{"id":"e12738e4-721b-424a-9bbb-6fcafc447fdc","arxiv_id":"2502.17451","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Relativistic coupled-cluster calculations with complete-basis-set extrapolation and higher-order corrections give ionization potentials for ScF, YF, LaF, and AcF with 10 to 16 meV uncertainties.","lead":"This paper calculates the ionization potentials of four metal monofluoride molecules (ScF, YF, LaF, AcF) using high-level relativistic quantum chemistry, with estimated uncertainties of 10 to 16 meV. The results give experimentalists precise targets for measuring these values, particularly for radioactive AcF, a candidate for fundamental symmetry experiments.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Breit correction is taken from atomic M+ ions with no molecular benchmarking; for ScF/YF/LaF it is +7 meV, comparable to the quoted 10-11 meV uncertainties, so an unquantified molecular-atomic difference could shift the central IPs.","rationale":"The conditional verdict is appropriate. This paper is otherwise careful: it uses CBS extrapolation, active-space and augmentation checks, explicit QED corrections, and a documented uncertainty budget, and it builds on a prior CaF/SrF/BaF study. The load-bearing gap is the Breit transferability: it is a systematic effect on the values, not just on the error bars, and it is not benchmarked. The proposed DCB SCF test is the minimal check that could falsify or support the transfer; if it fails, a Breit uncertainty must be added and the LaF and AcF confidence statements would need revision. I considered the 10%-of-triples uncertainty rule and the LaF v4z->cv4z 50 meV basis-set anomaly; both are weaknesses, but the Breit transfer is more directly tied to the numerical values and is explicitly unquantified in the uncertainty budget.","tokens_in":14288,"tokens_out":9037,"duration_ms":89350,"concrete_test":"Perform a four-component Dirac-Coulomb-Breit (DCB) SCF calculation for ScF, ScF+, AcF, and AcF+ using the same s-aug-dyall.cv4z basis sets and equilibrium geometries as in the paper. Define the molecular Breit contribution as ΔBreit_mol = [E_DCB(MF+) - E_DC(MF+)] - [E_DCB(MF) - E_DC(MF)]. Compute the analogous atomic M+ Breit contribution at the same level of approximation (SCF) and compare. If the molecular and atomic SCF Breit differences disagree by more than 2 meV for either molecule, the transferability assumption used for Table IV is not supported and a Breit uncertainty must be added. If a correlated molecular DCB calculation is feasible (e.g., via MRCC or a future DIRAC implementation), repeat the comparison at CCSD level for ScF as a stronger test.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section III (Breit paragraph, Table IV row '+ΔBreit') applies the Breit contribution to the atomic M+ ionization potential as a stand-in for the molecular MF -> MF+ Breit correction, with the only justification being that the electronic structure of MF resembles M+. No benchmark is provided. The correction is not negligible: +6.9 meV (ScF), +6.9 meV (YF), +7.0 meV (LaF), and -3.0 meV (AcF), while the reported total uncertainties are 10.7, 10.3, 16.0, and 13.9 meV (Table V). For ScF and YF, the Breit term is roughly two-thirds of the uncertainty, and Table V contains no Breit-specific error term, so an error of 5-10 meV in transferability would shift the central values outside the claimed error bars. The ionized electron has predominant metal s-character, but in the molecule it occupies a bonding orbital with F contributions and a different screening environment than in the bare M+ ion; Breit is a short-range two-electron effect sensitive to exactly these details. The paper itself notes that no straightforward four-component molecular Breit implementation exists at the correlation level, so this transfer is an untested modeling choice rather than a controlled approximation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports theoretical adiabatic and vertical ionization potentials for ScF, YF, LaF, and AcF using four-component relativistic CCSD(T) with complete-basis-set extrapolation, followed by a posteriori corrections for active-space truncation, diffuse augmentation, full triples, Breit (taken from atomic M+ FSCC calculations), and QED effective potentials. The final recommended adiabatic IPs are 7.067(11) eV for ScF, 6.694(10) eV for YF, 5.929(16) eV for LaF, and 6.044(14) eV for AcF (Table VI). The authors compare with older mass-spectrometric measurements, which they argue are unreliable at the 0.3 eV level, and with a recent AcF calculation, and they provide an uncertainty budget with contributions from basis set, correlation treatment, and QED.","tokens_in":14568,"tokens_out":8067,"duration_ms":81199,"significance":"If the results withstand the concerns below, this would be the most accurate determination of the ionization potentials of these four molecules and would provide a concrete target for the planned AcF spectroscopy. The paper has notable strengths: it uses a well-established four-component coupled-cluster protocol, it does not fit any target IP, it is consistent with the authors' earlier validated treatment of CaF, SrF, and BaF, and it gives a transparent breakdown of most uncertainty sources. The main limitation is that the Breit correction is transferred from atomic ions without a molecular benchmark, and two entries in the uncertainty table are not derived in the text. These issues are fixable but currently affect the central claim of 10–16 meV accuracy.","major_comments":[{"comment":"The Breit correction for the molecular IP is approximated by the atomic M+ Breit correction, but no molecular benchmark is provided. The corrections in Table IV are +6.9 meV (ScF), +6.9 meV (YF), +7.0 meV (LaF), and -3.0 meV (AcF), while the total uncertainties in Table V are 10.7, 10.3, 16.0, and 13.9 meV; for ScF and YF the Breit term is roughly two-thirds of the total uncertainty and Table V contains no separate Breit uncertainty. Because the ionized electron in the molecule occupies a bonding orbital with F contributions rather than an atomic s-orbital, the atomic-to-molecular transfer could reasonably differ by more than a few meV. I request either a molecular-level estimate of the Breit contribution (for example, a DCB calculation at the CCSD level on at least one of the molecules) or an explicit conservative uncertainty for this transfer before the claimed uncertainties are accepted.","section":"Section III, Breit paragraph; Table IV row '+ΔBreit'; Table V"},{"comment":"The text states that the uncertainty from neglected higher excitations is 'a conservative 10% fraction of the triples contribution,' but the numbers in Table V do not match 10% of the residual full-triples correction reported in Table IV. For ScF, Table IV lists +ΔT = 2.3 meV, 10% of which is 0.23 meV, whereas Table V lists 8.76 meV; the latter is close to 10% of the full CCSD-to-CCSD(T) difference of about 83 meV. The 'triples contribution' used for the uncertainty estimate must be defined explicitly, and the derivation of each entry in Table V should be stated, since this is the dominant uncertainty for all four molecules.","section":"Section IV.b, Table V row 'higher excitations'"},{"comment":"The QED uncertainty entries in Table V (0.19, 0.56, 1.25, and 4.01 meV) are never derived in the text. For AcF, the assigned QED uncertainty of 4.01 meV is two-thirds of the total uncertainty contribution from QED corrections, and the total uncertainty depends sensitively on this choice. Please document how these values were obtained, for example from basis-set dependence of the effective potentials, variation of the self-energy potential parameters, or comparison with atomic QED benchmarks.","section":"Section IV.c, Table V row 'QED'"},{"comment":"The 50 meV shift in the LaF IP when going from dyall.v4z to dyall.cv4z is three times larger than the final quoted uncertainty of 16 meV for LaF, yet the text only states that this behavior 'is under investigation.' Since the final LaF value is obtained from the cv basis family, the manuscript should either explain quantitatively why the v4z-to-cv4z difference does not contribute to the uncertainty or enlarge the LaF uncertainty accordingly; an unresolved basis-set issue cannot support a 16 meV claim.","section":"Section III, Table II and following paragraph"}],"minor_comments":[{"comment":"The heading reads 'METHOD AND AND COMPUTATIONAL DETAILS'; the duplicated 'AND' should be removed.","section":"Section II heading"},{"comment":"There are several typographical errors, including 'similarlto' and 'mdedicated ea-surements'; please proofread the final text.","section":"Section V"},{"comment":"The text states that the diffuse-function uncertainty is evaluated with 'singly- and doubly-augmented dyall.v4z basis sets,' but Table II reports s-aug- and d-aug-dyall.cv4z; align the text with the actual calculations.","section":"Section IV.a"},{"comment":"The text refers to values 'shown in Table II in bold font,' but the displayed table does not show bold formatting; ensure the final typeset version marks the CBS-extrapolated row clearly.","section":"Table II"}],"recommendation":"major_revision","confidential_remarks":"The numerical results are plausible and the methodology is generally sound, but the Breit transfer and the unresolved LaF basis-set issue are the main risks to the quoted uncertainties. Both can be addressed within the manuscript's scope by adding conservative uncertainty terms or performing a benchmark calculation, so I do not recommend rejection. The higher-excitation and QED uncertainty entries also need more transparent derivation before the uncertainty budget can be considered reproducible."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is a solid, workmanlike computational chemistry paper that delivers the first reliable ionization potentials for ScF, YF, LaF, and a new AcF value that should directly help the planned Schiff-moment spectroscopy. It is not methodologically new — the protocol is the same CCSD(T)+corrections machinery the authors used for CaF/SrF/BaF — but the target molecules and the AcF prediction are genuinely new results.\n\nWhat it does well: the uncertainty budget is detailed and honest. Basis-set convergence, active space, augmentation, triple excitations, and QED corrections are checked separately, and each contribution is listed in Table V. The comparison with the earlier AcF calculation [37] is within combined uncertainties. They also correctly warn that the 1968 Knudsen-cell measurements carry ±0.3 eV uncertainties and an outdated mercury calibration, so their numbers supersede those by more than an order of magnitude in precision.\n\nThe main soft spot is the Breit correction. The paper takes the Breit contribution to the atomic M+ IP as a proxy for the molecular MF→MF+ Breit shift, because no four-component molecular Breit implementation is available at the correlation level. The stress-test note is right: this transfer is unbenchmarked, and for ScF and YF the Breit term (+7 meV) is roughly two-thirds of the total claimed uncertainty (10–11 meV). If the molecular-atomic difference is a few meV, the central values shift outside the error bars. The paper is transparent that this is a modeling choice, but it does not put any error bar on it. That is a real gap, though it does not undermine the qualitative conclusions — the IPs are certainly much better than the 0.3 eV experiments, and the AcF value should still be a reliable guide for spectroscopy.\n\nMinor issues: the 10% rule for higher-excitation uncertainty is a heuristic, and the large v4z→cv4z basis-set jump for LaF (50 meV) is left as \"under investigation\" without further comment. The QED corrections use effective potentials; for AcF they reach 6 meV, so a dedicated QED treatment would be welcome but is unlikely to change the picture.\n\nBottom line: this paper deserves a serious referee. It is transparent, technically sound at the level claimed, and the AcF IP is exactly the kind of input experimental groups need. I would send it to review, with a request that the authors either benchmark the Breit transfer on a lighter molecule where a molecular Breit calculation is feasible, or at least assign an explicit uncertainty to it.","headline":"A transparent, carefully budgeted relativistic CCSD(T) study that gives the first accurate IPs for ScF/YF/LaF and a credible AcF prediction; the main soft spot is the unbenchmarked atomic Breit transfer, which the authors disclose but do not fully quantify.","tokens_in":15116,"tokens_out":2125,"would_cite":true,"duration_ms":21753,"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":"The paper reports relativistic coupled-cluster values for the adiabatic ionization potentials of ScF, YF, LaF, and AcF — 7.067(11), 6.694(10), 5.929(16), and 6.044(14) eV — with the AcF value serving as the prediction for planned…","keywords":["adiabatic ionization potential","actinium monofluoride","relativistic coupled cluster","complete basis set extrapolation","QED corrections","Breit interaction","Schiff moment","spectroscopic uncertainties"],"falsifier":"A sub-10 meV measurement of the AcF ionization threshold, for instance by Rydberg-series or threshold photoelectron spectroscopy, that disagrees with the recommended 6.044(14) eV would refute the prediction; before any experiment, a direct four-component relativistic molecular calculation of the Breit contribution would test the weakest assumption by comparing the neutral-minus-cation Breit shift with the atomic M+ value.","tokens_in":14085,"feed_emoji":"⚛️","tokens_out":15694,"duration_ms":138232,"temperature":0.7,"pith_summary":"The paper sets out to fix the ionization potentials of four group-III monofluorides — ScF, YF, LaF, and AcF — at the few-meV level, with actinium fluoride as the principal motivation because AcF is a candidate for nuclear Schiff-moment searches. It uses relativistic coupled-cluster theory with single, double, and perturbative triple excitations, extrapolates the molecular energies to the complete basis set limit, and then adds corrections for a larger active space, extra diffuse functions, full triples, Breit interaction, and QED effects. The recommended adiabatic values are 7.067(11) eV for ScF, 6.694(10) eV for YF, 5.929(16) eV for LaF, and 6.044(14) eV for AcF. If these numbers are right, they are more accurate than the available mass-spectrometric measurements for the first three molecules and provide the first reliable spectroscopic target for AcF.","feed_headline":"Ionization energies of four metal fluorides are predicted to 10-16 meV","feed_subtitle":"Theoretical values for ScF, YF, LaF and AcF give spectroscopy a target, with actinium fluoride the payoff.","key_machinery":"The load-bearing construction is a layered correction identity: $\\mathrm{IP} = E_{\\mathrm{DC\\text{-}CCSD(T)}}^{\\mathrm{CBS}} + \\Delta_{\\mathrm{active}} + \\Delta_{\\mathrm{aug}} + \\Delta_T + \\Delta_{\\mathrm{Breit}} + \\Delta_{\\mathrm{QED}}$. The baseline term is the adiabatic energy difference between the neutral $X^{1}\\Sigma$ state and the cation $X^{2}\\Sigma_{1/2}$ state, each evaluated at its own equilibrium bond length, with energies extrapolated to the complete basis set limit through a three-cardinality extrapolation formula [41]. The five $\\Delta$ terms are small a posteriori corrections — a larger active space, extra diffuse functions, full triples, Breit interaction, and QED effects — each computed with an independent variant of the method; their main job is to decompose and quantify every source of error so that a total uncertainty can be assembled by summing independent contributions.","core_discovery":"The central claim is that a four-component Dirac–Coulomb CCSD(T) calculation, extrapolated to the complete-basis-set limit on singly augmented core-valence basis sets and corrected for active-space incompleteness, basis augmentation, residual triples, Breit, and QED contributions, determines the adiabatic ionization potentials of ScF, YF, LaF, and AcF to within 10–16 meV. The paper presents the final numbers — 7.067(11), 6.694(10), 5.929(16), and 6.044(14) eV respectively — as the best available values: distinctly more precise than the roughly 0.3 eV mass-spectrometric determinations for the first three, and a new prediction for AcF. The authors explicitly interpret the rise from LaF to AcF as a relativistic stabilization of the s-orbital from which the electron is removed.","pith_inferences":["Beyond the paper, the same correction-stack strategy could be extended to other actinide and lanthanide monofluorides, monoxides, or monohydrides whose IPs are unmeasured; the main bottleneck would be the same molecular Breit calculation that the paper approximates atomically.","If a future experiment confirms the AcF value, it would simultaneously validate the molecular effective-QED treatment at the heavy end of the periodic table, because the QED correction is largest for AcF at about −6 meV.","A direct four-component relativistic molecular Breit calculation for any one of the four systems would be the sharpest available check on the weakest assumption and could be performed before any new experiment on AcF."],"forward_implications":["The AcF prediction gives experimental campaigns a concrete search target: the adiabatic threshold should lie near 6.044 eV, with the zero-point-corrected 0–0 band near 6.048 eV.","If the new values are right, the older mass-spectrometric IPs for ScF, YF, and LaF are low by roughly 0.3–0.5 eV, so those measurements should be revisited with higher-resolution techniques.","The predicted rise from LaF at 5.929 eV to AcF at 6.044 eV sharpens the test of relativistic s-orbital stabilization in the actinide fluoride series.","For the three lighter molecules, the theoretical values are precise enough to serve as temporary benchmarks until modern measurements replace the older ones."],"supporting_citations":[{"why":"Supplies the only existing experimental IPs for ScF, YF, and LaF, with ±0.3 eV uncertainties and an outdated mercury calibration; these are the baseline the new values disagree with and motivate re-measurement.","marker":"[20]"},{"why":"Provides the previous relativistic coupled-cluster study of AcF, including an IP of 6.058(16) eV that the present AcF value is compared against.","marker":"[37]"},{"why":"Is the program package used for all baseline four-component CCSD(T) calculations of the molecular energies and properties.","marker":"[38]"},{"why":"Supplies the relativistic basis sets in three cardinalities used for the complete-basis-set extrapolation and for the core-valence and augmentation tests.","marker":"[39, 40]"},{"why":"Gives the complete-basis-set extrapolation formula used to convert the three-cardinality sequence into CBS-limit ionization energies.","marker":"[41]"},{"why":"Documents the earlier application of the same methodology to CaF, SrF, and BaF, establishing that alternative extrapolation schemes agree to within 1 meV and providing the uncertainty-estimation approach reused here.","marker":"[42]"},{"why":"Supplies the separate coupled-cluster implementation used to compute the CCSDT minus CCSD(T) difference that fixes the residual-triples correction.","marker":"[45]"},{"why":"Provides the atomic Fock-space coupled-cluster code used to evaluate the Breit contribution on the M+ ions, which is then transferred to the molecules.","marker":"[46]"},{"why":"Supplies the effective QED potentials used to compute the self-energy and vacuum-polarization corrections to the ionization energies.","marker":"[47]"}],"fun_headline_variants":["Metal fluoride ionization potentials pinned to 10-16 meV","ScF, YF, LaF, AcF ionization energies: theory leads with meV precision","AcF ionization potential predicted: theory sets benchmark for spectroscopy","Relativistic effects shape ionization energies of ScF, YF, LaF, AcF"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the Breit contribution to the molecular ionization energy can be taken from the atomic M+ ion because the molecule's electronic structure resembles that ion; if the true molecular Breit shift differs from the atomic one by more than a few meV, the stated 10–16 meV uncertainties are too small.","fun_headline_variants_meta":{"raw":{"variants":["Metal fluoride ionization potentials pinned to 10-16 meV","ScF, YF, LaF, AcF ionization energies: theory leads with meV precision","AcF ionization potential predicted: theory sets benchmark for spectroscopy","Relativistic effects shape ionization energies of ScF, YF, LaF, AcF"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000559,"raw_usage":{"total_tokens":2596,"prompt_tokens":825,"completion_tokens":1771,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":441,"completion_tokens_details":{"reasoning_tokens":1686}},"tokens_in":441,"tokens_out":1771,"duration_ms":12797,"temperature":1.0,"reasoning_tokens":1686,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T19:38:02.048591+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A sub-10 meV measurement of the AcF ionization threshold, for instance by Rydberg-series or threshold photoelectron spectroscopy, that disagrees with the recommended 6.044(14) eV would refute the prediction; before any experiment, a direct four-component relativistic molecular calculation of the Breit contribution would test the weakest assumption by comparing the neutral-minus-cation Breit shift with the atomic M+ value.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the only existing experimental IPs for ScF, YF, and LaF, with ±0.3 eV uncertainties and an outdated mercury calibration; these are the baseline the new values disagree with and motivate re-measurement."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the previous relativistic coupled-cluster study of AcF, including an IP of 6.058(16) eV that the present AcF value is compared against."},{"cited_title":"Shenyavskaya and B","cited_arxiv_id":null,"evidence_quote":"Is the program package used for all baseline four-component CCSD(T) calculations of the molecular energies and properties."},{"cited_title":"Dyall, Relativistic double-zeta, triple-zeta, and quadruple-zeta basis sets for the light elements H–Ar, Theor Chem Acc 135 (2016)","cited_arxiv_id":null,"evidence_quote":"Documents the earlier application of the same methodology to CaF, SrF, and BaF, establishing that alternative extrapolation schemes agree to within 1 meV and providing the uncertainty-estimation approach reused here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the separate coupled-cluster implementation used to compute the CCSDT minus CCSD(T) difference that fixes the residual-triples correction."},{"cited_title":"Martin, The total atomization energy and heat of for- mation of HCN(g), Chemical Physics Letters 259, 679 (1996)","cited_arxiv_id":null,"evidence_quote":"Provides the atomic Fock-space coupled-cluster code used to evaluate the Breit contribution on the M+ ions, which is then transferred to the molecules."},{"cited_title":"Lesiuk and B","cited_arxiv_id":null,"evidence_quote":"Supplies the effective QED potentials used to compute the self-energy and vacuum-polarization corrections to the ionization energies."}],"review_version":1}