{"id":"2bf901fb-a6d1-44b9-ab31-37c45d901d4c","arxiv_id":"2501.01162","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"An STM study resolves the d-pi-d exchange coupling in a surface Fe2ReA4 tetramer and predicts the two-radical spin-excitation spectrum from one fitted coupling parameter.","lead":"Researchers built a tetramer of two iron atoms and four retinoic-acid molecules on a gold surface, then switched individual molecules into spin-carrying radicals and measured the magnetic coupling between them. The result shows a specific d-pi-d spin-coupling mechanism in which the radicals align the iron spins, a design-relevant step for molecule-based magnets.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 1.25 excitation ratio verifies |j| and S_AB=4, but it is invariant under j→−j; the experimental spectra do not establish the antiferromagnetic (ferrimagnetic) sign, which is imported from DFT/VBT.","rationale":"The reader's CONDITIONAL verdict is appropriate. I agree with the overall assessment, but I would elevate the sign ambiguity over the S_AB=4/Fe–Fe-exchange concern as the weakest point. The ratio 1.25 is actually strong evidence for S_AB=4: for any fixed dimer spin S the ratio is (S+1)/S, so the measured value singles out S=4 (or, equivalently, the maximum-spin sector) regardless of individual Fe moments; direct Fe–Fe exchange shifts different S sectors but does not change the within-sector gaps as long as S=4 remains the ground sector. The same cannot be said for the sign. Eq. (2) with j<0 gives the same two excitation energies 4|j| and 5|j| from the S_tot=5 ground state, so the measured 22.0/27.5 meV double step is equally compatible with a ferromagnetic ground state. Since the paper's central narrative is 'ferrimagnetic coupling' (antiferromagnetic radical–Fe, parallel Fe spins), and since STM-IETS at zero field is sign-blind, the crucial experimental evidence for ferrimagnetism is missing. This is a correctness risk rather than an internal inconsistency: the DFT/VBT sign may well be right, but it is not tested by the headline ratio. The proposed field-dependence or line-shape-amplitude check would settle it. Thus no verdict change from the reader's conditional; the condition should explicitly require sign-sensitive evidence or tempering the ferrimagnetic claim.","tokens_in":9276,"tokens_out":17826,"duration_ms":179993,"concrete_test":"Re-fit the dI/dV line shapes of Fig. 3(e) with the Ternes scattering model under the two sign assignments (ground S_tot=3 vs S_tot=5), comparing the predicted step intensities and line shapes; if the two fits are not clearly distinguishable, measure the inelastic signals in a magnetic field up to a few tesla and track the Zeeman shifts/fan diagram of the two steps, from which the ground-state spin and sign of j can be read off. If both signs remain equally consistent, the abstract should be revised to state that the antiferromagnetic sign is a DFT prediction, not an experimentally established fact.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step in the central claim is not the S_AB=4 structure of Eq. (2), but the sign assignment j>0. The measured two-step spectrum at ΔE1=22.0 meV and ΔE2=27.5 meV has ratio 1.25, matching the Heisenberg ladder for a spin-1/2 pair coupled to S_AB=4. However, the whole level scheme is even under j→−j: replacing j by −j inverts the spectrum and swaps the order of the two S_tot=4 excited states, but the first two excitation energies from the new ground state are again 4|j| and 5|j|, with identical ratio. Zero-field inelastic tunneling measures only energy differences, so the observed steps cannot distinguish the ground-state total spin S_tot=3 (antiferromagnetic radical–Fe coupling, ferrimagnetic state) from S_tot=5 (ferromagnetic coupling). The ferrimagnetic claim in the abstract and conclusions therefore rests entirely on the DFT/VBT assertion that the radical–Fe exchange is antiferromagnetic, not on the headline experimental ratio. If the DFT sign were wrong, the identical data would correspond to a ferromagnetic ground state. The paper gives no magnetic-field or line-shape-amplitude analysis that would select the ground-state spin.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports low-temperature STM/AFM and dI/dV measurements on coordinated Fe–retinoic-acid (ReA) tetramers on Au(111), in which individual ReA molecules are dehydrogenated into spin-1/2 radicals. Single bridge-site radicals show a spin-excitation step at ΔE_B1 = 22.9 meV, single wing-site radicals show a much smaller step near 2.6 meV, and when the two bridge-site molecules are both radicals the spectrum shows two steps at ΔE1 = 22.0 meV and ΔE2 = 27.5 meV with ratio 1.25. The authors model the system with the Heisenberg Hamiltonian of Eq. (1), in which each bridge radical couples to the Fe dimer via j s·S_AB; for S_AB = 4 this gives ΔE2/ΔE1 = 5j/4j = 1.25, which they take as verification of the magnetic coupling between the two radical spins mediated by the Fe dimer. DFT and valence-bond theory are used to assign the antiferromagnetic (ferrimagnetic) sign of the radical–Fe exchange and to support the parallel alignment of the two Fe spins.","tokens_in":9545,"tokens_out":13328,"duration_ms":120391,"significance":"The parameter-free ratio ΔE2/ΔE1 = 1.25 is a clean fingerprint of a spin-1/2 pair exchange-coupled to an S = 4 spin center, and its agreement with the measured two-step spectrum is a genuine quantitative result that does not depend on any fitted exchange parameter. If the sign assignment is accepted, the work provides a striking atomic-scale realization of d–π–d-mediated ferrimagnetic coupling, with DFT connecting the measured excitation spectrum to the underlying electronic structure. The manuscript also contains useful control measurements (wing-wing and bridge-wing configurations) and a demonstration that tip manipulation can switch a coupled radical into a Kondo-dominated regime. The main weakness is that the zero-field inelastic tunneling data determine only the magnitude |j| and the S_AB = 4 manifold; the sign of j, and hence the claimed ferrimagnetic ground state, is imported from DFT/valence-bond theory and is not experimentally discriminated by the headline ratio.","major_comments":[{"comment":"The measured two-step spectrum is invariant under the sign change j → −j. For j > 0 the ground state is S_tot = 3 and the first two excitations are at 4j and 5j; for j < 0 the ground state is S_tot = 5 and the first two excitation energies from the new ground state are again 4|j| and 5|j|, with the two S_tot = 4 states merely ordered differently. Since zero-field inelastic tunneling spectroscopy measures only energy differences, the steps at 22.0 and 27.5 meV do not by themselves prove either the antiferromagnetic radical–Fe sign or the S_tot = 3 ground state. The ferrimagnetic claim in the Abstract and Conclusions therefore rests entirely on the DFT/valence-bond assignment of the sign, not on the experimental ratio; the text should state this separation of measured and inferred content explicitly, or provide an experimental discriminator such as a magnetic-field dependence of the step positions that selects the ground-state total spin.","section":"Theoretical analysis (Eq. 2) and Fig. 3(e); Abstract; Conclusions"},{"comment":"The DFT value for the wing-site excitation, ΔE_W1 = 5.2 meV, falls outside the observed experimental range 0.1–3 meV, and the paper attributes this to substrate effects without a quantitative or substrate-including calculation. Because the antiferromagnetic sign of both j and J is taken from the same DFT framework, this quantitative failure at the wing site weakens the theoretical support for the sign assignment; the manipulation experiment in Fig. 5 shows that j changes from roughly 5.1 meV to 1.2 meV under a local environment change, so the robustness of the computed sign to the acknowledged substrate and conformational variability should be addressed.","section":"Experimental and DFT Results, Fig. 2(f)"},{"comment":"The manuscript reports no uncertainty or linewidth for the fitted excitation energies ΔE1 and ΔE2, yet the claim of 'perfect agreement' with the ratio 1.25 is asserted to three significant figures. Given the visible width of the steps in Fig. 3(e) and the site-to-site scatter of the single-bridge excitations (19–25 meV), the paper should report the fitting errors and state explicitly that the ratio test confirms the model only within those uncertainties; it should also note that the absolute magnitudes predicted with j = 5.1 meV (20.4 and 25.5 meV) agree with the measured 22.0 and 27.5 meV only at the 8% level.","section":"Fig. 3(e) and fitting procedure (SM Sec. 1.2)"},{"comment":"The model treats the Fe dimer as a rigid S_AB = 4 spin and neglects direct Fe–Fe exchange; this is justified only if the Fe–Fe exchange exceeds j and keeps the dimer in its maximal-spin manifold. The observed ratio is indeed consistent with S_AB = 4 (the predictions for S_AB = 3, 2, and 1 would be 4/3, 3/2, and 2 respectively), but the paper should state the relevant energy scales and, ideally, report the DFT-computed Fe–Fe exchange, so that the 'verification of S_AB = 4' claim is not presented as if the model's assumption were itself the evidence.","section":"Theoretical analysis, Eq. (2)"}],"minor_comments":[{"comment":"The phrase 'in the x-↓-x form' appears to be a rendering artifact and should be corrected, as should the missing space in the title 'observation ofd-π-d' in the arXiv version.","section":"Introduction"},{"comment":"The measurement temperature is not given in the main text; it should be stated because it determines the thermal broadening of the 22 meV steps.","section":"Experimental methods / Figs. 2–3"},{"comment":"The relative intensities of the two inelastic steps in the two-bridge case are not discussed; the model predicts specific transition matrix elements, and comparing the measured step heights with the model would strengthen the identification of the two S_tot = 4 final states.","section":"Fig. 3(e)"},{"comment":"The paper should explicitly state that the single-bridge fit j = 5.1 meV predicts two-bridge steps of 20.4 and 25.5 meV, which fall within the observed site-to-site variation of the bridge-site excitation energy; the text currently emphasizes only the ratio, which is parameter-free.","section":"Theoretical analysis"},{"comment":"The statement that the wing-site DFT overestimate is explained by substrate interaction is plausible but qualitative; the Fig. 5 result (j reduced from 5.0 to 1.2 meV) supports the mechanism and should be used explicitly as the quantitative comparison.","section":"Effect of substrate"}],"recommendation":"major_revision","confidential_remarks":"The central experimental result is strong and the parameter-free ratio is a genuine asset, but the sign-blindness of the zero-field measurement is the main gap: the abstract and conclusions assert a ferrimagnetic ground state that the experiment cannot distinguish from a ferromagnetic one. A careful reframing that separates measured quantities (|j|, S_AB = 4 manifold) from theoretically assigned ones (sign of j) would make the paper publishable; field-dependent data, if available, would be the cleanest way to close the gap. The manuscript also needs a modest editorial pass for the garbled text in the introduction."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is worth your time: it reports a clean, parameter-free consistency test for a d-π-d spin coupling model in a surface-supported Fe2(ReA)4 tetramer. The genuinely new step is site-selective generation of two bridge-site radicals, which gives two inelastic tunneling steps at 22.0 and 27.5 meV with ratio 1.25. Fitting j from the single-bridge spectrum (j=5.1 meV) predicts Δ2/Δ1 = 5/4 exactly. That is a nice test, and the bridge-site DFT excitation energy (23.1 meV) matches experiment well.\n\nSoft spots, in proportion. Most important: the ratio is invariant under j → −j. Zero-field IETS only measures energy differences, so the data cannot distinguish the ferrimagnetic ground state (Stot=3, radical spins antialigned with Fe) from a ferromagnetic one (Stot=5). The paper's claim that the experiment verifies antiferromagnetic coupling is not supported; the sign is imported from DFT and valence bond theory. The authors should either run magnetic-field or line-shape analysis to select the ground-state spin, or state plainly that the sign is theory-assigned. Second, the wing-site DFT value (5.2 meV) lies outside the measured 0.1–3 meV range. The substrate-competition explanation is plausible, and they show one manipulation switching the spectrum to Kondo-like, but the discrepancy remains unquantified. Third, the Heisenberg model assumes the Fe dimer is locked in S_AB=4 with no direct Fe–Fe exchange term. The ratio test does probe S_AB, but if Fe–Fe exchange were weak or antiferromagnetic, the level scheme would not be so simple. Fourth, no error bars for the fitted energies, no data availability statement; the site-dependent spread is mentioned but only representative spectra are shown.\n\nThese are not fatal. The coupling between the two bridge radicals through the Fe dimer is genuinely demonstrated, and the absolute energies (predicted 20.4/25.5 meV vs measured 22.0/27.5 meV) agree within the spread of single-molecule measurements. I would send this for peer review. The paper is for researchers in on-surface molecular magnetism, especially those combining STM with model Hamiltonians. The central d-π-d mechanism holds up; the sign of j is an interpretation, not a measurement.\n\nRecommendation: engage with it. Ask for a field or lineshape test of the sign, or an explicit caveat, and for quantification of the wing-site discrepancy.","headline":"Clean ratio test for d-π-d coupling, but the ferrimagnetic sign is theory-assigned, not experimentally verified.","tokens_in":10185,"tokens_out":6412,"would_cite":true,"duration_ms":51415,"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 1.25 spin-energy ratio reveals d-pi-d coupling","keywords":["d-pi-d spin coupling","metal-organic coordination","scanning tunneling microscopy","spin excitation spectroscopy","Heisenberg model","ferrimagnetism","organic radical","retinoic acid"],"falsifier":"If the model is wrong, the ratio of the two spin-excitation energies in a doubly bridge-radicalized tetramer will deviate from 1.25. A direct test is to measure the spin-excitation spectrum while applying an external magnetic field: the ground state is predicted to be total spin $S=3$, so the Zeeman splitting pattern of the two steps should match that assignment; a different total spin would produce a different pattern. Alternatively, replacing the Fe dimer with a nonmagnetic metal pair should make the two-step spectrum disappear.","tokens_in":1883,"feed_emoji":"🧲","tokens_out":3106,"duration_ms":97292,"temperature":0.7,"pith_summary":"This paper reports atomic-scale observation of d-pi-d spin coupling in a metal-organic coordination structure built from two iron atoms and four retinoic acid molecules on Au(111). By dehydrogenating selected molecules with the STM tip, the authors turn them into spin-1/2 radicals and measure spin excitations in differential conductance spectra. When the two bridging molecules are both radicalized, two excitation steps appear at 22.0 and 27.5 meV, and their ratio of 1.25 exactly matches a Heisenberg model in which the radical spins couple antiferromagnetically to a parallel Fe dimer. This establishes the ferrimagnetic ground state of the tetramer and provides a quantitative, parameter-free fingerprint for d-pi-d coupling.","feed_headline":"A 1.25 spin-energy ratio reveals d-pi-d coupling","feed_subtitle":"Two spin-excitation steps at 22.0 and 27.5 meV match a model of antiferromagnetic radical-iron coupling.","key_machinery":"The central object is the Heisenberg spin Hamiltonian $H_{bb} = j\\,\\mathbf{s}_{13}\\cdot\\mathbf{S}_{AB}$ (Eq. (2)), written for the two bridge-site radical spins (collectively $\\mathbf{s}_{13}$) and the Fe dimer treated as one spin $\\mathbf{S}_{AB}$ with length 4. Using the identity $\\mathbf{S}_{\\text{tot}} = \\mathbf{s}_{13} + \\mathbf{S}_{AB}$, the energies of the four lowest states become $4j$, $-j$, $0$, and $-5j$, so the first two excitation energies are $4j$ and $5j$. The ratio $\\Delta E_2/\\Delta E_1 = 5/4 = 1.25$ is independent of the coupling strength $j$, which is why the authors can claim a direct test of the spin structure rather than a fit parameter.","core_discovery":"By constructing a coordination tetramer of two Fe atoms and four retinoic acid molecules on Au(111) and then dehydrogenating selected molecules into spin-1/2 radicals with the STM tip, the authors observe spin excitations in dI/dV spectra. A single bridge-site radical gives one spin-flip step at about 22.9 meV, while a wing-site radical gives about 2.6 meV. When both bridge-site molecules are radicalized, two steps appear at $\\Delta E_1 = 22.0$ meV and $\\Delta E_2 = 27.5$ meV, with $\\Delta E_2/\\Delta E_1 = 1.25$. The central claim is that this 1.25 ratio is a fingerprint of d-pi-d coupling: in the Heisenberg Hamiltonian $H_{bb} = \\frac{j}{2}[S_{\\text{tot}}(S_{\\text{tot}}+1) - s_{13}(s_{13}+1) - S_{AB}(S_{AB}+1)]$, with the Fe dimer treated as one spin $S_{AB}=4$ and the two bridge radicals as a composite spin $s_{13}\\in\\{0,1\\}$, the four lowest states have energies $4j$, $-j$, $0$, and $-5j$, so the first two excitation energies are $4j$ and $5j$, giving exactly the measured ratio. The ground state is the ferrimagnetic $S=3$ configuration in which the radical pair is antiferromagnetically coupled to the Fe dimer while the two Fe spins are forced parallel.","pith_inferences":["If the ratio identity is general, any dimer of two $S=2$ metal ions bridged by two $S=1/2$ radicals should show the same 5/4 ratio regardless of chemical details, so a measured deviation would signal direct metal-metal exchange or asymmetric coupling and could serve as a diagnostic tool.","The substrate-competition result implies that the same molecular unit on an insulating surface or in a bulk crystal could exhibit much stronger d-pi-d coupling, potentially raising magnetic ordering temperatures as the introduction argues.","The STM dehydrogenation method could be extended to other carboxylate radicals to build larger coupled spin arrays, where the 1.25 ratio could act as a built-in check that the intended d-pi-d coupling path is active."],"forward_implications":["The 1.25 ratio is a coupling-strength-independent signature of two metal spins bridged by two radical spins, so it can be searched for in other metal-organic coordination geometries.","The ground state of the double-bridge-radical tetramer is a ferrimagnetic $S=3$ unit with a parallel Fe dimer ($S_{AB}=4$) and an anti-aligned radical pair, demonstrating atomic-scale construction of a ferrimagnetic building block.","The large difference between bridge-site coupling ($j \\approx 5.1$ meV) and wing-site coupling ($J \\approx 1.0$ meV) shows that coordination geometry controls the strength of d-pi-d exchange.","Competition between radical-Fe coupling and molecule-substrate coupling, observed when a radical is moved by STM pulses, explains the spread of measured excitation energies and indicates that weaker substrate coupling would make the d-pi-d interaction more dominant.","The spectroscopic two-step pattern offers a practical readout for verifying the integrity of d-pi-d coupled units in larger engineered spin arrays."],"supporting_citations":[{"why":"Provides the prior ReA radical system whose hydrogen-bond-mediated spin coupling and radical creation method this work extends to metal-coordinated structures.","marker":"[28]"},{"why":"Gives the perturbative scattering framework used to fit the dI/dV spectra and extract the spin-excitation energies.","marker":"[37]"},{"why":"Supplies the exchange model of zero-bias tunneling anomalies behind the spectral fitting.","marker":"[36]"},{"why":"Demonstrates creation of single-molecule radical spins by STM manipulation, the technique used here to switch ReA molecules into radicals.","marker":"[34]"},{"why":"Shows how a sigmatropic reaction generates a spin in a closed-shell molecule on a metal surface, supporting the dehydrogenation mechanism.","marker":"[35]"},{"why":"Reports a similar coordination structure of radical-bearing molecules with metal atoms on Au(111) used as a structural reference.","marker":"[11]"}],"fun_headline_variants":["1.25 ratio fingerprints d-pi-d spin coupling","Atomic-scale d-pi-d coupling in Fe-radical tetramers","Two spin steps reveal d-pi-d coupling in Fe-organic structures","Ferrimagnetic spin coupling imaged at atomic scale","d-pi-d spin coupling seen via 1.25 energy ratio"],"cache_read_input_tokens":12160,"weakest_assumption_plain":"The prediction rests on treating the two iron atoms as one rigid spin of length 4, with no direct iron-iron exchange and no substrate-induced renormalization large enough to alter the level ordering.","fun_headline_variants_meta":{"raw":{"variants":["1.25 ratio fingerprints d-pi-d spin coupling","Atomic-scale d-pi-d coupling in Fe-radical tetramers","Two spin steps reveal d-pi-d coupling in Fe-organic structures","Ferrimagnetic spin coupling imaged at atomic scale","d-pi-d spin coupling seen via 1.25 energy ratio"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001193,"raw_usage":{"total_tokens":4985,"prompt_tokens":1071,"completion_tokens":3914,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":687,"completion_tokens_details":{"reasoning_tokens":3825}},"tokens_in":687,"tokens_out":3914,"duration_ms":25601,"temperature":1.0,"reasoning_tokens":3825,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:33:37.108837+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"If the model is wrong, the ratio of the two spin-excitation energies in a doubly bridge-radicalized tetramer will deviate from 1.25. A direct test is to measure the spin-excitation spectrum while applying an external magnetic field: the ground state is predicted to be total spin $S=3$, so the Zeeman splitting pattern of the two steps should match that assignment; a different total spin would produce a different pattern. Alternatively, replacing the Fe dimer with a nonmagnetic metal pair should make the two-step spectrum disappear.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the prior ReA radical system whose hydrogen-bond-mediated spin coupling and radical creation method this work extends to metal-coordinated structures."},{"cited_title":"Bocquet, N","cited_arxiv_id":null,"evidence_quote":"Supplies the exchange model of zero-bias tunneling anomalies behind the spectral fitting."},{"cited_title":"Langner, S","cited_arxiv_id":null,"evidence_quote":"Demonstrates creation of single-molecule radical spins by STM manipulation, the technique used here to switch ReA molecules into radicals."},{"cited_title":"Karan, N","cited_arxiv_id":null,"evidence_quote":"Shows how a sigmatropic reaction generates a spin in a closed-shell molecule on a metal surface, supporting the dehydrogenation mechanism."},{"cited_title":"Minamitani, N","cited_arxiv_id":null,"evidence_quote":"Reports a similar coordination structure of radical-bearing molecules with metal atoms on Au(111) used as a structural reference."}],"review_version":1}