{"id":"fa6e49e9-0014-4dcf-bdad-53cf16aa0b83","arxiv_id":"2501.06632","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper presents an analytic-derivative formulation and implementation of MP2 atomic axial tensors for VCD, validated against finite-difference results and used for (S)-methyloxirane spectra.","lead":"Researchers derived and implemented the first analytic-gradient way to compute vibrational circular dichroism spectra at the MP2 level of theory, replacing a slower numerical-differentiation approach. The new method has roughly O(N^6) scaling, and they used it to produce the first fully analytic MP2 VCD spectrum for (S)-methyloxirane.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The analytic MP2 AAT is validated only against the authors' own finite-difference MagPy code on one molecule/basis; a shared error in the common CPHF/normalization layer would leave Eq. (31) wrong while preserving the 1e-7 agreement.","rationale":"The paper's derivation and implementation are credible: the working equation is explicit, no parameters are fitted, the finite-difference cross-check is a legitimate first test, and the O(N^6) scaling of a single AAT element is reasonable. The Reader's CONDITIONAL verdict is appropriate. The single most load-bearing concern is not an internal inconsistency in the algebra, but the absence of an independent validation for the correlated part of the AAT: the only comparison is against a same-group finite-difference code sharing the same integral layer, for one molecule and one basis set. A systematic error in the magnetic-field CPHF, in the required canonical-orbital convention for dependent pairs, or in the normalization derivative of Eq. (30) would cancel between the two implementations. The acknowledged lack of GIAOs and the unstated gauge origin add a reproducibility problem for the reported spectra, but that is a limitation rather than a fatal flaw. The recommended check is an independent finite-difference or finite-field MP2 AAT calculation for the same molecule; if it reproduces Table 1, the shared-error concern is resolved and the central claim stands. The verdict should remain CONDITIONAL until that independent check or a stated gauge-origin convention is provided.","tokens_in":14127,"tokens_out":15721,"duration_ms":165210,"concrete_test":"Recompute the (P)-H2O2/6-31G electronic MP2 AAT with an independent finite-difference implementation that does not share MagPy/Psi4's magnetic-field CPHF and MP2 amplitude-response routines, for example by scripting a finite-field complex-SCF/MP2 calculation in PySCF or using another package's property machinery, and compare all 36 AAT elements against Table 1. Discrepancies above roughly 1e-6 a.u. would establish that the shared-error concern is real; agreement at the 1e-7 level would substantially strengthen the evidence for Eq. (31).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on Eq. (31) being the correct analytic MP2 AAT, but the only direct evidence is the comparison in Table 1 between the new analytic code and MagPy's finite-difference procedure for (P)-H2O2 with 6-31G. Both codes are by the same group and use Psi4 for integrals, and the paper states (Section 3) that the comparison requires assuming canonical perturbed HF orbitals and including U^chi_pq for both dependent and independent pairs. A systematic error in the magnetic-field CPHF response, in the treatment of dependent-pair U coefficients, or in the normalization derivative Eq. (30) would appear in both routes and the 1e-7 agreement would not expose it. The finite-difference check is therefore an internal consistency test, not an independent validation of the correlated amplitude-response terms. In addition, the paper explicitly acknowledges that omitting GIAOs makes the rotatory strengths origin-dependent, yet no gauge origin is stated for the (S)-methyloxirane results, so the reported 'fully analytic' spectrum is not uniquely reproducible. These gaps are load-bearing for the correctness and physical meaningfulness of the headline claim, even though the derivation itself is plausible and the O(N^6) per-element scaling is defensible.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript derives an analytic-gradient expression for the MP2 atomic axial tensor in Stephens's VCD formulation. Starting from the intermediately normalized MP2 wave function, the authors use second-quantized derivative determinants and Wick's theorem to obtain Eq. (31), implement it in the open-source apyib package, and validate it against their finite-difference MagPy code for (P)-hydrogen peroxide with the 6-31G basis. They report an O(N^6) scaling for the analytic route, several orders of magnitude below the O(N^11) finite-difference route, and present the first fully analytic MP2 VCD spectra for (S)-methyloxirane with four basis sets under two geometry/Hessian protocols.","tokens_in":14340,"tokens_out":8956,"duration_ms":86529,"significance":"If Eq. (31) is correct, this is a significant methodological advance: it removes the numerical-differentiation bottleneck in correlated VCD and greatly extends the size of molecules and basis sets accessible at the MP2 level. The paper's strengths are its self-contained derivation, the explicit treatment of intermediate versus full normalization, and a machine-implemented comparison that reaches roughly 1e-7 a.u. agreement. The main caveats are that the validation is internal to the authors' own code stack and that the reported spectra are origin-dependent with no stated gauge origin. These caveats do not invalidate the derivation, but they do affect how strongly the central claim can be accepted as published.","major_comments":[{"comment":"The numerical validation is an internal consistency check rather than an independent one: MagPy and apyib are written by the same group and both obtain integrals from Psi4, so the finite-difference and analytic routes share the CPHF response layer, the dependent-pair U treatment mentioned in §3, and the normalization derivative in Eq. (30). The 1e-7 to 1e-9 a.u. agreement therefore does not exclude a systematic error common to both codes. Because Eq. (31) is the central claim, I ask for validation against an independent implementation or, at minimum, additional tabulated cases with different molecule/basis combinations and a demonstration that the magnetic-field CPHF response and normalization derivative are correct in isolation (for example, by comparing the HF limit of the analytic AAT against an independent HF VCD code).","section":"§4.1, Table 1"},{"comment":"The rotatory strengths and spectra are origin-dependent because GIAOs are not used, as the authors acknowledge in §5, yet no gauge origin is reported for any calculation. Without this information, the numerical values in Tables 2 and 3 and the spectra in Figs. 1 and 2 are not uniquely reproducible, and the physical interpretation of basis-set sign changes is incomplete. Please state the gauge origin used, quantify the origin dependence (for example, by comparing AATs and rotatory strengths at the molecular center of mass and at another origin), or provide an origin-invariant formulation.","section":"§4.2, Tables 2–3; §5"},{"comment":"The text states that 'several small molecular test cases' were compared, but Table 1 reports only (P)-H2O2 with the 6-31G basis. If additional validation data exist, they should be included in the main text or the SI; as written, the reader cannot assess whether the agreement is robust across basis sets and molecules. This is especially important given that the only comparison is to the authors' own finite-difference code.","section":"§3"}],"minor_comments":[{"comment":"The phrase 'hydrogen molecule dimer' is unusual; presumably 'H2 dimer' or 'hydrogen molecule' is intended.","section":"§3"},{"comment":"Operator strings such as 'a†aχaa' are hard to parse; explicit brackets or a short sentence defining the action of the core-derivative creation operators would improve readability.","section":"§2, Eq. (9)"},{"comment":"Several rotatory strengths contain stray spaces (for example, '2 .495' and '3 .302'); the formatting should be corrected.","section":"Tables 2–3"},{"comment":"Reference 34 gives the page range as '1–25' and should include the volume and article number or issue information as appropriate for Molecular Physics.","section":"References"},{"comment":"The SI is said to contain only Cartesian coordinates; adding the full set of validation AATs, input keywords, and convergence settings would substantially improve reproducibility.","section":"Supporting Information"}],"recommendation":"major_revision","confidential_remarks":"The main gate for this paper is validation. The 1e-7 agreement between two codes from the same group sharing the Psi4 integral layer is strong evidence against implementation bugs but weaker evidence for the physical correctness of the magnetic-field response terms. An independent benchmark, or at least additional tabulated validation cases, would significantly de-risk acceptance. The gauge-origin omission is a straightforward and important fix to make before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague — this is a real methods paper. The authors derive the first analytic MP2 atomic axial tensor for VCD, Eq. (31), implement it in apyib, and show agreement with their own finite-difference code at the 1e-7 a.u. level. The O(N^6) scaling versus O(N^11) for finite difference is defensible, and the workload comparison is dramatic. The (S)-methyloxirane spectra are a useful first look at basis-set sensitivity at MP2, and the paper is honest about the GIAO limitation.\n\nWhat is genuinely new: no prior analytic MP2 AAT formulation exists in the cited literature; the derivation follows established second-quantized algebra and the normalization treatment is explicit. Credit where due: the algebra is not a black box. They show enough of the Wick contractions and the final expression for a careful reader to re-derive it.\n\nSoft spots, in order of weight. First, validation rests entirely on agreement with MagPy, a finite-difference code from the same group sharing Psi4 integrals. The stress-test worry is real: a systematic error in the magnetic-field CPHF response or in the normalization derivative would appear in both routes, and the 1e-7 agreement would not expose it. That said, the derivation is self-contained and the finite-difference protocol is standard; this is an internal consistency check, not an independent validation. A single tabulated molecule/basis is thin but representative. I would not call it fatal, but I would want an independent implementation or at least a second molecule and a GIAO-based comparison at a lower level.\n\nSecond, the lack of GIAOs makes rotatory strengths origin-dependent, and the paper does not state the gauge origin used for the methyloxirane spectra. The authors acknowledge the origin dependence in the conclusion, but without stating the gauge origin the reported values are not uniquely reproducible. That is a defect for a methods paper, easy to fix.\n\nThe derivation is compressed in places—the reduction from Eqs. (12)–(13) to the final overlaps is left to the reader—but the pieces are consistent and the equations are internally coherent. I see no circularity: no parameters are fitted to the target.\n\nWho is this for: anyone computing VCD with correlated methods, and method developers. It deserves serious peer review. My recommendation: send it out, with a request for a stated gauge origin and ideally an independent validation point. The central method is likely correct.","headline":"A genuine first: analytic MP2 AATs for VCD with O(N^6) scaling, but the only validation is same-group internal consistency and the reported spectra lack a stated gauge origin.","tokens_in":14949,"tokens_out":1871,"would_cite":true,"duration_ms":16698,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper presents the first analytic-derivative formulation of MP2 vibrational circular dichroism atomic axial tensors, replacing an O(N^11) finite-difference method with an O(N^6) closed-form expression.","keywords":["vibrational circular dichroism","MP2","atomic axial tensor","analytic derivatives","rotatory strength","chirality","coupled-perturbed Hartree-Fock","O(N^6) scaling"],"falsifier":"Compute the analytic MP2 AAT for a small chiral molecule using an independent implementation with gauge-including atomic orbitals and a different integral package, then compare individual tensor elements to the values reported here; a disagreement beyond about 1e-7 a.u. in any element that cannot be attributed to gauge origin would falsify the working equation or its implementation.","tokens_in":13859,"feed_emoji":"🌀","tokens_out":6970,"duration_ms":63634,"temperature":0.7,"pith_summary":"The paper presents the first analytic-derivative derivation of vibrational circular dichroism (VCD) atomic axial tensors at the second-order Møller-Plesset (MP2) level, replacing the previous finite-difference approach that scaled as O($N^{11}$) with an O($N^{6}$) analytic scheme. The derivation yields a closed-form expression built from coupled-perturbed Hartree-Fock coefficients, half-derivative overlap integrals, and derivatives of the MP2 amplitudes; the authors implement it and show that individual tensor elements agree with their finite-difference results to about 1e-7 a.u. on a test molecule. With this method they compute the first fully analytic MP2 VCD spectrum of (S)-methyloxirane across four basis sets. If correct, the advance makes MP2-level VCD practical for larger chiral molecules and opens the way to analytic chiroptical spectra at higher correlation levels.","feed_headline":"New analytic formula cuts VCD cost from O(N^11) to O(N^6)","feed_subtitle":"MP2-level circular dichroism previously required slow finite differences; now it runs in seconds on small chiral molecules.","key_machinery":"The central object is Eq. (31), the fully normalized analytic MP2 AAT expression, which is built from the first-order MP2 wave function with intermediate normalization and then corrected by the derivative of the normalization factor. The required ingredients are first-order CPHF coefficients for nuclear displacements and magnetic-field perturbations, half-derivative overlap integrals, and derivatives of the T2 amplitudes; the derivation uses Wick's theorem with additional contraction rules for the half-derivative overlaps to reduce the many overlap terms to a compact formula. The cancellation structure—several seemingly contributing terms vanish because single-excitation projections cannot fully contract, and paired terms in Eqs. (22)-(23) cancel—leaves the final expression that the implementation evaluates.","core_discovery":"The paper's central discovery is that the MP2 atomic axial tensor—the part of the VCD rotatory strength arising from the magnetic-dipole transition moment—can be evaluated analytically, without complex arithmetic or non-orthonormal molecular orbital overlaps, by differentiating the intermediately normalized MP2 wave function and then restoring full normalization through a closed-form derivative of the normalization factor. The resulting working expression (Eq. 31 of the paper) collects the Hartree-Fock-level AAT term, a pure amplitude-derivative term, and several coupling terms between CPHF coefficients and amplitude derivatives; all other terms cancel by Wick's theorem contractions. The expression agrees with finite-difference AATs to roughly 1e-7 a.u. for (P)-hydrogen peroxide with the 6-31G basis, and the method is validated as far less expensive: a full MP2 AAT that took hours per tensor element numerically takes seconds analytically. Using this method, the paper reports the first fully analytic MP2 VCD spectrum of (S)-methyloxirane for several basis sets.","pith_inferences":["A natural test is whether the analytic and finite-difference routes share a hidden systematic error; comparing the analytic AATs to an independent implementation that uses gauge-including atomic orbitals would separate origin-dependence effects from the response equations themselves.","The O(N^6) scaling suggests MP2 VCD could be applied to molecules of 20-30 heavy atoms with modest basis sets, a size regime where computed versus measured VCD spectra are often used for assignment; this is a testable prediction about practical applicability.","The same formal machinery could be transferred to coupled-cluster doubles or other response properties, such as Raman optical activity, where analytically differentiated wave functions would yield similar savings.","Because the paper's validation used a single molecule and basis set, a broader benchmark across several chiral molecules with different functional groups would firm up the claimed agreement level."],"forward_implications":["MP2-level VCD spectra can now be computed analytically for molecules and basis sets that were impractical with the previous O(N^11) finite-difference method, which required hours per tensor element.","The analytic route eliminates complex wave-function arithmetic and non-orthonormal molecular-orbital overlaps, removing a major practical barrier in correlated VCD calculations.","The same strategy—differentiate the wave function, restore normalization, and contract with Wick's theorem—can be applied to higher correlation methods, which the paper states as future work.","The computed (S)-methyloxirane rotatory strengths show substantial basis-set dependence, including sign changes for weak modes, so practical MP2 VCD simulations will need carefully chosen basis sets and geometries.","The analytic method provides a new reference standard for validating cheaper approximations to VCD and for benchmarking finite-difference implementations."],"supporting_citations":[{"why":"Defines the atomic axial tensor formulation of VCD rotatory strengths that the paper differentiates.","marker":"[11]"},{"why":"Pioneered analytic-derivative AATs at the Hartree-Fock level; the paper extends this strategy to MP2.","marker":"[14]"},{"why":"Reports the finite-difference MP2/CID AAT method used as the validation reference.","marker":"[20]"},{"why":"Shows gauge-origin-independent VCD at the MCSCF level, establishing the GIAO treatment the paper notes as future work.","marker":"[15]"},{"why":"Demonstrates full VCD simulation with density functional theory including GIAOs, a key comparative standard.","marker":"[19]"},{"why":"Supplies the coupled-perturbed Hartree-Fock equations for nuclear-displacement and magnetic-field perturbations.","marker":"[21]"},{"why":"Gives the Wick's theorem contraction rules used to evaluate overlaps of derivative determinants.","marker":"[22]"},{"why":"Provides the geometries and Hessians used for the (S)-methyloxirane VCD spectra calculations.","marker":"[33]"}],"fun_headline_variants":["MP2 VCD goes analytic, cuts cost to O(N^6)","First analytic MP2 VCD spectrum: cost drops to O(N^6)","Analytic MP2 VCD: O(N^6) beats O(N^11) finite differences","MP2 VCD without numeric derivatives: analytic AATs, O(N^6)","Analytic MP2 AATs make VCD spectra fast: O(N^6)"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The validation of the analytic expression rests on agreement with a finite-difference implementation that shares the same integral code and was tested on a single molecule with one basis set; if both routes contain the same systematic error—for example in the magnetic-field response or in the normalization derivative—the close numerical agreement would not expose it.","fun_headline_variants_meta":{"raw":{"variants":["MP2 VCD goes analytic, cuts cost to O(N^6)","First analytic MP2 VCD spectrum: cost drops to O(N^6)","Analytic MP2 VCD: O(N^6) beats O(N^11) finite differences","MP2 VCD without numeric derivatives: analytic AATs, O(N^6)","Analytic MP2 AATs make VCD spectra fast: O(N^6)"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00086,"raw_usage":{"total_tokens":3693,"prompt_tokens":869,"completion_tokens":2824,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":485,"completion_tokens_details":{"reasoning_tokens":2715}},"tokens_in":485,"tokens_out":2824,"duration_ms":17460,"temperature":1.0,"reasoning_tokens":2715,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:56:08.449640+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the analytic MP2 AAT for a small chiral molecule using an independent implementation with gauge-including atomic orbitals and a different integral package, then compare individual tensor elements to the values reported here; a disagreement beyond about 1e-7 a.u. in any element that cannot be attributed to gauge origin would falsify the working equation or its implementation.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the atomic axial tensor formulation of VCD rotatory strengths that the paper differentiates."},{"cited_title":"D.; Handy, N","cited_arxiv_id":null,"evidence_quote":"Pioneered analytic-derivative AATs at the Hartree-Fock level; the paper extends this strategy to MP2."},{"cited_title":"M.; Crawford, T","cited_arxiv_id":null,"evidence_quote":"Reports the finite-difference MP2/CID AAT method used as the validation reference."},{"cited_title":"L.; J rgensen, P.; Helgaker, T.; Ruud, K.; Jensen, H","cited_arxiv_id":null,"evidence_quote":"Shows gauge-origin-independent VCD at the MCSCF level, establishing the GIAO treatment the paper notes as future work."},{"cited_title":"R.; Frisch, M","cited_arxiv_id":null,"evidence_quote":"Demonstrates full VCD simulation with density functional theory including GIAOs, a key comparative standard."},{"cited_title":"D.; Osamura, Y.; Schaefer, H","cited_arxiv_id":null,"evidence_quote":"Supplies the coupled-perturbed Hartree-Fock equations for nuclear-displacement and magnetic-field perturbations."},{"cited_title":"D.; Schaefer, H","cited_arxiv_id":null,"evidence_quote":"Gives the Wick's theorem contraction rules used to evaluate overlaps of derivative determinants."},{"cited_title":"u n. C . Z hang, X . Z heng, and the integral packages MOLECULE ( J . A lml \\","cited_arxiv_id":null,"evidence_quote":"Provides the geometries and Hessians used for the (S)-methyloxirane VCD spectra calculations."}],"review_version":1}