{"id":"51a8954f-7f79-4c5f-92c3-863536f37e42","arxiv_id":"1908.04578","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"First large-scale ab initio prediction of the Lr+ spectrum, with two independent methods agreeing and two strong ground-state transitions identified for experiment.","lead":"This paper calculates the low-lying spectrum of singly ionised lawrencium, an element produced one atom at a time, using two independent relativistic quantum chemistry methods. The predicted transition energies, lifetimes, and branching ratios are meant to guide upcoming laser-spectroscopy experiments on this heaviest known ion.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Uncertainty prescription under-covers its own Lu+ benchmark: for 5d6s 1D2 both methods give ~17884 cm^-1, experiment is 17332 cm^-1, so the 389 cm^-1 floor is not conservative and shared systematic error is not bounded.","rationale":"The paper's goal is to provide reliable search-window predictions for Lr+ spectroscopy, and the central recommendation hinges on the quoted uncertainties. The reader's conditional verdict already flags the homologue extrapolation. Reading Table I carefully, I found a concrete internal stress test that the reader's 'can only be tested by experiment' claim overlooks: the Lu+ calibration set itself contains a level where the proposed uncertainty prescription fails. This supports the conditional verdict but does not change it, because the underlying predictions are not invalidated: two independent methods bracket most Lu+ levels and the method differences for the headline Lr+ states are small. What the finding does is sharpen the condition: before the 389 cm^-1 uncertainty is used to set the laser scan window, the paper should either apply a leave-one-out calibration multiplier or explicitly downgrade the uncertainty from 'conservative' to 1-sigma. Since the reader already asked for conditional acceptance, my read leaves the verdict unchanged.","tokens_in":9383,"tokens_out":15462,"duration_ms":146009,"concrete_test":"Back-test the uncertainty prescription on the eight Lu+ levels of Table I: for each level compute the recommended value and uncertainty exactly as done for Lr+ (mean of FSCC and CI+MBPT; uncertainty = max(|Δmethod|, 389 cm^-1)), and tabulate which experimental energies fall outside the quoted intervals. Confirm whether 5d6s 1D2 (experiment 17332 cm^-1; methods 17892/17875 cm^-1; interval 17884±389 cm^-1) is outside; if so, recompute the minimum multiplier k such that all eight Lu+ intervals (mean ± k×389) contain experiment. If k > 1, the Lr+ uncertainties should be scaled by k, or the paper should explicitly state that 389 cm^-1 is a 1-sigma, not a conservative, bound.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section III derives the Lr+ uncertainties as max(method difference, 389 cm^-1), where 389 cm^-1 is the standard deviation of the CI+MBPT errors on the eight Lu+ levels in Table I. The claim that this is 'conservative' is contradicted by an in-sample case. For Lu+ 5d6s 1D2 the two methods return 17892 and 17875 cm^-1 (a 17 cm^-1 method difference), while the experimental value is 17332 cm^-1. The recommended mean is 17884 cm^-1, so the true error is about 550 cm^-1, i.e. ~160 cm^-1 outside the quoted ±389 interval. This is exactly the failure mode that matters for Lr+: the method-difference term is insensitive to a systematic bias shared by both calculations, and the floor taken from Lu+ does not cover even one of the calibration levels. The headline Lr+ lines (7s7p 3P1 at 31540±389 and 7s7p 1P1 at 47295±1048) may still be close, but the stated uncertainties—especially the 389 cm^-1 quoted for the primary search line—cannot be described as conservative bounds on the true error. The experiment will scan a finite window; an understated error can put the line outside the scan before any theory refinement is attempted.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports relativistic Fock-space coupled-cluster (FSCC) and CI+MBPT calculations of the low-lying excitation spectrum, g-factors, transition rates, branching ratios, and lifetimes of Lr+ and its lighter homologue Lu+. The authors benchmark both methods against experimental Lu+ energy levels and A coefficients, then use the Lu+ comparison to assign uncertainties to their recommended Lr+ energies, which are taken as the mean of the two methods with an uncertainty of max(method difference, 389 cm^-1). The main output is a set of recommended Lr+ excitation energies, of which the two strongest ground-state transitions are identified as 7s7p 3P1 at 31540 cm^-1 and 7s7p 1P1 at 47295 cm^-1, meant to guide the planned optical spectroscopy of Lr+. No experimental Lr+ data are used in the calculations.","tokens_in":9710,"tokens_out":5823,"duration_ms":59055,"significance":"If the predictions are reliable, this is the first systematic theoretical spectrum of Lr+ and will directly inform an ongoing experimental search, making the paper valuable to both atomic theory and superheavy-element spectroscopy. The study combines two independent state-of-the-art methods, presents explicit basis-convergence and MBPT partial-wave-convergence tests, and does not fit any Lr+ experimental data, so the recommended energies are genuinely ab initio. The Lu+ comparison is a useful validation of both methods and of the predicted A coefficients. However, the paper's central uncertainty prescription is not supported by its own benchmark, as one Lu+ level shows an error of about 550 cm^-1 while the quoted uncertainty floor is 389 cm^-1; because the experimental search window will be set by these quoted errors, this issue is load-bearing for the paper's main application.","major_comments":[{"comment":"The uncertainty prescription described in Section III is not conservative by the paper's own Lu+ benchmark. For Lu+ 5d6s 1D2, CI+MBPT gives 17892 cm^-1 and FSCC gives 17875 cm^-1, while the experimental value is 17332 cm^-1. The recommended mean is therefore 17884 cm^-1, an error of about 552 cm^-1, yet the formula max(method difference, 389 cm^-1) assigns this level an uncertainty of only 389 cm^-1. Because 389 cm^-1 is the standard deviation of the CI+MBPT residuals, it is expected that some individual residuals exceed it; calling that value 'conservative' is contradicted by an in-sample case. This matters directly for the primary Lr+ search line at 31540 cm^-1, whose quoted ±389 cm^-1 is set by the same floor and would not cover a shared bias of the magnitude seen in Lu+.","section":"Section III, Table I"},{"comment":"Using the difference between the two calculated energies as an error estimate is not sensitive to systematic errors common to both methods. The Lu+ 5d6s 1D2 case shows that the two methods can agree to 17 cm^-1 while both being about 550 cm^-1 from experiment, so a small method difference cannot be taken as evidence of small total error. Both calculations start from the same projected Dirac-Coulomb-Breit Hamiltonian and use similar QED-model corrections, so shared systematic errors are plausible. I ask the authors to replace or supplement the max(method-difference, standard-deviation) rule with a bound that covers the observed Lu+ residuals, for example the maximum residual or an expanded uncertainty, and to discuss which physical effects could produce a common bias in Lr+ but not in Lu+.","section":"Section III"},{"comment":"The statement 'We expect similar accuracy for the calculated transition energies of the heavier homologue of Lu+, Lr+' is the only justification for transferring the Lu+ error estimate to Lr+. The paper should make this transferability argument more concrete, for example by comparing the Lr+ energies with the independent prior calculations of Dzuba et al. and Cao and Dolg, or by estimating the sensitivity of the 7s7p levels to higher-order correlation, higher partial waves, and the QED-model uncertainty. Without such a check, the quoted uncertainties for Lr+ rest on an untested assumption; this is not a fatal flaw, but it is load-bearing for the experimental search-window application.","section":"Section III"}],"minor_comments":[{"comment":"In the first two rows of Table V, the lower level is labeled '7s2 2 S0' but the ground state is 7s2 1S0; the superscript should be 1.","section":"Table V"},{"comment":"The phrase 'Land` e g-factors' contains a formatting artifact; it should read 'Landé g-factors'.","section":"Section II B"},{"comment":"The definition of the even-tempered exponents in Eq. (4) is hard to read because the value of gamma is split across lines; please format it as a single number.","section":"Section II A, Eq. (4)"},{"comment":"The hyphen in 'branching-ratios' should be removed for consistency with the main text and standard usage.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":"The paper is technically solid in its methods and the Lu+ benchmark is a valuable check, but the uncertainty estimate is contradicted by an in-sample Lu+ level and must be revised before the predictions can responsibly guide an experimental search. The fix is straightforward—for example, using a maximum residual or an expanded uncertainty—so I would support publication after that revision. There are no concerns about citation patterns or scope fit."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nYou should know this paper is genuinely new: it gives the first large-scale ab initio spectrum of Lr+, including energies, g-factors, Einstein A coefficients, branching ratios, and lifetimes, with no experimental input for Lr+ itself. The validation on Lu+ is an honest attempt to gauge accuracy, and the two independent methods (FSCC and CI+MBPT) mostly bracket the known Lu+ levels. The convergence tests for basis size and MBPT partial waves are reported, not just waved at. For a superheavy ion where the planned optical search needs a narrow window, this is exactly the kind of predictive work that moves the field forward.\n\nThe main soft spot is the uncertainty prescription. The paper calls the 389 cm−1 floor “conservative,” but the Lu+ data in their own Table I contradict that. For 5d6s 1D2, both methods give ~17884 cm−1 while experiment is 17332 cm−1—a true error of ~550 cm−1, well outside the ±389 interval. The method difference is only 17 cm−1 there, so the max(method difference, 389) rule cannot catch the shared systematic bias. That matters for Lr+: the headline line at 31540 cm−1 could easily be off by more than the quoted ±389, and an experimental scan planned around that width might miss it. This does not invalidate the central prediction—the two methods agree and the Lu+ benchmark is broadly within a few hundred cm−1—but it means the stated uncertainties are not rigorous bounds, just rough estimates. The paper should say so plainly and maybe widen the search guidance to a few thousand cm−1.\n\nMinor issues: the A-coefficient uncertainties are not quantified (10–30% is quoted, but that is from Lu+), and the g-factor of the Lu+ 3P2 level sits far from experiment, which the authors flag as a possible misassignment in the data. Those are side notes, not fatal. The citation pattern is clean: prior Lr+ work is limited to ionization potentials or a misidentified ground state, and the authors correctly situate themselves. No circularity or fabrication concerns.\n\nNet: this paper deserves a serious referee. It is a solid, useful contribution for the atomic-physics and superheavy-element community. My advice to the editor: send it out, but ask the authors to revise the uncertainty discussion—either remove the word “conservative” or re-derive the error from the full Lu+ benchmark distribution, including the 1D2 outlier. The predictions themselves should stand.\n\n— [Your name]","headline":"First real Lr+ spectrum calculation, with a useful slate of atomic properties for the planned experiment—but the uncertainty bars are not as conservative as claimed, and one benchmark level sits well outside the quoted 389 cm−1 floor.","tokens_in":10239,"tokens_out":1690,"would_cite":true,"duration_ms":19183,"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 predicts the first systematic, high-precision spectrum of Lr+, identifying two strong ground-state transitions at 31540 and 47295 cm^-1 with uncertainties of at least 389 cm^-1.","keywords":["lawrencium","superheavy elements","atomic spectroscopy","Fock space coupled cluster","CI+MBPT","energy levels","Einstein coefficients","laser spectroscopy"],"falsifier":"A laser scan of the predicted $7s7p\\,{}^3P_1$ transition near $31540\\,\\mathrm{cm}^{-1}$ that finds no resonance within the stated uncertainty band, or finds the $7s7p\\,{}^1P_1$ line at $47295\\,\\mathrm{cm}^{-1}$ displaced by more than the recommended uncertainty, would refute the assumption that Lu+ accuracy transfers to Lr+.","tokens_in":9214,"feed_emoji":"⚛️","tokens_out":10704,"duration_ms":96351,"temperature":0.7,"pith_summary":"Singly ionized lawrencium (Lr+), with 103 protons, is made only one atom at a time, and no experimental spectrum exists. This paper tries to close that gap by computing the low-lying spectrum from first principles with two independent relativistic many-body methods and using the lighter homolog lutetium (Lu+) as a benchmark. It predicts that two electric-dipole transitions from the ground state are strong enough for laser searches: the $7s7p\\,{}^3P_1$ state at $31540\\,\\mathrm{cm}^{-1}$ and the $7s7p\\,{}^1P_1$ state at $47295\\,\\mathrm{cm}^{-1}$, with recommended uncertainties of at least $389\\,\\mathrm{cm}^{-1}$. If right, the theory shrinks the experimental search window from an unknown spectrum to a few specific wavelengths, and it provides lifetimes and branching ratios needed to plan the measurement.","feed_headline":"Lawrencium ion lines pinpoint two laser targets","feed_subtitle":"Two strong ground-state transitions at 31540 and 47295 cm⁻¹ mark where to search.","key_machinery":"The central device is homologue benchmarking with two independent relativistic methods: Fock space coupled cluster (FSCC), a multireference approach that builds excited states by adding electrons to a closed-shell reference, and configuration interaction plus many-body perturbation theory (CI+MBPT), which combines a CI expansion with second-order core-valence corrections. Both start from the projected Dirac-Coulomb-Breit Hamiltonian and include QED corrections through separate operators. Agreement between the two methods, validated by reproducing measured Lu+ levels and transition rates, is used to set the Lr+ error bars and to justify transferring the demonstrated accuracy to the heavier ion.","core_discovery":"The paper claims to provide the first large-scale systematic calculation of the Lr+ spectrum. Both methods place the ground state at $7s^2\\,{}^1S_0$ and agree on the ordering of the low-lying $6d7s$ and $7s7p$ levels; the recommended energies are the mean of the FSCC and CI+MBPT results, with uncertainty at least $389\\,\\mathrm{cm}^{-1}$ or the inter-method difference. For the two transitions proposed for experiment, the paper gives $7s7p\\,{}^3P_1$ at $31540 \\pm 389\\,\\mathrm{cm}^{-1}$ (lifetime 14.5 ns, branching ratio 0.90 to ground) and $7s7p\\,{}^1P_1$ at $47295 \\pm 1048\\,\\mathrm{cm}^{-1}$ (lifetime 1.1 ns, branching ratio 0.96). It also finds that the $6d7s\\,{}^3D_1$ level decays only by a suppressed M1 transition and has a lifetime of about 25 days. The accuracy argument rests on Lu+: average theory-experiment differences of $-263\\,(348)\\,\\mathrm{cm}^{-1}$ for FSCC and $16\\,(389)\\,\\mathrm{cm}^{-1}$ for CI+MBPT over the eight lowest relevant levels.","pith_inferences":["The paper does not spell this out, but the same dual-method benchmarking could be used to set search windows for the next even heavier ions, where no homologous experimental anchor exists.","A precise remeasurement of the Lu+ $3P_2$ g-factor would be a cheap indirect test of the wavefunction-quality assumption: the paper predicts 1.50 against a tabulated 1.66 and notes the assignment may be erroneous.","The metastable $6d7s\\,{}^3D_1$ state, with its 25-day lifetime, could in principle serve as an optical-clock or trapping state for Lr+ if the ion can be held long enough; this goes beyond the paper's experimental guidance."],"forward_implications":["Experimenters can begin a laser search by scanning a window around $31540\\,\\mathrm{cm}^{-1}$ for the $7s7p\\,{}^3P_1$ state, the strongest accessible ground-state transition with the smallest recommended uncertainty.","The $7s7p\\,{}^1P_1$ line at $47295\\,\\mathrm{cm}^{-1}$, with a branching ratio of 0.96 and a lifetime near 1 ns, provides a second, independent resonance to confirm the spectrum.","The predicted lifetimes and branching ratios allow estimates of detector sensitivity and required beam time for an experiment producing about one ion per second.","The very long predicted lifetime of the $6d7s\\,{}^3D_1$ level (about 25 days) means population can accumulate in a metastable state, which any excitation scheme must take into account or could exploit."],"supporting_citations":[{"why":"Defines the relativistic Fock space coupled cluster method used to compute the Lr+ and Lu+ excitation energies.","marker":"[6, 7]"},{"why":"Describes the CI+MBPT implementation used for g-factors, transition amplitudes, and the second set of energies.","marker":"[8]"},{"why":"Supplies the uncontracted universal Gaussian basis set on which the FSCC calculations are built.","marker":"[15]"},{"why":"Provides the model Lamb shift operator used to add QED corrections to the FSCC energies.","marker":"[16]"},{"why":"Gives the radiative potential used to incorporate QED corrections in the CI+MBPT calculation.","marker":"[25]"},{"why":"Supplies the experimental Lu+ energy levels against which both methods are benchmarked.","marker":"[29]"},{"why":"Provides the measured Lu+ transition rates used to benchmark calculated Einstein coefficients and branching ratios.","marker":"[30]"},{"why":"Introduces the extrapolated intermediate Hamiltonian that allows the large FSCC model spaces whose convergence is verified.","marker":"[14]"}],"fun_headline_variants":["First ab initio Lr+ spectrum maps two laser lines","Theory pinpoints two Lr+ transitions for laser search","Lr+ spectrum from first principles flags two lasers","Ab initio Lr+ energies give two laser search windows"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire error budget for Lr+ rests on the assumption that the accuracy demonstrated for Lu+ transfers to Lr+ unchanged, since no Lr+ experimental data exists to test it.","fun_headline_variants_meta":{"raw":{"variants":["First ab initio Lr+ spectrum maps two laser lines","Theory pinpoints two Lr+ transitions for laser search","Lr+ spectrum from first principles flags two lasers","Ab initio Lr+ energies give two laser search windows"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000836,"raw_usage":{"total_tokens":3666,"prompt_tokens":987,"completion_tokens":2679,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":603,"completion_tokens_details":{"reasoning_tokens":2611}},"tokens_in":603,"tokens_out":2679,"duration_ms":19212,"temperature":1.0,"reasoning_tokens":2611,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:38:07.803003+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A laser scan of the predicted $7s7p\\,{}^3P_1$ transition near $31540\\,\\mathrm{cm}^{-1}$ that finds no resonance within the stated uncertainty band, or finds the $7s7p\\,{}^1P_1$ line at $47295\\,\\mathrm{cm}^{-1}$ displaced by more than the recommended uncertainty, would refute the assumption that Lu+ accuracy transfers to Lr+.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Describes the CI+MBPT implementation used for g-factors, transition amplitudes, and the second set of energies."},{"cited_title":"Eliav, M","cited_arxiv_id":null,"evidence_quote":"Supplies the uncontracted universal Gaussian basis set on which the FSCC calculations are built."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the model Lamb shift operator used to add QED corrections to the FSCC energies."},{"cited_title":"Beloy and A","cited_arxiv_id":null,"evidence_quote":"Gives the radiative potential used to incorporate QED corrections in the CI+MBPT calculation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the experimental Lu+ energy levels against which both methods are benchmarked."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the measured Lu+ transition rates used to benchmark calculated Einstein coefficients and branching ratios."},{"cited_title":"High-accuracy relativistic coupled-cluster calculations for the heaviest elements,","cited_arxiv_id":null,"evidence_quote":"Introduces the extrapolated intermediate Hamiltonian that allows the large FSCC model spaces whose convergence is verified."}],"review_version":1}