreted as the magnetic exchange contribution only because the two sites t_i and t'_i are assumed chemically equivalent (p. 6: 'As the Mn layer is planar due to pseudomorphic growth on W(110), atomic contrast emerging from chemical interaction is expected to result in the same variation of Δfch at all atomic sites'). This assumption is load-bearing: the reported site- and distance-dependent exchange forces, the 40 pN magnitudes, and the headline sensitivity claim all depend on Δfex,i containing no non-magnetic signal. The paper's defense is the planarity of the Mn layer, but the spin spiral is non-collinear, and spin-orbit coupling / DMI can induce site-dependent magnetoelastic displacements or local electronic reorganization even in a nominally planar layer. A few-picometer buckling or a site-dependent hybridization change would break the chemical equivalence of t_i and t'_i and would feed directly into Δfex,i. The DFT calculations model a collinear c(2x2) antiferromagnetic cell, so they cannot capture any spin-spiral-induced symmetry breaking in the chemical force; the experimental tip is not chemically characterized at the atomic scale. Thus the isolation of exchange forces is an unverified premise rather than a demonstrated result, and the central claim of enhanced magnetic sensitivity is only as strong as this premise.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports combined scanning tunneling and atomic force microscopy (SPEX) measurements on a Mn monolayer on W(110) exhibiting a cycloidal spin spiral. From constant-height images, the authors extract a site-resolved frequency-shift difference Δfex,i = Δfi - Δf'i between crystallographically equivalent top and hollow sites, assign it to magnetic exchange, and convert it to exchange forces Fex,i(z) as a function of tip-sample distance. They compare these forces with the spin-polarized current asymmetry Ai(z) and conclude that the exchange force field is more sensitive to atomic-scale magnetization variations than the tunneling current, particularly at top versus hollow sites and along the spin spiral. Density functional theory calculations for Fe-, Co-, and Mn-terminated tips are used to reproduce the site- and distance-dependent exchange forces; only the Mn-terminated tip gives qualitative agreement, which the authors interpret as evidence for a crossover between direct antiferromagnetic and Zener-type ferromagnetic exchange. The weak spin polarization of the tunneling current is explained by the dominance of low-spin-polarized pz states near the Fermi energy.","tokens_in":12932,"tokens_out":3884,"duration_ms":43784,"significance":"If the central claim holds, this would be the first quantitative atomic-scale exchange force field measurement on a chiral spin spiral and would establish a force-based route complementary to spin-polarized current methods for studying chiral magnetic structures. The experimental work has notable strengths: the measurements use careful constant-height imaging with simultaneous current and frequency-shift detection, the raw Δfex and IT quantities are reported with explicit error margins, and data affected by tip magnetization changes are identified and excluded. The DFT calculations are genuinely first-principles and the comparison across three tip terminations is systematic, though the theory-experiment match ultimately depends on an assumed apex termination. The significance of the result, if confirmed, is high for the MExFM and spintronics communities.","major_comments":[{"comment":"The central quantity Δfex,i = Δfi - Δf'i is interpreted as the magnetic exchange contribution only because the two sites ti and t'i are assumed to have identical chemical forces. The manuscript argues that the Mn layer is planar and therefore chemical contrast is identical, but this is not independently verified. A non-collinear spin spiral with DMI could in principle induce site-dependent magnetoelastic displacements or electronic reorganizations that break the chemical equivalence even in a nominally planar layer, and the DFT calculations use a collinear c(2x2) antiferromagnetic cell, so they cannot capture such spiral-induced symmetry breaking. This assumption is load-bearing for all quantitative exchange-force magnitudes, including the reported 40 pN values. Please provide a concrete test of chemical equivalence (for example, a non-magnetic reference measurement, a z-dependent chemical contrast analysis, or a non-collinear calculation of the chemical force), or explicitly rephrase the results as exchange-force differences under an unverified equivalence assumption.","section":"Section 4, p. 6"},{"comment":"The data for i = 7 to i = 14 were excluded because of a spontaneous change of the tip magnetization during acquisition. Consequently, the claim that the exchange force varies monotonically across the spin spiral is supported only by the half-period i = 1 to i = 7, and the full-period behavior was not experimentally demonstrated. The manuscript acknowledges this exclusion in the text, but the abstract and discussion nevertheless state a general sensitivity to atomic-scale magnetization variations along the spiral. Please state explicitly in the abstract or conclusions that the full-period dependence was not obtained, and temper the claim to the measured half-period.","section":"Section 4, p. 8-9"},{"comment":"The reported exchange forces Fex,i(Δz) are derived from Δfex,i using the Sader-Jarvis inversion, but no uncertainty is propagated to the force values. The manuscript gives ±0.2 Hz for Δfex and ±3% for IT, yet the central 'exchange force is more sensitive than the current' conclusion is based on comparisons of Fex and A without error bars on Fex. Please propagate the Δfex uncertainty through the force inversion, or provide a sensitivity analysis showing that the stated uncertainties translate into a force uncertainty small enough to support the site-dependent and distance-dependent claims.","section":"Section 4, Figure 2"},{"comment":"The agreement between experiment and DFT is established only after choosing a Mn-terminated tip and a distance offset Δd = 0.0 nm ↔ Δz = -0.16 nm. The manuscript states that Fe- and Co-terminated tips cannot explain the observations, but the experimental tip is not chemically characterized at the atomic scale, so the Mn-terminated assignment is selected by qualitative agreement rather than independently determined. This makes the claim that first-principles calculations 'reveal' the exchange mechanisms contingent on an untested assumption about the tip apex. Please present the tip-termination dependence as a model-dependent interpretation rather than a confirmed identification, or provide independent evidence for the Mn termination, for example from the measured total-force magnitude or from adsorption experiments.","section":"Section 5, p. 9-11"}],"minor_comments":[{"comment":"There is a typo in the sentence about the symmetric slab: 'modelled using by a symmetric slab' should be 'modelled using a symmetric slab'.","section":"Methods, p. 15"},{"comment":"The caption is inconsistent with the panel labels: it says 'Mn-terminated (a, c) tips' but panels (a) and (c) are both top-site results, while the hollow-site results appear in (e) and (g). Please correct the caption so that each tip termination is assigned to the correct panel.","section":"Figure 4 caption"},{"comment":"The DFT model approximates the spin spiral as a collinear c(2x2) antiferromagnetic order. This is a practical approximation, but its effect on the calculated exchange forces, especially for the hollow sites where multiple neighbors contribute, should be stated more prominently as a limitation.","section":"Section 5, p. 9"},{"comment":"The discussion of the surface tilt and its influence on Δfex is mentioned in the main text, but the resulting uncertainty on the absolute distance scale is not quantified in the main text. A brief quantitative statement would help the reader assess the robustness of the distance-dependent comparison with DFT.","section":"Supplementary Note 2"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is a solid experimental advance: it quantifies atomic-scale exchange forces on a chiral spin spiral and shows the force signal is more site-sensitive than the spin-polarized current. The main soft spot is the cancellation assumption, which is reasonable but not independently verified.\n\nThe genuinely new thing is the measurement itself. No one has quantified atomic-scale exchange forces on a non-collinear spin spiral before. The authors use their SPEX setup to acquire simultaneous constant-height current and frequency-shift images, then extract a site-resolved exchange force field along the Mn/W(110) cycloid. The top-site versus hollow-site difference is a clear experimental result, and the DFT calculations offer a plausible mechanism: a transition from long-range Zener exchange to short-range direct exchange, depending on tip termination. The experiments look careful. They report error bars on Δf and current, and they honestly state that data for half the spin spiral was excluded because the tip magnetization changed.\n\nThe load-bearing assumption is the usual one in exchange force subtraction: t_i and t’_i are crystallographically equivalent, so the chemical force cancels in Δf. The paper defends this by the planarity of the Mn layer, and for a long-period magnetic modulation that is plausible. But it is not empirically verified. A site-dependent magnetoelastic displacement of even a few picometers would leak into the exchange-force numbers. I don’t think this is a fatal flaw—the spin spiral is a weak perturbation on the lattice, and the authors’ argument is sound—but it should be acknowledged explicitly in the paper.\n\nTwo smaller issues: the error bars are not propagated to the derived Fex curves, so the 40 pN magnitude carries no uncertainty. And the DFT-experiment agreement is only obtained for a Mn-terminated tip, with Fe and Co terminations failing to reproduce the data. That is model selection, not an independent prediction. The mechanism story is therefore suggestive, not closed.\n\nThis paper is for the MExFM and SP-STM community, and for anyone working on chiral magnetism in thin films. It deserves a serious referee. With a short addition on the cancellation assumption and error propagation, it would be a solid publication. I would send it out.","headline":"Solid experimental extension of SPEX to a chiral spin spiral; the exchange-force extraction rests on a plausible but unverified cancellation assumption.","tokens_in":13443,"tokens_out":3519,"would_cite":true,"duration_ms":34231,"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":"On a cycloidal Mn spin spiral, the atomic-scale magnetic exchange force field measured with a combined force-and-current probe is more sensitive to local magnetization than the spin-polarized tunneling current.","keywords":["exchange force microscopy","spin spiral","Mn/W(110)","chiral magnetism","Dzyaloshinskii-Moriya interaction","spin-polarized scanning tunneling microscopy","Zener exchange","density functional theory"],"falsifier":"Apply the same constant-height SPEX protocol to a non-magnetic monolayer with the same lattice structure on W(110): if the frequency-shift difference between $t_i$ and $t'_i$ does not vanish, the $\\Delta f_{\\mathrm{ex}}$ channel is contaminated by non-magnetic interactions, and the quoted exchange forces would not be purely magnetic.","tokens_in":12493,"feed_emoji":"🧲","tokens_out":9426,"duration_ms":80139,"temperature":0.7,"pith_summary":"This paper claims that the magnetic exchange force between a magnetic tip and a chiral spin spiral can be measured atom by atom, and that this force signal is more sensitive to the local magnetization than the spin-polarized tunneling current recorded at the same probe position. It demonstrates this on a single monolayer of Mn on W(110), whose Dzyaloshinskii–Moriya interaction drives a cycloidal spin spiral. If correct, the work establishes a route to quantify direct and indirect exchange mechanisms in chiral magnetic structures at the atomic scale, complementary to current-based spin sensing.","feed_headline":"Exchange forces beat tunneling current at the atomic scale","feed_subtitle":"Force and current measured in one pass quantify the Mn spin spiral's exchange field atom by atom.","key_machinery":"The central object is the frequency-shift difference $\\Delta f_{\\mathrm{ex},i}$ recorded at two crystallographically equivalent sites ($t_i$ and $t'_i$) in each magnetic unit cell at constant height; because the Mn layer is planar and chemical forces are assumed identical at these sites, this difference isolates the magnetic exchange interaction. The paper combines this with simultaneous current detection (SPEX) and distance-dependent MExFIS spectroscopy, converts $\\Delta f_{\\mathrm{ex}}$ into exchange forces $F_{\\mathrm{ex}}$ using a standard frequency-shift-to-force inversion, and compares the result against DFT supercell calculations that define the exchange energy as $E_{\\mathrm{ex}} = E_{\\mathrm{ap}} - E_{\\mathrm{p}}$.","core_discovery":"On a single Mn layer on W(110), the frequency-shift difference between two crystallographically equivalent atomic sites within the c(2 x 2) magnetic unit cell, namely $\\Delta f_{\\mathrm{ex},i} = \\Delta f_i - \\Delta f'_i$, traces the out-of-plane component of the cycloidal spin spiral, reaching exchange forces up to 40 pN. The exchange force field varies between top and hollow sites and with tip-sample distance, indicating that different exchange mechanisms dominate at different atomic positions, while the spin polarization of the tunneling current stays small and site-independent. First-principles calculations reproduce the site- and distance-dependent exchange forces when the tip apex is manganese-terminated, and attribute the observed behavior to a competition between short-range antiferromagnetic direct d-d exchange and longer-range ferromagnetic Zener-type exchange mediated by s electrons.","pith_inferences":["If exchange forces are systematically more sensitive than tunneling current to local magnetization, force-based detection could characterize antiferromagnetic or chiral textures even on samples where spin-polarized STM contrast vanishes, including insulating magnets.","Since DFT predicts distinct distance-dependent force curves for Fe-, Co-, and Mn-terminated tips, site- and distance-resolved exchange force measurements might serve as a fingerprint of the tip apex chemistry.","The same SPEX/MExFIS protocol could be extended to measure anisotropic (Dzyaloshinskii–Moriya) exchange forces by probing along different crystallographic directions, potentially quantifying spin-orbit contributions to the force field."],"forward_implications":["The exchange force channel resolves the atomic lattice in both surface directions within the magnetic unit cell, where the constant-height current resolves only one direction.","The exchange force field at top and hollow sites shows distinct distance dependences, revealing that different exchange mechanisms (direct d-d versus indirect Zener-type) dominate at different sites.","Along the spin spiral, both the exchange force magnitude and the current asymmetry decay gradually from the parallel to the perpendicular magnetic alignment, tracking the out-of-plane magnetization component.","The weak spin polarization of the tunneling current is explained by the dominance of weakly spin-polarized pz states at the Fermi energy, a consequence of the large exchange splitting of Mn d-bands."],"supporting_citations":[{"why":"Establishes the cycloidal spin-spiral ground state of a Mn monolayer on W(110) and its periodicity, defining the sample under study.","marker":"[17]"},{"why":"Provides the antiferromagnetic c(2 x 2) unit cell and lattice constants used to assign top and hollow atomic sites.","marker":"[14]"},{"why":"Introduces the SPEX combination of magnetic exchange force and spin-polarized current detection that this work extends to a spin spiral.","marker":"[27]"},{"why":"Demonstrates atomic-resolution magnetic exchange force microscopy, the foundational detection scheme.","marker":"[24]"},{"why":"Supplies quantitative measurement of the magnetic exchange force across a vacuum gap, a baseline and method for converting frequency shifts to forces.","marker":"[30]"},{"why":"Provides the physical picture of competing direct antiferromagnetic and indirect Zener-type exchange used to interpret the site-dependent force curves.","marker":"[31]"},{"why":"Gives the frequency-shift-to-force inversion used to derive exchange forces from the measured frequency-shift differences.","marker":"[37]"},{"why":"Provides the standard tunneling model used to relate the measured current asymmetry to the spin polarization of the vacuum local density of states.","marker":"[41, 42]"}],"fun_headline_variants":["Exchange forces mapped atom by atom on a spin spiral","40 pN exchange forces reveal spin spiral's atomic texture","Force beats current: exchange interactions measured at atomic scale","Atomic-scale exchange forces probe chiral magnetism","New force-based method quantifies magnetic exchange on atomic scale"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The extraction of the exchange force rests on the assumption that the chemical (non-magnetic) force is identical at the two sites $t_i$ and $t'_i$, so that only magnetic exchange contributes to the measured frequency-shift difference.","fun_headline_variants_meta":{"raw":{"variants":["Exchange forces mapped atom by atom on a spin spiral","40 pN exchange forces reveal spin spiral's atomic texture","Force beats current: exchange interactions measured at atomic scale","Atomic-scale exchange forces probe chiral magnetism","New force-based method quantifies magnetic exchange on atomic scale"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000206,"raw_usage":{"total_tokens":1355,"prompt_tokens":862,"completion_tokens":493,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":478,"completion_tokens_details":{"reasoning_tokens":418}},"tokens_in":478,"tokens_out":493,"duration_ms":5163,"temperature":1.0,"reasoning_tokens":418,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:26:42.572465+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Apply the same constant-height SPEX protocol to a non-magnetic monolayer with the same lattice structure on W(110): if the frequency-shift difference between $t_i$ and $t'_i$ does not vanish, the $\\Delta f_{\\mathrm{ex}}$ channel is contaminated by non-magnetic interactions, and the quoted exchange forces would not be purely magnetic.","supporting_citations":[],"review_version":1}