{"id":"9d0c03f2-aff5-465a-8e64-882385b96a35","arxiv_id":"1908.05802","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Simulated X-ray Raman spectra of iron-sulfur dimers highlight low-lying d-d states that stay dark in absorption, suggesting a route to selectively excite them.","lead":"Researchers simulated what X-ray Raman spectroscopy would see for [2Fe-2S] iron-sulfur clusters, using accurate quantum chemistry calculations. They found that the Raman signal highlights different low-energy states than ordinary absorption, which could let experiments selectively excite states that are otherwise invisible.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation (9) expands each field separately and keeps only the (k_s·r)(k_p·r) cross term; the exact phase is e^{i q·r}, whose consistent second-order term is -(q·r)^2/2. The SXRS intensities and central selective-excitation claim may be unsupported.","rationale":"The reader's weakest assumption was the accuracy of the state-averaged DMRG results at D=2000, especially for states 10–19 where the partial D=3000 comparison shifts energies by up to -0.28 eV. That concern is legitimate and the paper itself labels the active-space model as 'qualitative', so the quantitative ordering of closely spaced states is not firmly established. However, it is not the most load-bearing weakness. The paper's central claim is a spectroscopic prediction: that SXRS and absorption access different states and that the intensity differences enable selective excitation. This claim rests on the expression for the SXRS signal, specifically on the effective transition charge density α_eg of Eq. (10). The derivation of α_eg from Eq. (9) is not a consistent long-wavelength expansion: the exact phase factor e^{i q·r} has a cancellations among the second-order terms that the paper ignores. Keeping only the cross term (k_s·r)(k_p·r) generates an artificially large, geometry-independent quadrupole-like coupling. The correct leading terms are either the dipole term i q·r or the quadrupole term -(q·r)^2/2, both of which are much smaller for near-collinear X-ray beams and have different state selectivity. Consequently, the computed SXRS spectra and the derived absorption-active versus Raman-active separation are not trustworthy. This is an internal inconsistency in the central observable, not merely an accuracy limitation. The DMRG convergence issue would only matter after the signal expression is corrected. A revision should re-derive the SXRS signal from the full σ_eg(q) or a consistent q expansion and then re-evaluate the predicted dark-state accessibility. Until then, the central claim is unsupported as written.","tokens_in":10367,"tokens_out":16383,"duration_ms":155637,"concrete_test":"Recompute the SXRS spectra in Fig. 3 using the exact transition charge density form factor σ_eg(q) = ∫ σ_eg(r) e^{i q·r} dr, with q = k_s - k_p determined by the stated beam geometry. If the geometry is unspecified, test the two extremes: (i) collinear beams with |q| = ω_eg/c, where the leading terms are i q·r and -(q·r)^2/2; (ii) a scattering geometry with large q, keeping the full exponential. Compare the Y-polarized SXRS band near 2.9–3.1 eV to the absorption band near 2.2–2.4 eV. If the relative enhancement disappears or inverts, the central claim of selective excitation of dark states fails.","verdict_should_be":"REJECT","load_bearing_attack":"The central claim depends directly on the computed SXRS spectra, which are built on the effective transition charge density α_eg defined in Eq. (10). The derivation of Eq. (9) is internally inconsistent. The exact spatial phase in the A^2 interaction is e^{i q·r} with q = k_s - k_p, but the paper expands each plane-wave factor to first order, obtaining 1 + i(k_s - k_p)·r + (k_s·r)(k_p·r), and then keeps only the last term while dismissing the first-order term. A consistent second-order expansion of e^{i q·r} is 1 + i q·r - (q·r)^2/2, and the (k_s·r)(k_p·r) term is only one of three equal-order contributions in the product expansion; the dropped diagonal terms -(k_s·r)^2/2 and -(k_p·r)^2/2 are of the same magnitude and, for collinear or nearly collinear X-ray beams, cancel almost entirely against it. Thus Eq. (10) is not the long-wavelength limit of the minimal-coupling A^2 interaction. Even if the DMRG wavefunctions were exact, the predicted Raman-active versus absorption-active intensity ordering would be artifacts of this expansion. The DMRG convergence concern raised by the reader is real but secondary: it affects the accuracy of the states, whereas the signal expression affects the very observable being predicted.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper computes stimulated X-ray Raman spectroscopy (SXRS) and X-ray absorption signals for a ferric-ferric and a ferric-ferrous [2Fe-2S] dimer, using electronic states obtained from state-averaged DMRG calculations with active spaces CAS(38e,36o) and CAS(39e,36o). The central claim is that, along one incidence axis, absorption and SXRS select different subsets of the dense low-lying d-d manifold, so that the two spectroscopies complement each other and may allow selective excitation of previously dark states. The SXRS signal is derived from a minimal-coupling Hamiltonian in terms of transition charge densities, and the spectra are presented for X, Y, and Z polarized incidence directions.","tokens_in":10651,"tokens_out":6081,"duration_ms":56905,"significance":"If the computed spectra are correct, the paper offers a concrete, falsifiable proposal: direction-dependent SXRS can reveal dark d-d states in iron-sulfur clusters, with predicted energy windows (for example, absorption-active around 2.2-2.4 eV versus Raman-active around 2.9-3.1 eV for the ferric-ferric dimer). The work combines state-of-the-art DMRG with explicit transition charge densities and natural transition orbital analysis, which is a strength: the state assignments are grounded in well-defined ab initio quantities rather than in the spectral features being predicted. The central prediction is specific enough to be tested by future experiments, and the computed state manifold itself is a useful benchmark for the electronic structure of these clusters.","major_comments":[{"comment":"The derivation of the effective transition charge density α_eg is internally inconsistent. The exact spatial phase in the σ(r)A^2 interaction is e^{i q·r} with q = k_s - k_p; a consistent second-order expansion is e^{i q·r} ≈ 1 + i q·r - (q·r)^2/2. Equation (9) instead expands each plane-wave factor separately, obtaining 1 + i q·r + (k_s·r)(k_p·r), and then keeps only the cross term while dropping the two diagonal terms -(k_s·r)^2/2 and -(k_p·r)^2/2. Those diagonal terms are of the same order as the kept term; for collinear, equal-magnitude wavevectors they cancel it identically (since then q = 0 and the exact phase is 1). Thus Eq. (10) is not the long-wavelength limit of the minimal-coupling A^2 interaction, and the SXRS intensities in Eq. (11) — including the predicted separation into absorption-active and Raman-active states — are built on an incorrect expression. The authors should re-derive the signal using the full q-dependent phase or a consistent multipole expansion, and recompute the spectra.","section":"Eq. (9)-(10), 'Simulation of stimulated X-ray Raman (SXRS) signals'"},{"comment":"The DMRG convergence check is not sufficient to support the quantitative energy separation that underlies the selective-excitation claim. The paper reports that for states 10-19 the average excitation-energy shift between D=2000 and D=3000 (the latter with only 2 sweeps) is -0.28 eV for the ferric-ferric dimer and -0.23 eV for the ferric-ferrous dimer. The ferric-ferric Raman-active band at 2.9-3.1 eV is separated from the absorption-active band at 2.2-2.4 eV by roughly 0.5 eV, so a 0.28 eV uncertainty could reorder the relevant states; moreover, the intensities entering the direction-dependent spectra are transition-charge-density matrix elements whose D=2000 convergence is not separately assessed. Since the central proposal depends on which specific states are absorption-bright versus Raman-bright, the authors should provide convergence evidence for the transition densities and for the relative intensities, or temper the selective-excitation conclusion accordingly.","section":"Table 1, Fig. 3, and 'Computational methods'"},{"comment":"The notation for the field vectors in Eqs. (8)-(10) is confusing and likely dimensionally inconsistent. In Eq. (9) the expansion is in the wavevectors k_si and k_pi, so the second-order term involves (k_si·r)(k_pi·r); in Eq. (10) the same term is written with (ϵ_si·r)(ϵ_pi·r), where ϵ_si and ϵ_pi are described as 'direction of propagation' but earlier in the text ϵ denotes polarization. If ϵ are unit vectors along k, the prefactor must contain the corresponding |k_s||k_p| = ω_s ω_p/c^2 factors; the current prefactor in Eq. (10) does not make this explicit. This should be clarified, since the numerical spectra depend on the definition of α_eg.","section":"Eq. (8)-(10), notation"}],"minor_comments":[{"comment":"The sentence 'The first term vanishes by the definition of the transition charge density' would benefit from an explicit justification: ∫ σ_eg(r) dr = ⟨e|g⟩ = 0 for e ≠ g, and the text should state this rather than relying on an implicit definition.","section":"Eq. (9)"},{"comment":"The phrase 'theoretically predicted dense low-lying excited states' is used without a citation to the earlier prediction; adding a reference (e.g., Ref. 18) at that point would help the reader connect the claim to the literature.","section":"Abstract and Conclusions"},{"comment":"There is a typo in the acknowledgment: 'gratefully acknowledges the the support' should read 'gratefully acknowledges the support'.","section":"Acknowledgment"},{"comment":"The acronym 'MCLT' appears where 'MLCT' (metal-to-ligand charge transfer) is presumably intended; this should be corrected.","section":"Page 7, 'Computational methods'"},{"comment":"The caption states that signals are normalized, so SXRS and absorption strengths cannot be directly compared; however, the text in the results section talks about 'signal enhancements' in SXRS relative to absorption. It would be helpful to state explicitly whether the claimed enhancements are relative intensities within the normalized spectra or absolute cross-section statements.","section":"Fig. 3 caption"}],"recommendation":"major_revision","confidential_remarks":"The central issue is the derivation of Eq. (10): the inconsistent second-order expansion of the spatial phase directly undermines the computed SXRS spectra and the main claim of absorption/Raman selectivity. This is a fixable problem in principle — the correct expansion e^{i q·r} or a full treatment of the q-dependent phase can be used — but it requires re-deriving the signal expression and recomputing the spectra, so the present version cannot be accepted as is. The DMRG convergence caveat is secondary but should also be addressed. The paper is within the journal's scope and the overall approach (minimal-coupling signal theory combined with ab initio DMRG transition densities) is worth pursuing."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. The DMRG study of the low-lying states of [2Fe-2S] dimers is careful and useful: enlarged active space, 20 states, natural-transition-orbital analysis, and the absorption spectra are a clean extension of the earlier work by Sharma et al. But the SXRS signal expression is derived incorrectly, and the central claim about accessing dark states rests on that error.\n\nThe problem is Eq. (9). The exact spatial phase in the A^2 interaction is e^{i q·r} with q = k_s - k_p. The paper instead writes e^{i k_s·r} e^{-i k_p·r} as (1 + i k_s·r)(1 - i k_p·r) and keeps only the cross term (k_s·r)(k_p·r), justifying it by saying the linear term is small. That is not a consistent second-order expansion. The correct expansion of e^{i q·r} to second order is 1 + i q·r - (q·r)^2/2. The cross term is one of three equal-order contributions; the diagonal terms are dropped. For the geometry used here—all pulses along the same axis, X, Y, or Z—q is nearly zero, so the true SXRS signal is a tiny dipole term, while the paper's α_eg has a large spurious (k·r)^2 factor. The computed Raman-active vs absorption-active state ordering, and therefore the selective-excitation proposal, are artifacts of this expansion.\n\nThat's a load-bearing flaw, not a minor approximation. The DMRG convergence caveats—partial D=3000 data showing -0.28 eV shifts for the higher states, and the qualitative active space—are secondary. The absorption spectra and state assignments are probably fine, and the analysis of the d-d manifold is a genuine contribution.\n\nOn the citation pattern: it's fair, no red flags. No code or data shipped, but that's common for this type of paper.\n\nBottom line: this paper should go to peer review, but the referee should ask for a re-derivation of the SXRS signal using the full q-dependent transition charge density or a proper multipole expansion. If the corrected calculation no longer shows the absorption/SXRS contrast, the central claim should be withdrawn. The electronic structure part can stand on its own.","headline":"The DMRG excited-state calculations are solid and worth having, but the SXRS signal derivation in Eq. (9) is wrong, so the central absorption/Raman contrast and selective-excitation claim do not stand as written.","tokens_in":11249,"tokens_out":14217,"would_cite":true,"duration_ms":126469,"reading_group":"yes","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Two X-ray spectroscopies complementarily expose the dipole-forbidden d-d states of [2Fe-2S] dimers, with the Raman channel revealing states hidden from absorption.","keywords":["iron-sulfur clusters","stimulated X-ray Raman spectroscopy","absorption spectroscopy","density matrix renormalization group","transition charge density","d-d excited states","selective excitation"],"falsifier":"Measure the Y-polarized SXRS and absorption spectra of an oriented synthetic [2Fe-2S] model complex across the predicted window: for the ferric-ferric dimer the absorption-active band should peak near 2.2–2.4 eV and the Raman-active band near 2.9–3.1 eV, and in the mixed-valence dimer states 2 and 4 should appear only in absorption while states 5, 7, and 8 appear only in Raman. Alternatively, a DMRG calculation at D=3000 or with dynamic correlation that shifts the 10th–19th excited states by more than the predicted energy gap between the two bands would overturn the selective-excitation proposal.","tokens_in":10150,"feed_emoji":"⚛️","tokens_out":9891,"duration_ms":77537,"temperature":0.7,"pith_summary":"Iron-sulfur [2Fe-2S] dimers are the electron-transfer cofactors of ferredoxins, and their dense low-lying d-d excited states are electric-dipole forbidden, making them hard to see with ordinary absorption. This paper computes, from first principles, the absorption and stimulated X-ray Raman (SXRS) spectra of the homovalent ferric-ferric and mixed-valence ferric-ferrous dimers, using excited states and transition charge densities from density matrix renormalization group (DMRG) calculations. The central claim is that the two spectroscopies are complementary: along one polarization axis, absorption-active states (2.2–2.4 eV in the ferric-ferric dimer) are clearly separated in energy from Raman-active states (2.9–3.1 eV), and the mixed-valence dimer shows states that are almost exclusively seen by one technique or the other. If these predictions hold, SXRS would provide a way to access previously dark states and to selectively excite them by tuning the pump, opening a window onto the electronic dynamics underlying electron transfer.","feed_headline":"Raman X-rays reveal dark states in iron-sulfur electron carriers","feed_subtitle":"DMRG simulation shows a Y-polarized pump can selectively excite states that absorption cannot reach.","key_machinery":"The central object is the transition charge density (TCD) $\\sigma_{eg}(\\mathbf{r}) = \\langle e | \\hat{\\sigma}(\\mathbf{r}) | g \\rangle$, the matrix element of the charge density operator between ground and excited states, computed here from DMRG wavefunctions. For off-resonant SXRS the authors use the minimal-coupling Hamiltonian and a long-wavelength approximation to obtain an effective TCD $\\alpha_{eg} = \\frac{e}{2m\\hbar c^2 \\omega_{s_i}\\omega_{p_i}} \\int (\\boldsymbol{\\epsilon}_{s_i}\\cdot\\mathbf{r})(\\boldsymbol{\\epsilon}_{p_i}\\cdot\\mathbf{r})\\,\\sigma_{eg}(\\mathbf{r})\\,d\\mathbf{r}$, which enters the signal as products $\\alpha_{ge}\\alpha_{eg}$. The absorption signal instead uses the transition dipole $\\mu_{eg}$. The mechanism that carries the argument is the contrast between which states have large $\\mu_{eg}$ versus large $\\alpha_{eg}$; the TCD therefore determines both the complementarity of the two spectroscopies and the predicted selective-excitation window.","core_discovery":"The paper establishes that for an oriented [2Fe-2S] dimer, the off-resonant stimulated X-ray Raman signal and the linear absorption signal are almost identical for X and Z polarizations but diverge sharply for Y polarization. In the ferric-ferric dimer, absorption is concentrated in states at 2.2–2.4 eV while the Raman response is dominated by states at 2.9–3.1 eV, with the 6th, 7th, 13th, 15th, and 16th excited states strongly enhanced in SXRS relative to absorption. In the mixed-valence ferric-ferrous dimer, the 2nd and 4th excited states are almost exclusively absorption-active, and the 5th, 7th, and 8th states are Raman-active in Y polarization. The authors attribute the contrast to the different molecular properties probed: absorption depends on the transition dipole moment $\\mu_{eg}$, while SXRS depends on products of effective transition charge densities $\\alpha_{ge}\\alpha_{eg}$ that weight different regions of the transition density. They conclude that this intensity difference provides a means to access previously dark states and to selectively excite states by tuning the excitation bandwidth, enabling studies of the ensuing electronic dynamics.","pith_inferences":["The same transition-charge-density contrast likely applies to other multinuclear transition-metal clusters with dense dipole-forbidden manifolds, such as [4Fe-4S] clusters or the manganese-calcium cluster of photosystem II, making the proposed SXRS scheme a general probe of catalytically relevant dark states (editorial inference).","The long-wavelength approximation drops the linear momentum-transfer term in the TCD expansion; at hard-X-ray photon energies with larger momentum transfer, this term could alter the predicted Y-polarization contrast, and a full multipolar calculation would test whether the complementarity persists (editorial inference).","The single-molecule orientation assumed here would be degraded by rotational averaging in solution; simulating the rotationally averaged signals would show how much of the absorption/Raman separation survives in isotropic samples (editorial inference)."],"forward_implications":["A pump tuned to the Raman-active band (~2.9–3.1 eV in the ferric-ferric dimer) could selectively populate states that are dark in absorption, allowing time-resolved observation of their dynamics.","In the mixed-valence dimer, combining absorption and SXRS gives a more complete picture of the d-d manifold than either technique alone, since some states are visible to only one.","The computed spectra carry assignment information: the first band of the ferric-ferrous dimer is dominated by local ferrous d-d excitations (0.04 eV splitting consistent with Mössbauer estimates), while the second band is dominated by inter-center charge-transfer d-d excitations.","Because the absorption/Raman contrast appears only for Y polarization, the technique can also serve as a probe of molecular orientation in aligned samples."],"supporting_citations":[{"why":"previous DMRG study establishing the dense low-lying d-d manifold that this work extends.","marker":"[18]"},{"why":"supplies the enlarged active-space construction and the estimate that metal-to-ligand charge-transfer states lie too high to matter.","marker":"[19]"},{"why":"derives the minimal-coupling Hamiltonian expression for off-resonant Raman signals used in the SXRS simulation.","marker":"[27]"},{"why":"the DMRG program used to compute the multiple excited states and transition densities.","marker":"[24]"},{"why":"experimental RIXS evidence of the dense low-energy manifold that motivates the SXRS probe.","marker":"[25]"},{"why":"provides the synthetic [2Fe-2S] complex geometry used as the model system.","marker":"[9]"},{"why":"the protocol for preparing active-space orbitals by split-localized unrestricted natural orbitals.","marker":"[31]"}],"fun_headline_variants":["Y-polarized X-ray Raman exposes hidden states in iron-sulfur dimers","Polarization trick reveals dark states in electron-transfer dimers","X-ray Raman and absorption diverge: selective excitation key","Polarization-selective X-ray Raman accesses absorption-dark states","Y-polarized Raman reveals states absorption cannot reach"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that state-averaged DMRG at bond dimension D=2000 with the CAS(38e,36o) and CAS(39e,36o) active spaces fixes which states are absorption-bright versus Raman-bright, although the paper's own D=3000 check shifts the higher states by -0.28 eV and -0.23 eV and the active-space model is described as qualitative.","fun_headline_variants_meta":{"raw":{"variants":["Y-polarized X-ray Raman exposes hidden states in iron-sulfur dimers","Polarization trick reveals dark states in electron-transfer dimers","X-ray Raman and absorption diverge: selective excitation key","Polarization-selective X-ray Raman accesses absorption-dark states","Y-polarized Raman reveals states absorption cannot reach"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000678,"raw_usage":{"total_tokens":3093,"prompt_tokens":964,"completion_tokens":2129,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":580,"completion_tokens_details":{"reasoning_tokens":2044}},"tokens_in":580,"tokens_out":2129,"duration_ms":14405,"temperature":1.0,"reasoning_tokens":2044,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:04:26.979611+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the Y-polarized SXRS and absorption spectra of an oriented synthetic [2Fe-2S] model complex across the predicted window: for the ferric-ferric dimer the absorption-active band should peak near 2.2–2.4 eV and the Raman-active band near 2.9–3.1 eV, and in the mixed-valence dimer states 2 and 4 should appear only in absorption while states 5, 7, and 8 appear only in Raman. Alternatively, a DMRG calculation at D=3000 or with dynamic correlation that shifts the 10th–19th excited states by more than the predicted energy gap between the two bands would overturn the selective-excitation proposal.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"previous DMRG study establishing the dense low-lying d-d manifold that this work extends."},{"cited_title":"G.; DeBeer, S.; Neese, F","cited_arxiv_id":null,"evidence_quote":"supplies the enlarged active-space construction and the estimate that metal-to-ligand charge-transfer states lie too high to matter."},{"cited_title":"Y.; Saurabh, P.; Mukamel, S","cited_arxiv_id":null,"evidence_quote":"derives the minimal-coupling Hamiltonian expression for off-resonant Raman signals used in the SXRS simulation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"the DMRG program used to compute the multiple excited states and transition densities."},{"cited_title":"E.; Hahn, A","cited_arxiv_id":null,"evidence_quote":"experimental RIXS evidence of the dense low-energy manifold that motivates the SXRS probe."},{"cited_title":"B.; Holm, R.; Ibers, J","cited_arxiv_id":null,"evidence_quote":"provides the synthetic [2Fe-2S] complex geometry used as the model system."},{"cited_title":"Electronic landscape of the P-cluster of nitrogenase as revealed through many-electron quantum wavefunctions","cited_arxiv_id":"1810.10196","evidence_quote":"the protocol for preparing active-space orbitals by split-localized unrestricted natural orbitals."}],"review_version":1}