{"id":"ca5d3586-6389-4b1c-bf28-d7fa177e64c7","arxiv_id":"2412.01231","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Self-energy corrections to E1 amplitudes in H-like ions are computed to all orders in Zα; the vertex+reducible part breaks the accuracy of effective QED operators for np-n'd transitions.","lead":"This paper computes the quantum electrodynamics self-energy correction to electric dipole transition amplitudes in hydrogen-like ions, to all orders in the nuclear binding strength. It finds that effective QED operators used for atomic structure calculations reproduce these corrections for s-to-p transitions but fail quantitatively for p-to-d transitions, which matters for precision atomic tests and astrophysical line ratios.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The np-n'd conclusion rests on vr values that are not independently checked; the only cross-check, 1s-2p1/2, is where vr is tiny, and Table I itself shows unexplained low-Z discrepancies.","rationale":"The reader's weakest assumption is close to mine, but I sharpen it: the danger is not only the low-Z cancellation in Table II; it is that the only cross-check is insensitive to exactly the contribution that matters. A separate observation from the manuscript is that Table I's agreement with Ref. [28] is not tight at low Z (up to 15% at Z=5), so calling it an 'important check of consistency' overstates the evidence. This does not prove the central claim wrong: the p-d conclusion could easily survive a clean recomputation, since QEDMOD visibly fails to reproduce even the po part for several entries (e.g., qmod-po/po of -9% to +109% in Table IV). But the current evidence for the vr part is not strong enough for ACCEPT. I would keep the reader's CONDITIONAL verdict: the paper should either provide the independent p-d check or document convergence of the vr evaluation before the 'cannot be reproduced' claim is used in many-electron applications.","tokens_in":15813,"tokens_out":13173,"duration_ms":124690,"concrete_test":"Compute the self-energy correction to the 3d3/2-2p1/2 decay rate in H-like Cs with the independent two-loop imaginary-part method of Ref. [28] and compare, via Eq. (24), with the Table IV entry for 2p1/2-3d3/2; agreement within the quoted uncertainties would support the vr part, while a discrepancy at the level of the QEDMOD deviation would undermine the central claim. As a cheaper internal check, rerun the vr evaluation for one p-d transition (e.g., 2p1/2-3d3/2) with the regulator rho varied from 10^-5 to 10^-7 and with the partial-wave expansion extended by several units; the result should remain stable within the quoted uncertainty.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that for np-n'd transitions in H-like Cs the perturbed-orbital part is not dominant and QEDMOD-type energy-level operators fail (Table IV, Summary). This rests on the vertex+reducible (vr) column of Table IV, not on the 1s-2p1/2 case. The only independent validation is the decay-rate comparison for 1s-2p1/2 (Table I, Eq. (24)) and the corresponding Table II sequence. In that case vr is only about 1% of the total self-energy correction, so even a large relative error in vr would barely change the checked observable; the comparison therefore cannot validate the p-d vr values that are comparable to or larger than po. The validation is also not quantitatively clean: Table I shows 3-15% deviations from Ref. [28] at Z=2-10, which the text assigns to 'numerical issues' without analysis. Since Table IV does not report the free/many decomposition or any convergence data (regulator rho, partial-wave cutoff, grid), a systematic error in the vr calculation could change the small se values and the qmod-se percentages that support the claim. This is the load-bearing assumption: the Table IV vr entries are accurate to the quoted uncertainties.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports ab initio calculations of the one-loop electron self-energy correction to E1 transition amplitudes in hydrogen-like ions, to all orders in the nuclear binding parameter Zα. The correction is split into a perturbed-orbital (po) part and a vertex-plus-reducible (vr) part. For the 1s-2p1/2 transition, the results are converted into corrections to the 2p1/2 decay rate and compared with the all-order calculation of Ref. [28]; good overall agreement is found. The authors then present an extensive set of results for ns-n'p and np-n'd transitions in H-like cesium (Z=55), including the frequency-dependent correction to the E1 operator, and compare the po part with the QEDMOD model-potential results. The central conclusion is that for ns-n'p transitions the po part dominates and QEDMOD reproduces the self-energy correction to within about 10% in most cases, while for np-n'd transitions the po and vr parts are of comparable magnitude and QEDMOD gives at best an order-of-magnitude estimate.","tokens_in":16041,"tokens_out":3162,"duration_ms":30211,"significance":"If the central claim is correct, the paper provides an important caution for the common practice of using effective QED potentials, such as QEDMOD, to estimate QED corrections to transition amplitudes in many-electron atoms: the method may be reliable for ns-n'p-type transitions but fails quantitatively for np-n'd transitions. The work also provides a large, systematic dataset of self-energy corrections for H-like cesium that could be used to construct model operators for transition amplitudes, and it quantifies the frequency-dependent correction to the length-gauge E1 operator. The comparison with the independent calculation of Ref. [28] is a valuable external consistency check, and the identification of the large cancellations between the free-electron and many-potential parts of the vr contribution is an honest warning about numerical difficulty. The main unmet need is direct evidence for the numerical accuracy of the vr values that underpin the central claim.","major_comments":[{"comment":"","section":"§III, Table IV, and discussion around Eq. (12)"},{"comment":"The low-Z deviations from Ref. [28] are described only as \"probably due to numerical issues\" without further analysis. For example, at Z=2 the present result is -0.00355(4) versus -0.00343 from Ref. [28], and at Z=10 it is -0.04762(5) versus -0.045, which are 3-6% level discrepancies. Since this comparison is the main external consistency check of the whole calculation, the authors should quantify the numerical uncertainty in the decay-rate conversion, explain the origin of these low-Z discrepancies, or at least demonstrate that the same numerical issue does not affect the vr values in Tables III and IV at the level claimed.","section":"§III, Table I"}],"minor_comments":[{"comment":"There are small presentational errors: \"Table III and IV presents\" should be \"present\", and \"ab inito\" appears in the text instead of \"ab initio\" in a couple of places. Please correct these.","section":"Section III, captions of Tables III and IV"},{"comment":"The quoted uncertainties are not consistent: several rows in Table II (e.g., Z=70, 80, 90, 100) and many entries in Tables III and IV carry no uncertainty, while the conclusions rely on comparisons at the level of a few percent. Please either provide uncertainties for all numerical values or state explicitly which entries are expected to be exact at the displayed digit level and why.","section":"Tables II, III, and IV"},{"comment":"The regulator ρ is stated to be typically 10^-6 and its error is said to be \"completely negligible,\" but no numerical sensitivity study is shown. A brief statement of how the final values of z_vr change when ρ is varied would make the claim verifiable and would also address the concern raised in the major comment.","section":"Section II.B, Eq. (14)"}],"recommendation":"major_revision","confidential_remarks":"The paper is technically substantial and the dataset is valuable, but the main conclusion about np-n'd transitions rests on a part of the calculation (the vr contribution in Table IV) that is not independently validated. The requested convergence checks or an alternative cross-check for one or more p-d cases are, in my view, necessary before publication. The use of QEDMOD as a benchmark is acceptable, but the fact that one of the authors is also a QEDMOD co-author should be stated explicitly in the manuscript to avoid any appearance of circularity."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this one. It is the first systematic all-order benchmark of self-energy corrections to E1 amplitudes across many hydrogenic transitions, and it fills a real gap: Sapirstein and Cheng only did ns-np1/2 in a few alkali atoms. The decay-rate comparison with Ref. [28] for 1s-2p1/2 is a meaningful independent check, and the two methods agree reasonably.\n\nThe main new result is the mapping of where the vertex+reducible (vr) part matters. For ns-n'p transitions, po dominates and QEDMOD works within ~10%, which is reassuring. For np-n'd, the paper argues that po is no longer dominant and QEDMOD fails even to reproduce po, so it is only order-of-magnitude. That conclusion is likely right, and it matters because many-electron calculations use QEDMOD for exactly these E1 amplitudes.\n\nThe soft spots are real but not fatal. Table I shows unexplained 3-15% deviations from Ref. [28] at Z=2-10, attributed to 'numerical issues' without analysis. Several table entries lack quoted uncertainties, and the nuclear model sensitivity at Z=90-100 is mentioned but not quantified. More importantly, the stress-test hits the right nerve: the only independent validation (the decay-rate comparison) is for 1s-2p1/2, where vr is about 1% of se. That check cannot validate the p-d vr values that drive the central claim. The free/many decomposition in Table II shows large cancellations, but Table IV gives no convergence data on the regulator, partial waves, or grid.\n\nThat said, the central qualitative claim does not collapse. Even if the vr numbers in Table IV were wrong by tens of percent, the fact that QEDMOD fails to reproduce the po part for p-d transitions is visible directly in the qmod-po column, independent of vr. So the paper's bottom line about QEDMOD is robust; the quantitative error bars on se and the qmod-se ratios are the fragile part.\n\nThis deserves a serious referee. The authors should be asked to provide data or at least convergence checks for the vr part in Table IV, and to explain the low-Z discrepancies. I would cite this for the benchmark values, and bring it to a reading group.","headline":"Serious all-order QED benchmark with a likely-right conclusion about QEDMOD for p-d transitions, but the vertex+reducible numbers for those transitions are not independently verified.","tokens_in":16632,"tokens_out":2725,"would_cite":true,"duration_ms":23527,"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":"For np-n'd E1 amplitudes in hydrogen-like ions, the perturbed-orbital part of the self-energy correction ceases to dominate, so energy-level QED operators cannot reproduce it.","keywords":["self-energy correction","E1 transition amplitude","hydrogen-like ions","vertex and reducible corrections","perturbed-orbital approximation","QEDMOD model potential","radiative corrections","hydrogen-like cesium"],"falsifier":"Compute the vr contribution to a specific $np$-$n'd$ transition in hydrogen-like cesium, for instance $2p_{1/2}$-$4d_{3/2}$ where the total correction nearly vanishes and changes sign, using an independent numerical scheme such as a different regularization of the dipole operator or a direct partial-wave summation that avoids the free/many split, and check that the remainder matches Table IV within the quoted uncertainty; a mismatch would overturn the claim that QEDMOD misses the correction by the stated amount.","tokens_in":1884,"feed_emoji":"⚛️","tokens_out":1916,"duration_ms":73817,"temperature":0.7,"pith_summary":"This paper computes the one-loop electron self-energy correction to electric-dipole (E1) transition amplitudes in hydrogen-like ions, to all orders in the nuclear binding parameter $Z\\alpha$. Its central result is that the correction splits into two parts with different behavior: a perturbed-orbital part that dominates for $ns$-$n'p$ transitions (about 99% for $1s$-$2p_{1/2}$), and a vertex+reducible part that becomes comparable for $np$-$n'd$ transitions. Since common effective QED potentials reproduce only the perturbed-orbital part, they estimate the self-energy correction to $p$-$d$ amplitudes only to order of magnitude, and sometimes with the wrong sign. The authors conclude that effective operators built to reproduce energy-level Lamb shifts cannot be trusted for $np$-$n'd$ E1 matrix elements, and they quantify this failure for hydrogen-like cesium.","feed_headline":"Perturbed-orbital dominance breaks down for np-n'd transitions","feed_subtitle":"In hydrogen-like cesium, the missing vertex part rivals the model-operator estimate for p-d E1 amplitudes.","key_machinery":"The central object is the decomposition $\\delta z_{\\rm se} = z_{\\rm po} + z_{\\rm vr}$, where $z_{\\rm po}$ sums intermediate-state matrix elements of the renormalized one-loop self-energy operator $\\Sigma_R(\\varepsilon)$ and $z_{\\rm vr}$ contains the vertex diagram minus reducible contributions. The vr part is further split into free-electron and many-potential contributions, with the free vertex part evaluated in momentum space after Fourier-transforming the electric dipole operator through a finite regulator $\\rho$. The paper's main diagnostic is the scaled correction $R_{\\rm se}(Z\\alpha) = (\\pi/\\alpha)\\,\\delta z_{\\rm se}/z_{ab}$, which is tabulated for many transitions and compared with the values obtained from the QEDMOD model potential.","core_discovery":"The paper establishes, through all-order-in-$Z\\alpha$ numerical calculations, that the one-loop self-energy correction to E1 amplitudes in hydrogen-like ions decomposes into a perturbed-orbital (po) part and a vertex+reducible (vr) part whose relative size depends on the orbital angular momentum change. For the $1s$-$2p_{1/2}$ transition across $Z = 2$ to $100$, the vr part is consistently about 1% of the total, so effective-potential approximations hold at that level. For $ns$-$n'p$ transitions in hydrogen-like cesium ($Z = 55$) the same is roughly true, with the QEDMOD model potential reproducing the ab initio total to within about 10% in most cases. For $np$-$n'd$ transitions, however, the po and vr parts are of the same order, the total self-energy correction is smaller and irregular in sign, and QEDMOD fails quantitatively, in several cases giving the wrong sign. The paper's conclusion is that effective QED operators constructed from energy-level data cannot well reproduce the self-energy correction to $np$-$n'd$ E1 matrix elements.","pith_inferences":["If the cesium pattern is generic, then in many-electron atoms the dominant QED effect on $p$-$d$ transitions may enter through correlation-induced configuration mixing rather than through the direct radial self-energy correction; the authors' earlier neon-like iron and nickel study is consistent with that route.","A testable extension is to compute the vr part for a few $np$-$n'd$ transitions in lighter and heavier hydrogen-like ions, for example $Z \\approx 30$ and $Z \\approx 80$, since the paper's detailed conclusions are for $Z = 55$ only.","The severe cancellation between free and many-potential parts of vr at low $Z$ suggests a cross-check with a completely different renormalization scheme; without such a check, the quoted sub-percent uncertainties at low $Z$ rest mainly on internal consistency.","A practical extension would be to fit the po part with the QEDMOD potential and tabulate vr residuals as a state-dependent operator; the sign reversal between $n_s < n_p$ and $n_s > n_p$ seen in the paper's figure indicates that such an operator would need to be nonlocal."],"forward_implications":["For $1s$-$2p_{1/2}$ decays in any hydrogen-like ion, the vertex and reducible self-energy parts shift the amplitude by about 1%, so effective-potential calculations are reliable at that level.","For $ns$-$n'p$ transitions in heavy hydrogen-like ions, the QEDMOD model potential reproduces the ab initio self-energy correction to within roughly 10% in most cases, with the worst relative errors occurring where the correction itself is abnormally small.","For $np$-$n'd$ transitions, the same model potential gives only the order of magnitude of the self-energy correction, and sometimes not even the sign, so many-electron $p$-$d$ amplitude calculations need explicit uncertainty estimates.","The frequency-dependent part of the E1 operator is a percent-level correction for hydrogen-like cesium and must be included when comparing with high-precision measurements.","The irregular dependence of the vr correction on principal quantum numbers indicates that a simple effective operator for E1 amplitudes will be difficult to construct."],"supporting_citations":[{"why":"Provides the prior ab initio E1 self-energy calculation for alkali $ns$-$np_{1/2}$ transitions whose conclusion about small vertex and reducible contributions this paper tests and qualifies.","marker":"[16]"},{"why":"Supplies all-order and $Z\\alpha$-expansion results for the $2p_{1/2}$ decay rate used as the benchmark for consistency of the new calculations.","marker":"[28]"},{"why":"Introduces the model-operator approach to Lamb shifts that underlies the approximate QED potential whose reproduction of the perturbed-orbital part is quantified here.","marker":"[13]"},{"why":"Provides the QEDMOD program used to generate the qmod comparison values in the paper's tables.","marker":"[14]"},{"why":"Describes the momentum-space calculation procedure for self-energy corrections that the vertex+reducible evaluation follows.","marker":"[22]"},{"why":"Defines the one-loop self-energy operator matrix elements used to build the perturbed-orbital part.","marker":"[19]"},{"why":"Gives the relativistic E1 transition operator in length gauge that yields the paper's frequency-dependent correction estimate.","marker":"[25]"}],"fun_headline_variants":["Perturbed-orbital dominance breaks for np-n'd E1 amplitudes","Self-energy correction flips sign for np-n'd transitions","QEDMOD fails for p-d E1 matrix elements in H-like cesium","Effective QED operators miss self-energy in np-n'd transitions","All-order self-energy: p-d transitions defy model operators"],"cache_read_input_tokens":18688,"weakest_assumption_plain":"The numerical evaluation of the vertex+reducible part rests on a delicate cancellation between the free-electron and many-potential contributions, with the first four digits cancelling at $Z = 2$, and the paper assumes the remaining finite difference is accurate to the quoted uncertainties without reporting an independent verification of that cancellation.","fun_headline_variants_meta":{"raw":{"variants":["Perturbed-orbital dominance breaks for np-n'd E1 amplitudes","Self-energy correction flips sign for np-n'd transitions","QEDMOD fails for p-d E1 matrix elements in H-like cesium","Effective QED operators miss self-energy in np-n'd transitions","All-order self-energy: p-d transitions defy model operators"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000297,"raw_usage":{"total_tokens":1727,"prompt_tokens":957,"completion_tokens":770,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":573,"completion_tokens_details":{"reasoning_tokens":679}},"tokens_in":573,"tokens_out":770,"duration_ms":6817,"temperature":1.0,"reasoning_tokens":679,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T04:33:16.478079+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the vr contribution to a specific $np$-$n'd$ transition in hydrogen-like cesium, for instance $2p_{1/2}$-$4d_{3/2}$ where the total correction nearly vanishes and changes sign, using an independent numerical scheme such as a different regularization of the dipole operator or a direct partial-wave summation that avoids the free/many split, and check that the remainder matches Table IV within the quoted uncertainty; a mismatch would overturn the claim that QEDMOD misses the correction by the stated amount.","supporting_citations":[{"cited_title":"Sapirstein and K","cited_arxiv_id":null,"evidence_quote":"Provides the prior ab initio E1 self-energy calculation for alkali $ns$-$np_{1/2}$ transitions whose conclusion about small vertex and reducible contributions this paper tests and qualifies."},{"cited_title":"Radiative Corrections to One-Photon Decays of Hydrogenic Ions","cited_arxiv_id":"hep-ph/0311134","evidence_quote":"Supplies all-order and $Z\\alpha$-expansion results for the $2p_{1/2}$ decay rate used as the benchmark for consistency of the new calculations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the model-operator approach to Lamb shifts that underlies the approximate QED potential whose reproduction of the perturbed-orbital part is quantified here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the QEDMOD program used to generate the qmod comparison values in the paper's tables."},{"cited_title":"Yerokhin, A","cited_arxiv_id":null,"evidence_quote":"Describes the momentum-space calculation procedure for self-energy corrections that the vertex+reducible evaluation follows."},{"cited_title":"Yerokhin and V","cited_arxiv_id":null,"evidence_quote":"Defines the one-loop self-energy operator matrix elements used to build the perturbed-orbital part."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the relativistic E1 transition operator in length gauge that yields the paper's frequency-dependent correction estimate."}],"review_version":1}