{"id":"c89655fe-cf9a-40dc-9d14-c26902a879c1","arxiv_id":"2505.00499","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"The paper implements relativistic two-component DIP-EOMCCSD/SDT with the Dirac-Coulomb-Breit Hamiltonian and shows that a composite scheme reproduces noble-gas double ionization potentials to better than 0.03 eV.","lead":"This paper implements a relativistic coupled-cluster method for calculating double ionization energies of heavy atoms and molecules, and tests it against full four-component calculations and experiment. A two-step composite scheme can match measured double ionization potentials of noble gas atoms to better than 0.03 eV, though part of that agreement comes from a known cancellation of errors.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Composite scheme in Eq. 9 rests on a basis-set convergence assumption that the paper's own Table III contradicts.","rationale":"The reader's weakest_assumption was the additivity/basis-convergence of the SDT-SD correction in Eq. 9, backed by the Table III basis-size dependence. I agree, and this is the most load-bearing concern: it directly undermines the headline quantitative achievement (sub-0.02 eV experimental agreement for Xe), not merely an implementation detail. Table III explicitly documents that +SDT(DCB/DZ) differs from +SDT(DCB/TZ) by 0.1-0.2 eV for Ar and Kr, and the +SDT(NR/full) points are 0.14-0.30 eV below experiment. The paper's own text calls the DZ result a fortunate cancellation of errors. The implementation claim (mmfX2C vs 4c agreement, Table I) is well-supported by direct comparison and is not under attack. The composite protocol, however, is presented as the main route to experiment, so a CONDITIONAL verdict is appropriate: the method is validated internally, but the experimental-accuracy claim is unsupported without a convergence test of the SDT-SD correction. This matches the reader's verdict, so no change is needed.","tokens_in":18952,"tokens_out":2261,"duration_ms":17247,"concrete_test":"Evaluate Eq. 9 with +SDT(DCB/TZ) or a larger-basis SDT correction for Xe (e.g., ANO-RCC-VTZP or VQZP if feasible) and compare to the DZ correction; if the Xe composite values shift by more than 0.1 eV, the advertised sub-0.02 eV agreement is not reproducible as a method statement. A cheaper check is to recalculate Ar and Kr with +SDT(DCB/TZ) but with the same frozen-core and virtual-space settings as the +SDT(DCB/DZ) data, and verify the correction vs basis-size trend. Alternatively, compute +SDT(DCB/DZ) in a second small basis (e.g., dyall.acvdz or x2c-SVPall-2c) and confirm whether the DZ correction is basis-family dependent.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central quantitative claim is the composite Eq. 9 agreement with experiment. Eq. 9 adds to a large-basis (full ANO-RCC) DCB-X2C-DIP-EOMCCSD value a correction [DIP(SDT) - DIP(SD)] evaluated in a small basis. This correction is implicitly assumed to be basis-size independent. Table III shows it is not: for Ar, +SDT(DCB/DZ) errors are -0.001 to +0.090 eV while +SDT(DCB/TZ) errors are -0.135 to -0.068 eV, a spread of roughly 0.1-0.2 eV. For Kr the analogous shift is -0.036/-0.047 to -0.083/-0.159 eV. The +SDT(NR/full) results are even more negative, -0.135 to -0.298 eV. The authors acknowledge the DZ correction is \"a fortunate cancellation of errors\" and that the correction is \"not converged with respect to the basis set size,\" yet the paper still highlights Xe errors below 0.02 eV as a success of the composite approach. Because the correction term swings by 0.1-0.2 eV depending on which basis it is evaluated in, the agreement to 0.02-0.09 eV is not a robust property of the method; it is a cancellation that could disappear for other systems, basis families, or with frozen-core choices. The claim that \"remaining high-order correlation effects are approximately modeled via small-basis ... SDT\" is therefore not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports an implementation of relativistic double-ionization-potential equation-of-motion coupled-cluster (DIP-EOMCCSD and DIP-EOMCCSDT, including 4h2p excitations) in the one-electron and molecular mean-field exact two-component (1eX2C and mmfX2C) frameworks, using Dirac-Coulomb, Dirac-Coulomb-Gaunt, and Dirac-Coulomb-Breit Hamiltonians. The central validation is a direct comparison with four-component DIP-EOMCCSD from the literature, which agrees to within 0.003 eV. The authors then study basis-set convergence of double IPs for noble gases and small molecules, find that large-basis DCB-X2C-DIP-EOMCCSD overestimates experiment, and propose a composite scheme in Eq. (9) that combines a large-basis SD result with a small-basis SDT-SD correction for Ar, Kr, and Xe. They report good agreement for the DZ-based correction but also explicitly note that this agreement relies on a cancellation of errors and that the correction is not basis-set converged. The abstract and conclusions highlight the composite results while acknowledging the poor convergence of the ANO-RCC basis family.","tokens_in":19456,"tokens_out":4074,"duration_ms":43704,"significance":"If the implementation is correct, the work is a useful methodological advance: it extends relativistic two-component coupled-cluster treatments to double ionization potentials with high-order correlation effects, and the machine-generated equations plus numerical checks against CCPy provide a concrete reproducibility basis. The direct mmfX2C versus 4c benchmark in Table I is a strong, non-circular validation and is the most valuable part of the paper. The composite-scheme accuracy claim, however, is not robust because it depends on a basis-set cancellation that the paper itself documents, so the significance of the composite results should be read as much more limited than the abstract suggests.","major_comments":[{"comment":"The composite scheme assumes that the SDT-SD difference computed in a small basis is a valid estimate of the missing high-order correlation effect at the large-basis level, but Table III shows this difference is not converged with basis size. For Ar, the +SDT(DCB/DZ) errors range from -0.001 to +0.090 eV, whereas the +SDT(DCB/TZ) errors range from -0.135 to -0.068 eV and the +SDT(NR/full) errors from -0.182 to -0.135 eV; Kr and Xe show similarly large swings. The small errors of the DZ-based composite are therefore a cancellation of errors, as the authors acknowledge in the text. The claim that Eq. (9) brings double IP values into excellent agreement with experiment is not established as a robust property of the method. Please either compute the correction in a sufficiently converged basis and report the residual basis error, or substantially soften the composite-scheme claim and present it as an empirical cancellation rather than a validated composite protocol.","section":"§IV, Eq. (9), Table III"},{"comment":"The statement that DCB-X2C-DIP-EOMCCSD tends to overestimate double IP values in the complete-basis limit by more than 0.25 eV on average is inferred from a single basis family, the full ANO-RCC basis, whose convergence the paper itself flags as poor. The non-relativistic +SDT(NR/full) data still show a 0.1-0.2 eV residual for Ar and Kr, and the x2c-type and Dyall families are only carried to triple-zeta quality, so the true CBS limit is not established by the data shown. The conclusion should either be restricted to the ANO-RCC family or supported by explicit extrapolations from more than one basis-set family.","section":"§IV, Fig. 1 and Conclusions"},{"comment":"For the molecular systems in Fig. 1, the calculated values are vertical electronic DIPs while the experimental references include zero-point vibrational energy differences and possible adiabatic/vertical distinctions; the paper lists these offsets (e.g., -0.123 eV for HBr) but does not include them in the plotted errors. The mean absolute errors quoted from Fig. 1 are therefore not direct electronic-structure errors for the diatomics. Please either apply the stated ZPVE corrections when reporting errors against experiment or explicitly exclude the diatomics from the quantitative MAE comparisons.","section":"§IV, Fig. 1 and comparison to experiment"}],"minor_comments":[{"comment":"The abstract states agreement with four-component calculations to within 0.001 eV, while Table I reports discrepancies up to 0.003 eV; these numbers should be made consistent.","section":"Abstract and Table I"},{"comment":"The sentence 'double IP values from are consistently underestimated' is missing the method label; it should read 'double IP values from +SDT(DCB/TZ) are consistently underestimated' or similar.","section":"§IV, near Table III"},{"comment":"The text frequently refers to the 'full ANO-RCC basis set' without a formal definition; please state explicitly what this set is (e.g., number of contracted functions or the ANO-RCC-VQZP plus additional functions) so the reader can reproduce the calculations.","section":"§III and §IV"},{"comment":"Several entries in the +Gaunt 4c column are blank or marked with a dash; please clarify whether those calculations were not performed or were omitted for other reasons.","section":"Table I"},{"comment":"The sentence 'the estimates computed using the non-relativistic DIP-EOMCC approach at the large basis set limit indicates' has a subject-verb agreement error and should be rephrased.","section":"Conclusions"}],"recommendation":"major_revision","confidential_remarks":"The direct 2c/4c benchmark in Table I is strong and the implementation appears sound; the main risk is overclaiming the composite scheme, since the authors' own Table III shows a 0.1-0.2 eV basis dependence in the correction term. If the authors reframe the composite results as an empirical cancellation and add explicit uncertainty estimates, the paper would be suitable for publication in this journal. The manuscript fits the journal's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, the implementation advance is genuine: DIP-EOMCCSDT with 4h2p excitations in the mmfX2C two-component framework with the Dirac–Coulomb–Breit Hamiltonian is new, and the direct benchmark against 4c DIP-EOMCCSD (Table I, max deviation 0.003 eV) is clean and convincing. That part holds up. Second, the composite scheme of Eq. 9 — the paper's route to experiment for Ar, Kr, Xe — rests on an additivity assumption that the paper's own Table III disproves. The SDT-minus-SD correction swings by roughly 0.1–0.2 eV depending on whether it is evaluated in double- or triple-zeta basis, and the authors say as much: the DZ correction is \"a fortunate cancellation of errors\" and not converged. So the sub-0.02 eV Xe numbers are not a robust property of the method; they are a basis-set- and system-dependent cancellation.\n\nThat said, the paper is honest about this in the text. It flags the ANO-RCC convergence problem and the 0.1–0.2 eV non-relativistic large-basis error. The abstract and conclusions, though, still lead with the Xe agreement, which overstates what the data establish.\n\nSoft spots beyond the composite scheme: the CBS limit is inferred almost entirely from a single basis family (ANO-RCC) that the paper itself shows is poorly converged; no code and no data are released beyond \"available upon request\"; and the authors do not reconcile the 0.1–0.2 eV error at the large basis limit in the non-relativistic calculations. These are addressable, and none of them undercut the core implementation and the 4c validation.\n\nWho is this for? People building relativistic EOMCC codes or needing double IPs of heavy elements with Breit and triples-level correlation. For that reader, the 4c cross-check is the main takeaway. The composite scheme should be treated as an illustration, not a recommended protocol.\n\nRecommendation: send it to peer review. The implementation claim deserves scrutiny and publication, and the paper is honest enough that a referee can work with it. The authors should be pushed to temper the composite-scheme language and, ideally, to test the correction in at least one additional basis family or system before presenting it as a predictive scheme.","headline":"Genuine implementation advance with a clean 4c validation; the composite correction for experiment relies on an error cancellation the paper itself documents.","tokens_in":19881,"tokens_out":2294,"would_cite":true,"duration_ms":21325,"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":"Two-component method matches 4c double ionization potentials to 0.003 eV.","keywords":["double ionization potential","equation-of-motion coupled cluster","exact two-component (X2C)","Dirac-Coulomb-Breit Hamiltonian","relativistic quantum chemistry","noble gas dications","composite correlation correction","coupled cluster triples"],"falsifier":"Compute the +SDT(DCB/TZ) correction for xenon, which the paper could not afford; if it lands near −0.1 eV like argon and krypton instead of near zero, the DZ-based composite agreement for xenon is coincidence. Alternatively, apply the composite scheme to a diatomic such as Cl2 and compare with vibrationally resolved experimental double IPs, where no error cancellation has been demonstrated.","tokens_in":18782,"feed_emoji":"⚛️","tokens_out":6100,"duration_ms":55025,"temperature":0.7,"pith_summary":"This paper implements double-ionization-potential equation-of-motion coupled-cluster (DIP-EOMCC) with up to 4-hole–2-particle excitations in a two-component relativistic framework built on the molecular mean-field exact two-component (mmfX2C) transformation. It seeks to show that mean-field treatment of two-electron relativistic effects is enough: for Dirac–Coulomb and Dirac–Coulomb–Gaunt Hamiltonians, mmfX2C-DIP-EOMCCSD reproduces full four-component double IPs to within 0.003 eV. Against experiment, plain DCB-X2C-DIP-EOMCCSD overshoots noble-gas double IPs by 0.1–0.4 eV in large bases, so the paper adds a composite correction that estimates missing higher-order correlation from a small-basis SDT-minus-SD difference. With that correction, Ar, Kr, and Xe double IPs land within 0.09 eV (often 0.02 eV) of experiment. The result matters because it points to a cheaper route to accurate dication energetics without sacrificing relativistic fidelity.","feed_headline":"Two-component method matches 4c double IPs to 0.003 eV","feed_subtitle":"A composite triples correction brings noble-gas double ionization potentials to within 0.09 eV of experiment.","key_machinery":"The central objects are the mmfX2C transformation, a unitary fold-down of the four-component Dirac picture built from molecular mean-field spinors, and the DIP-EOMCC operators that remove two electrons with up to 3-hole–1-particle (SD) or 4-hole–2-particle (SDT) excitations. The load-bearing identity is the composite additivity of Eq. 9, DIP = large-basis SD value + (small-basis SDT minus small-basis SD), which transfers a higher-order correlation correction computed in a small basis to a large-basis result.","core_discovery":"The central claim is that mmfX2C-DIP-EOMCCSD is quantitatively equivalent to full 4c-DIP-EOMCCSD for double ionization potentials of noble gases and small diatomics, with the largest deviation being 0.003 eV for the Dirac–Coulomb Hamiltonian and Gaunt-term shifts agreeing to 0.001 eV or better. At the Dirac–Coulomb–Breit level, the method systematically overestimates double IPs in the complete-basis limit by more than 0.25 eV on average, a deficit the paper attributes to missing higher-order correlation rather than to the relativistic treatment. To close the gap, it constructs a composite estimate, Eq. 9, that adds to a large-basis DCB-X2C-DIP-EOMCCSD double IP the difference between DIP-EOMCCSDT and DIP-EOMCCSD computed in a small basis. For Ar, Kr, and Xe this composite brings double IPs to within 0.09 eV of experiment, and below 0.02 eV for Xe. The paper is explicit that the double-zeta correction is not basis-set converged, and that the agreement relies partly on error cancellation.","pith_inferences":["If the additivity holds, the composite recipe could extend to other properties of heavy-element dications, but each new system needs its own check that the small-basis correlation correction is size-converged.","A testable expectation: for molecules and for elements beyond Xe, the +SDT(DCB/DZ) correction will tend to over-correct and underestimate double IPs the way +SDT(DCB/TZ) does for Ar and Kr, because the double-zeta basis lacks the flexibility to describe the triples effect fully.","The basis-family dependence seen here suggests part of the residual 0.1–0.3 eV error in composite schemes is an artifact of basis contraction and recontraction; repeating the analysis with an uncontracted basis would isolate that contribution.","The success of mean-field Breit treatment for double IPs hints that mmfX2C-DIP-EOMCC could also describe spin-orbit splittings in p4 and d8 configurations, where DIP-EOMCC is a natural tool."],"forward_implications":["mmfX2C-DIP-EOMCCSD reproduces 4c-DIP-EOMCCSD double IPs to within 0.003 eV for the systems studied, so four-component calculations are not needed for these properties.","The Gaunt term lowers double IPs by 0.004–0.043 eV and the gauge term raises them by 0.001–0.005 eV, so both two-electron relativistic terms are small but not negligible at the 0.01 eV scale.","DCB-X2C-DIP-EOMCCSD in the complete-basis limit overestimates double IPs by more than 0.25 eV on average; missing triples correlation, not the relativistic treatment, causes the overshoot.","The Eq. 9 composite brings Ar, Kr, and Xe double IPs to within 0.09 eV of experiment, and to under 0.02 eV for Xe.","The ANO-RCC basis family converges poorly for DCB-X2C-DIP-EOMCC, with non-relativistic large-basis results implying a residual 0.1–0.2 eV error."],"supporting_citations":[{"why":"Supplies the four-component DIP-EOM-CCSD benchmark values used in the direct comparison of Table I.","marker":"[87]"},{"why":"Introduces the Dirac–Coulomb–Breit molecular mean-field X2C equation-of-motion coupled-cluster framework that this work extends to double ionization.","marker":"[78]"},{"why":"Defines the molecular mean-field exact two-component approach that is the foundation of the mmfX2C transformation.","marker":"[61]"},{"why":"Provides the four-component DC/DCG equation-of-motion coupled-cluster methodology that motivates the two-component treatment.","marker":"[70]"},{"why":"Supplies the non-relativistic DIP-EOMCCSDT reference implementation and the 4-hole–2-particle DIP-EOMCC formalism used for testing and composite corrections.","marker":"[50]"},{"why":"Provides the experimental double ionization potentials used as the accuracy benchmark for the composite scheme.","marker":"[106]"}],"fun_headline_variants":["Two-component method matches 4c double IPs to 0.003 eV","Composite triples fix noble-gas double IPs to <0.02 eV","Relativistic two-component DIP-EOMCCSD rivals 4c to 0.003 eV","Xe double IP within 0.02 eV via composite DCB-X2C scheme","Mean-field two-component DIP-EOMCCSD matches 4c for noble gases"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The additivity assumption that the SDT-minus-SD difference computed in a small basis is a faithful estimate of the missing triples correlation at the large-basis level.","fun_headline_variants_meta":{"raw":{"variants":["Two-component method matches 4c double IPs to 0.003 eV","Composite triples fix noble-gas double IPs to <0.02 eV","Relativistic two-component DIP-EOMCCSD rivals 4c to 0.003 eV","Xe double IP within 0.02 eV via composite DCB-X2C scheme","Mean-field two-component DIP-EOMCCSD matches 4c for noble gases"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000366,"raw_usage":{"total_tokens":2058,"prompt_tokens":1128,"completion_tokens":930,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":744,"completion_tokens_details":{"reasoning_tokens":817}},"tokens_in":744,"tokens_out":930,"duration_ms":8218,"temperature":1.0,"reasoning_tokens":817,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:40:13.630472+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the +SDT(DCB/TZ) correction for xenon, which the paper could not afford; if it lands near −0.1 eV like argon and krypton instead of near zero, the DZ-based composite agreement for xenon is coincidence. Alternatively, apply the composite scheme to a diatomic such as Cl2 and compare with vibrationally resolved experimental double IPs, where no error cancellation has been demonstrated.","supporting_citations":[{"cited_title":"Pathak , author S","cited_arxiv_id":null,"evidence_quote":"Supplies the four-component DIP-EOM-CCSD benchmark values used in the direct comparison of Table I."},{"cited_title":"Zhang , author S","cited_arxiv_id":null,"evidence_quote":"Introduces the Dirac–Coulomb–Breit molecular mean-field X2C equation-of-motion coupled-cluster framework that this work extends to double ionization."}],"review_version":1}