{"id":"2353f067-9465-4573-841e-ec1b39f44a46","arxiv_id":"2506.18984","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A Green's function topological invariant distinguishes symmetry-allowed from symmetry-forbidden 4pi electrocyclizations through crossings of poles or zeros at the chemical potential.","lead":"The paper shows that, in strongly correlated molecules, a reaction that is 'symmetry-forbidden' by the Woodward-Hoffmann rules can be identified by a crossing of Green's function zeros at the Fermi level rather than a crossing of molecular orbital energies. It introduces a symmetry-resolved topological invariant that changes only at such crossings and demonstrates it on the ring-closing of butadiene to cyclobutene.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Zero-crossing diagnostic is defined relative to a geometry-dependent chemical potential; the paper does not test whether a different (equally valid) mu changes the zero crossings or Delta N +/-.","rationale":"The reader's CONDITIONAL verdict is appropriate. The physical picture is plausible and the CASSCF/TRIQS calculations support the qualitative statement on the example system. However, the central invariant N(R) depends on a geometry-dependent zero-frequency origin chosen via the Mulliken chemical potential. Topological invariants should be robust to continuous deformations that do not close the gap; the paper does not demonstrate that the zero-crossing at omega=0 and the sign of Delta N+/- are independent of this convention. The active-space issue is secondary but real, since the CASSCF(4,4) Green's function is a truncated model of the full many-body problem. A direct numerical check with shifted mu and a larger active space would settle the concern. I do not see an internal inconsistency or a reason to reject; the appropriate verdict remains conditional on these robustness tests.","tokens_in":18037,"tokens_out":13709,"duration_ms":152891,"concrete_test":"Recompute the disrotatory and conrotatory Green's-function maps and the invariants N+/- (R) for reactant and product using a set of constant shifts mu'(R)=mu(R)+delta with delta from -0.1 to +0.1 hartree (staying within the spectral gap at every geometry) and also using a fixed mu0 equal to the reactant value. If Delta N_+ and Delta N_- remain +1/-1 for the forbidden pathway and 0 for the allowed pathway for all delta, the diagnostic is gauge-robust; if any shift changes the invariant differences or moves a zero crossing off omega=0, the classification is convention-dependent and the central claim requires qualification.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The topological classification rests on the convention omega=0 equivalent to mu(R), with mu(R)=-(IP+EA)/2 (SI Eq. S4) recomputed at every geometry. All zero/pole crossings and the invariants N+/- (R) are measured against this geometry-dependent frequency origin. If a different but equally valid chemical potential were used, such as a fixed mu0 or another value inside the HOMO-LUMO gap, the locations of the zero crossings and the values of Delta N_+ and Delta N_- could change, potentially erasing the claimed distinction between the forbidden (Delta N_+=+1, Delta N_-=-1) and allowed (Delta N=0) pathways. The paper performs no such check: no scan over mu, no comparison with a fixed chemical potential, and no analysis of the zero trajectory in an absolute energy frame. In addition, the correlation-induced zeros are computed from a minimal CASSCF(4,4) active space without convergence tests; because the zeros are an artifact of static correlation, their position near omega=0 could shift with active-space size. Since the central claim is that the Green's-function zero crossing serves as an unambiguous reaction classifier, this untested gauge and numerical sensitivity is the weakest link in the argument.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a Green's-function-based topological invariant for classifying orbital-symmetry-controlled chemical reactions. The central object is N(R) = ∮ dω/(2πi) ∂_ω ln Det G(ω|R), whose symmetry-resolved versions N_+(R) and N_-(R) count Green's-function poles and zeros above ω=0 in each symmetry sector. For the 4π electrocyclization of butadiene, the authors compute CASSCF(4,4)/def2-SVP Green's functions along conrotatory and disrotatory IRCs and find that the symmetry-forbidden disrotatory pathway exhibits a crossing of Green's-function zeros at ω=0 (with ΔN_+=+1 and ΔN_-=-1), whereas the symmetry-allowed conrotatory pathway shows no such crossing (ΔN=0). The paper also applies the formalism to an asymmetric photoswitch where spatial symmetry is broken, reporting a maintained zero crossing at CASSCF(8,8) level.","tokens_in":18296,"tokens_out":3793,"duration_ms":44533,"significance":"If the diagnostic is robust, this work provides a conceptually novel way to identify symmetry-forbidden reactions in strongly correlated regimes where molecular-orbital crossings are not well defined. The external benchmark against the Woodward-Hoffmann classification is appropriate, and the paper contains enough computational detail (Cartesian coordinates, active-space definitions, winding-number evaluation procedure) to be reproducible. The main value lies in connecting modern topological invariants for Green's functions to mainstream quantum chemistry; the presentation is clear and the illustrative examples are well chosen.","major_comments":[{"comment":"The entire classification is defined relative to a geometry-dependent chemical potential μ(R)=-(IP+EA)/2. The zero-crossing events and the values of N_+(R), N_-(R) and their differences are evaluated at ω=0 in this shifted frequency frame. The authors do not test whether a different equally valid reference, such as a fixed μ_0 or another point inside the HOMO-LUMO gap, changes the number or location of zero crossings and hence the claimed ΔN_+=+1, ΔN_-=-1 versus ΔN=0 distinction. This is load-bearing because the invariant's meaning changes when the frequency origin moves with geometry; a numerical scan over μ is needed to establish that the classification is not an artifact of the Mulliken convention.","section":"SI Eq. (S4) and Sec. II.D"},{"comment":"The central calculations for the main reaction use a minimal CASSCF(4,4)/def2-SVP active space. Since the Green's-function zeros are presented as a consequence of static correlation, their position near ω=0 could shift with active-space size or basis set; no convergence test with a larger active space or basis is reported. The photoswitch calculation uses CASSCF(8,8) but there is no systematic comparison between the two levels, so the robustness of the zero-crossing diagnostic across levels of theory is unestablished.","section":"Sec. II.B and Sec. II.C"},{"comment":"The invariant involves a frequency cutoff C, but the paper does not report how N_+ and N_- depend on C. The text says the contour is the complex upper half plane, while the SI integrates along [−C,C] with a semicircular closure; the resulting winding number is a function of C unless convergence is demonstrated. Since the difference between allowed and forbidden pathways is the whole claim, the cutoff dependence should be checked and stated.","section":"Eq. (2) and SI Sec. I.A"},{"comment":"The sentence 'The topological invariant counts the number of zeros and poles on the real line contained within the plane' is imprecise: a winding number along a contour in the upper half-plane counts zeros minus poles enclosed by the contour, not simply those on the real line. This becomes relevant when the charge-conservation statement ΔN≡0 is invoked, because the relation between the invariant and the electron number requires a fixed frequency origin; the paper should clarify this point given the geometry-dependent μ(R).","section":"Eq. (2) and surrounding text"}],"minor_comments":[{"comment":"The caption refers to 'multireference' Green's functions but the main text distinguishes 'no interactions' and 'multireference'; a phrase such as 'with interactions' would be clearer for readers not familiar with the terminology.","section":"Fig. 2 caption"},{"comment":"The notation 'ΔN_+− ΔN_- ≠ 0' would benefit from an explicit definition of what 'switching of poles (zeros) between two different molecular geometries' means quantitatively, e.g., a statement that the pole/zero that crosses ω=0 changes symmetry sector.","section":"Sec. II.D, Eq. (3)"},{"comment":"The sentence 'It is therefore sufficient to compute the change of one block' should specify whether this sufficiency relies on the exact relation ΔN_+ + ΔN_- = 0, which is only exact in the limit of a converged contour; this qualification would prevent misinterpretation.","section":"Sec. II.D"},{"comment":"The phrase 'poles are replaces by zeros' in the paragraph before Fig. 3 contains a typo; it should read 'poles are replaced by zeros'.","section":"Sec. I"}],"recommendation":"major_revision","confidential_remarks":"The paper is a good candidate for the journal if the authors add the robustness checks suggested in the major comments. The main concern is not the central idea but whether the zero-crossing classification survives changes in the chemical-potential reference and in the active-space size; both are standard checks and should be achievable within the manuscript's scope. I would also encourage the authors to clarify the relationship between their geometry-dependent μ(R) and the Luttinger-type charge-conservation argument, since this is the most conceptually delicate point for the target readership."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know about this paper is that it takes the Green's function pole-zero invariant from correlated topological matter (Gurarie; Zhou-Liu) and applies it to orbital-symmetry-controlled reactions. The clean result is that the symmetry-forbidden disrotatory butadiene→cyclobutene path shows a zero of the Green's function crossing ω=0 near the transition state, with ΔN_+=+1 and ΔN_-=-1 in the symmetry-resolved invariants, while the allowed conrotatory path shows nothing. That is a new way to encode Woodward-Hoffmann character when molecular orbital theory breaks down, and it deserves to be taken seriously.\n\nThe paper does several things well. The CASSCF/TRIQS calculations are standard, and the SI includes Cartesian coordinates for all stationary points, so the numbers are in principle reproducible. The symmetry-resolved winding number is explained clearly, and the use of the Woodward-Hoffmann rules as an external benchmark is sensible. The discussion of how static correlations replace pole crossings with zero crossings in molecules like cyclobutadiene is pedagogically nice and connects to a real physical phenomenon.\n\nThe main soft spot is exactly what the stress-test note flags: the zero of frequency is set by a geometry-dependent chemical potential μ(R)=-(IP+EA)/2. All the pole and zero crossings are defined relative to that moving reference. If you used a fixed μ or a different value inside the gap, the number and location of zero crossings at ω=0 could change, and the claimed ΔN_± distinction might weaken or disappear. The paper does not scan μ, does not compare with a fixed chemical potential, and does not track zero trajectories in an absolute energy frame. This is a legitimate concern, not a manufactured one. It is not necessarily fatal—there are reasonable arguments that the molecular chemical potential is the natural zero for particle addition/removal—but the robustness claim is unestablished.\n\nThe second soft spot is numerical: the main results use a CASSCF(4,4) active space without any convergence check against larger active spaces or basis sets. For butadiene that might be adequate, but the paper's central claim is that the zero-crossing diagnostic is method-independent and general. The asymmetric photoswitch example is suggestive, but it only shows a zero crossing—no invariants are computed there, so the generalization beyond symmetry is not demonstrated. No code or raw data are shipped, which makes independent verification slower.\n\nAll that said, the central idea is plausible and the demonstration is clean enough to warrant a serious referee. The right outcome is a major-revision request: test the chemical-potential dependence, add at least one larger active-space check, and either compute invariants for the photoswitch or state explicitly that it is only an illustration. If those tests hold, this could become a useful tool for thinking about correlated reaction paths.\n\nI would bring this to a reading group. It is not a desk reject.","headline":"A genuinely new application of the many-body pole-zero winding invariant to organic reactions, with a clean demonstration on butadiene electrocyclization, but the geometry-dependent chemical potential gauge is untested and could shift the claimed zero crossings.","tokens_in":18782,"tokens_out":2672,"would_cite":true,"duration_ms":32915,"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":"The paper claims that symmetry-forbidden electrocyclizations are marked by Green's-function zeros crossing zero frequency, while allowed paths show no such crossing, and a winding-number invariant classifies the reaction.","keywords":["Green's function zeros","topological invariant","orbital-symmetry selection rules","electrocyclization","strong electronic correlations","reaction coordinate","CASSCF"],"falsifier":"Recompute the symmetry-resolved Green's functions for the disrotatory pathway with the chemical potential shifted by a small constant and with a larger active space and basis set, and follow the zeros near $\\omega=0$; if the crossing geometry shifts or the invariant changes $\\Delta N_+=+1$ and $\\Delta N_-=-1$ are not reproduced, the zero-crossing diagnostic is an artifact of the chosen reference and level of theory rather than a stable topological classification.","tokens_in":17804,"feed_emoji":"⚛️","tokens_out":17451,"duration_ms":160940,"temperature":0.7,"pith_summary":"Orbital-symmetry selection rules classify a reaction as allowed or forbidden by whether occupied and unoccupied molecular orbitals cross along the reaction path, but those orbitals are ill-defined precisely when the reaction is forbidden and correlations are strong. This paper claims that the distinction survives in the many-body Green's function: for the $4\\pi$ electrocyclization of butadiene to cyclobutene, a symmetry-forbidden disrotatory pathway shows zeros crossing $\\omega=0$ before the transition state, with symmetry-resolved invariants changing by $\\Delta N_+=+1$ and $\\Delta N_-=-1$, while the symmetry-allowed conrotatory pathway shows no pole or zero crossing and $\\Delta N_+=0$. The classifying object is $N(R)=\\oint \\frac{d\\omega}{2\\pi i}\\,\\partial_\\omega \\ln \\det G(\\omega|R)$, a winding number computed separately in each symmetry sector. The result matters because single-reference and molecular-orbital pictures fail at the near-degeneracies that define forbidden reactions; Green's-function zeros restore a well-defined topological signal.","feed_headline":"Zero crossings reveal which ring closures are forbidden","feed_subtitle":"A Green's-function invariant recovers the orbital-symmetry rule when correlations break molecular-orbital theory.","key_machinery":"The load-bearing object is the many-body Green's function $G(\\omega|R)$ for a fixed molecular geometry $R$, defined through the spectral representation over states with $N$, $N+1$, and $N-1$ electrons. Its determinant defines the invariant $N(R)=\\oint_C \\frac{d\\omega}{2\\pi i}\\,\\partial_\\omega \\ln \\det G(\\omega|R)$, a winding number whose value changes by $\\pm1$ whenever a pole or zero of the Green's function crosses $\\omega=0$; zeros appear where the self-energy diverges and encode static correlation. Spatial symmetry block-diagonalizes $G$ into $G_+\\oplus G_-$, giving invariants $N_+(R)$ and $N_-(R)$ that can only change at a pole or zero crossing at $\\omega=0$. The crossing of zeros at $\\omega=0$ is therefore the interacting counterpart of the HOMO-LUMO level crossing of molecular-orbital theory.","core_discovery":"On the paper's own terms, the central discovery is that the orbital crossing predicted by one-electron theory for a symmetry-forbidden reaction is replaced, in a correlated Green's function, by a crossing of Green's function zeros at $\\omega=0$, while the allowed path has neither poles nor zeros crossing $\\omega=0$. The zeros appear where the natural-orbital occupation numbers of the HOMO and LUMO become degenerate, and they are protected by the same spatial symmetry that protected the molecular-orbital crossing. Symmetry block-diagonalizes the Green's function as $G_+\\oplus G_-$; the winding invariants $N_+(R)$ and $N_-(R)$ change by $\\Delta N_+=+1$ and $\\Delta N_-=-1$ across the forbidden reaction and by $\\Delta N_+=0$ across the allowed one, so comparing reactants and products is sufficient to classify the reaction. A substituted planar $4\\pi$ photoswitch with no spatial symmetry still shows the zero crossing, indicating that static correlation can dominate over symmetry-breaking in such systems.","pith_inferences":["A natural extension the paper does not make is to search for a chemical-potential-independent invariant, because the zero crossings are defined relative to $\\mu(R)=-(IP+EA)/2$; the current data do not test whether a shifted reference would preserve the classification.","If the zero-crossing geometry coincides with degeneracy of the HOMO and LUMO natural-orbital occupation numbers, those occupation numbers alone could serve as a cheaper diagnostic for forbiddenness in larger systems; the paper reports the coincidence but does not propose it as a criterion.","The survival of the zero crossing in the substituted photoswitch suggests the classification may remain meaningful for asymmetric reactions when static correlation dominates the symmetry-breaking scale, a scale competition the paper does not quantify.","A nonzero invariant difference between reactant and product implies the two endpoints are not adiabatically connected in the correlated sense, which hints at a link between this classification and nonadiabatic reaction dynamics."],"forward_implications":["Reactant and product geometries alone suffice: computing $N_+(R)$ and $N_-(R)$ at the two endpoints yields $\\Delta N_+=+1$ and $\\Delta N_-=-1$ for the forbidden path and $\\Delta N_+=0$ for the allowed path, so a full scan of the reaction coordinate is unnecessary for classification.","A nonzero value of $\\Delta N_+ - \\Delta N_-$ forces at least one pole or zero to cross $\\omega=0$ between the two geometries, so a symmetry-forbidden reaction necessarily passes through a geometry where one symmetry sector loses its gapped single-particle description.","The framework replaces molecular-orbital correlation diagrams with Green's-function zeros, so it applies to strongly correlated reactions where single-reference methods and well-defined molecular orbitals do not exist.","The same invariant construction can be transferred to spin-resolved or excited-state reactions by treating spin or state sectors as the symmetry blocks."],"supporting_citations":[{"why":"Supplies the Green's-function winding-number invariant that counts zeros and poles, the central diagnostic of the paper.","marker":"70"},{"why":"Second source for the pole-and-zero winding invariant, cited alongside reference 70 for equation (2).","marker":"90"},{"why":"Establishes that the forbidden pathway's orbital crossing is symmetry-protected because no symmetry-allowed term can gap it out.","marker":"41"},{"why":"Original statement of the orbital-symmetry selection rules for pericyclic reactions that this work generalizes.","marker":"38"},{"why":"Shows that correlations turn the molecular-orbital crossing into an avoided crossing of many-electron states, motivating the Green's-function analysis.","marker":"49–51"},{"why":"Provides the substituted planar 4π photoswitch used as the symmetry-broken test case where the zero crossing persists.","marker":"97"}],"fun_headline_variants":["Green's function zeros reveal forbidden electrocyclizations","Correlated theory: forbidden reactions show zero crossings","Topological invariant for symmetry-controlled reactions","Orbital-symmetry rules hold via Green's function zeros","Zero crossings in correlated Green's functions flag forbidden paths"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The classification fixes the zero of frequency by a geometry-dependent chemical potential $\\mu(R)=-(IP+EA)/2$ (the negative average of the ionization potential and electron affinity); if a different chemical potential or reservoir were used, the number and location of $\\omega=0$ zero crossings could change, and the paper does not test this gauge dependence or convergence with active-space size or basis set.","fun_headline_variants_meta":{"raw":{"variants":["Green's function zeros reveal forbidden electrocyclizations","Correlated theory: forbidden reactions show zero crossings","Topological invariant for symmetry-controlled reactions","Orbital-symmetry rules hold via Green's function zeros","Zero crossings in correlated Green's functions flag forbidden paths"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000408,"raw_usage":{"total_tokens":2128,"prompt_tokens":966,"completion_tokens":1162,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":582,"completion_tokens_details":{"reasoning_tokens":1088}},"tokens_in":582,"tokens_out":1162,"duration_ms":12659,"temperature":1.0,"reasoning_tokens":1088,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T18:41:19.583539+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the symmetry-resolved Green's functions for the disrotatory pathway with the chemical potential shifted by a small constant and with a larger active space and basis set, and follow the zeros near $\\omega=0$; if the crossing geometry shifts or the invariant changes $\\Delta N_+=+1$ and $\\Delta N_-=-1$ are not reproduced, the zero-crossing diagnostic is an artifact of the chosen reference and level of theory rather than a stable topological classification.","supporting_citations":[{"cited_title":"Zhou \\ and\\ author J","cited_arxiv_id":null,"evidence_quote":"Second source for the pole-and-zero winding invariant, cited alongside reference 70 for equation (2)."},{"cited_title":"Muechler ,\\ 10.1103/PhysRevB.101.045123 journal journal Phys","cited_arxiv_id":null,"evidence_quote":"Establishes that the forbidden pathway's orbital crossing is symmetry-protected because no symmetry-allowed term can gap it out."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Original statement of the orbital-symmetry selection rules for pericyclic reactions that this work generalizes."},{"cited_title":"Mirzanejad \\ and\\ author L","cited_arxiv_id":null,"evidence_quote":"Provides the substituted planar 4π photoswitch used as the symmetry-broken test case where the zero crossing persists."}],"review_version":1}