{"id":"8585515f-1d01-4b07-8121-0a6ea8ae2155","arxiv_id":"2507.19421","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In lepton-flavour-violating new physics models where heavy states do not couple to right-handed charged leptons, one-loop electric dipole moments vanish in both the mass and Yukawa bases, leaving two-loop effects as the leading contribution.","lead":"This paper studies heavy new particles that change lepton flavour and have complex couplings, but do not interact with right-handed charged leptons. It argues that in such models, one-loop electric dipole moments of electrons, muons, and tau leptons still vanish, so only two-loop effects contribute.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Exact 'vanish' overstates the proof: C_EH = H Y_l is established only up to explicitly neglected non-hermitian terms of relative order m_l^2/v^2, so one-loop EDMs are suppressed, not exactly zero.","rationale":"The reader's verdict is CONDITIONAL and the weakest assumption points to exactly the place where the proof is soft: the C_EH = H'Y_l structure is assumed up to dropped terms. I agree. The basis-definition worry is less severe because the absence of NP couplings to e_R is a flavor-blind statement preserved under e_R rotations, and the matching is performed in the NP mass basis where the condition is well-defined; the paper is explicit about this. The real load-bearing step is eq (19): the non-hermitian correction [m_l m_l^dagger][H][m_l] is dropped as numerically small, but it is precisely the kind of term that can produce an imaginary diagonal dipole after rediagonalisation. The same applies to the RGE terms [Y_lY_l^dagger][C_EH] in eq (A3). Because these terms are suppressed by m_l^2/v^2, the one-loop EDM does not literally vanish; it is merely tiny. For the electron and muon the suppression is so strong that the two-loop contributions dominate; for the tau the residual one-loop piece could be at the percent level of the two-loop contribution, so the exact wording of the abstract overstates the result. A dedicated model calculation keeping the dropped terms would settle the magnitude. This does not overturn the reader's conditional acceptance; it reinforces it, and the paper would be improved by replacing 'vanish' with 'vanish up to corrections of order m_l^2/v^2' and by stating explicit bounds.","tokens_in":18265,"tokens_out":10702,"duration_ms":109027,"concrete_test":"Take a concrete UV model in the stated class (e.g., the Type-II seesaw with a complex triplet or the minimal scotogenic model), compute C_EH at one loop keeping the full Green-basis reduction including the -1/2 [m_l m_l^dagger][G_lD][m_l] term and the [Y_lY_l^dagger][C_EH] RGE terms, then evaluate the one-loop figure-2 diagram for the electron EDM in the mass basis. Compare Im C_ee^D with the two-loop estimate of eq (1); if the one-loop residual is below the two-loop value for all leptons, the practical conclusion stands and only the wording 'vanish' needs qualification; if it is comparable (likely only for tau), the claim needs a stated bound.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that one-loop EDMs vanish exactly, but the proof of the key structural condition C_EH = H'Y_l (needed to show that mass- and Yukawa-basis misalignment does not induce EDMs) drops terms that break that structure. In eq (19), after reducing the Green basis, the authors write C_EH = [H][m_l] + [m_l m_l^dagger][H][m_l], and then say: 'We neglect these unwelcome terms, which contradict the claim we aim to prove, because they are numerically small.' The second term is not of the form H'Y_l; writing it as (D_m H D_m) D_m leaves a non-hermitian prefactor, so after mass-basis diagonalisation it can generate Im C_D^ii at one loop. Similarly, the SMEFT RGE (A3) contains [Y_lY_l^dagger][C_EH] and [C_EH][Y_l^dagger Y_l] terms, which are dropped for the same reason; these also violate the H Y_l form unless H commutes with Y_lY_l^dagger. The relative size of all dropped terms is at most m_tau^2/v^2 ~ 1e-4, so the practical suppression is real, but the abstract's 'vanish' is not literally proven. There is also an unstated assumption that O_LEDQ and O_LEQU are not generated; if they are, the RGE terms proportional to them in (A3) can feed C_EH in forms not proportional to H Y_l. The exact cancellation is therefore conditional, not exact.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies heavy new physics models with lepton flavour-changing couplings and complex phases, in which the new heavy mass eigenstates do not interact with the SM singlet charged leptons e_R. It claims that in such models the one-loop dipole coefficient matrix has the structure C_D = (1/(16 pi^2 v)) H m_l with H hermitian, so the diagonal imaginary parts—the electric dipole moments—vanish at one loop in both the mass and Yukawa eigenstate bases, and EDM contributions first appear at two loops. The argument is organised in two steps: one-loop heavy-particle contributions are shown to be hermitian-coupling times mass, and NP corrections to the charged lepton mass matrix, computed in SMEFT at dimension six, are shown to preserve the same structure up to terms of relative order m_l^2/v^2, which the paper drops. The manuscript also discusses phenomenological patterns for l_j -> l_i gamma and EDM observables.","tokens_in":18749,"tokens_out":7312,"duration_ms":68080,"significance":"The paper's structural observation is useful: if the heavy NP interacts only with lepton doublets, the one-loop dipole matrix is forced into a hermitian-times-mass form, which protects the EDMs from one-loop contributions even after flavour-basis rotations. The two-step proof is well organised and the phenomenological discussion is clear, with a nice summary of expected suppression patterns. However, the central claim as stated in the abstract is 'vanish', whereas the proof establishes only a suppression by relative order m_l^2/v^2, because the key structural condition C_EH = H Y_l is proven only up to explicitly neglected non-hermitian corrections. This mismatch between the headline and the proven content is the main obstacle to acceptance.","major_comments":[{"comment":"The proof of [C_EH] = [H][Y_l] is not exact. After reducing the Green basis, eq. (19) gives [C_EH] = [H][m_l] + [m_l m_l^dagger][H][m_l] + ..., and the paper says 'We neglect these unwelcome terms, which contradict the claim we aim to prove, because they are numerically small.' In the mass basis, [m_l m_l^dagger][H][m_l] = D_m H D_m, which is not of the form H' D_m with H' hermitian, since D_m H is generally not hermitian. Therefore the effective Yukawa [Y_l] = (1/v)([m_l] + (v^2/Lambda^2)[C_EH]) is not of the form H'' D_m when this term is retained, and the vanishing of the diagonal imaginary parts of C_D in eq. (21) is not exact. The proven statement is a suppression by at most m_tau^2/v^2 ~ 1e-4 relative to the naive one-loop amplitude, not an exact zero. This should be stated honestly in the abstract and Section V, or the model class should be restricted so that the offending terms are absent (e.g., by a symmetry forcing [H,m_l] to vanish).","section":"Section IIIC, eq. (19)"},{"comment":"The RGE for C_EH contains the terms +4[C_EH][Y_l^dagger Y_l] + 5[Y_l Y_l^dagger][C_EH], which are not of the form H Y_l unless [C_EH] commutes with Y_l Y_l^dagger. These terms are dropped in the mass-matrix RGE (A4), as the text acknowledges. This is another place where the structural condition is only approximate: even if [C_EH] = H' Y_l at the matching scale, running to m_W generates non-hermitian corrections of relative order y_tau^2 or so. The paper should either prove the commutator vanishes for the model class or formulate the result as a one-loop-suppressed EDM rather than a vanishing one.","section":"Appendix A, eqs. (A2)-(A3)"},{"comment":"The argument assumes 'we suppose that neither O_LEDQ nor O_LEQU are generated in matching the model to SMEFT.' This is a non-trivial model assumption. The stated model condition—heavy states do not interact with singlets at tree level—does not by itself exclude loop-generated O_LEDQ or O_LEQU. If these operators are present, their Yukawa-type mixing into C_EH through the RGE terms in (A3) can violate the H Y_l structure. The theorem should therefore state this as an explicit hypothesis, and the paper should comment on whether the neutrino-mass models it cites (type I/II/III seesaw, scotogenic) satisfy it.","section":"Section IIIC, paragraph after eq. (19)"}],"minor_comments":[{"comment":"The sentence 'Symmetries can be inposed to inhibit flavour change' contains a typo; 'inposed' should be 'imposed'.","section":"Section IV"},{"comment":"The manuscript refers to Figure 3 but the figure does not appear in the text; please check that the submission includes it.","section":"Figure 3"},{"comment":"Reference [46] duplicates reference [9] (both cite Buras, hep-ph/9806471); please remove the duplicate.","section":"References"},{"comment":"The notation '(eQ)' inside the matrix entries of eq. (10) is confusing; the units should be defined in the caption or in the surrounding text.","section":"Eq. (10)"}],"recommendation":"major_revision","confidential_remarks":"For the editor: the manuscript's abstract overstates the strength of the theorem, and the body already contains the admission that the crucial terms are dropped as numerically small. The result is likely publishable after the claims are revised to 'suppressed' and the hypotheses are stated precisely, but the current wording is not acceptable for publication as an exact vanishing theorem."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nQuick take: the paper is right in substance but overclaims the proof. If you work on LFV/EDM model building, the structural result is worth having: in singlet-blind models, one-loop l_j -> l_i gamma amplitudes cannot be rotated into one-loop EDMs via NP-induced mass mixing, because the dipole coefficient keeps the form H m_l with hermitian H. That is a real gap in the literature—previous model-specific two-loop calculations didn't state it this way.\n\nThe first step, matching at Lambda_NP, is clean. The couplings that appear are hermitian combinations, so each separate diagram's diagonal imaginary part vanishes; off-diagonal complex phases stay off-diagonal. No issue there.\n\nThe soft spot is the second step, and it is visible in the text. Eq (19) writes C_EH = H m_l + m_l m_l^dagger H m_l, then says: \"We neglect these unwelcome terms, which contradict the claim we aim to prove, because they are numerically small.\" That sentence does a lot of work. The dropped term is not of the form H' Y_l, and after mass-basis diagonalisation it can in principle feed Im C_D^ii. The same happens with the Y_l Y_l^dagger C_EH and C_EH Y_l^dagger Y_l terms in the RGE (A3). Each is relatively O(m_l^2/v^2) <= ~1e-4, so the practical suppression is real and the phenomenological message survives. But the abstract says \"vanish,\" and that is not literally established. It should say \"suppressed below one-loop level by at least m_l^2/v^2,\" or better, the paper should bound the neglected terms explicitly.\n\nThere are also stated assumptions that limit the generality: O_LEDQ and O_LEQU are assumed not generated, and the singlet-blind condition is imposed in the massless SM basis at Lambda_NP. Both are reasonable for the intended neutrino-mass models, but they are assumptions, not derived.\n\nThe phenomenology section is honest: the patterns are \"expected, not obligatory,\" and cancellations are acknowledged. Citations are appropriate, including the self-citations, which are not load-bearing.\n\nBottom line: deserves a serious referee. The question is worth asking, and the answer is probably right for all practical purposes. The main revision needed is to align the claim with the proof—soften \"vanish\" or derive explicit bounds on the neglected terms. I'd bring it to reading group; it's a good case study in why \"hermitian times mass\" arguments need care with non-hermitian corrections.\n\nRecommendation: send to peer review, with a referee asked to focus on eq (19) and the RGE dropping in Appendix A.\n\nBest.","headline":"Right in substance but 'vanish' overstates the proof: the dropped non-hermitian terms of relative order m_l^2/v^2 mean one-loop EDMs are suppressed, not exactly zero.","tokens_in":19121,"tokens_out":2795,"would_cite":true,"duration_ms":27601,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["11.30.Er","13.40.Em","12.60.Cn"],"model":"deepseek-v4-flash","headline":"This paper claims that in new-physics models where heavy particles couple to lepton doublets but not to right-handed singlet leptons, one-loop electric dipole moments vanish exactly because the one-loop dipole coefficient matrix is a…","keywords":["electric dipole moments","lepton flavour violation","SMEFT","hermitian flavour structure","charged lepton mass matrix","radiative lepton decays","CP violation","neutrino mass models"],"falsifier":"Take any concrete model in this class (for instance the type-II seesaw, where the triplet does not couple to $e_R$), compute the complete one-loop dipole coefficient matrix in the mass basis without dropping the $m_l^2/v^2$ and $Y_l Y_l^\\dagger C_{EH}$ terms, and check the diagonal imaginary part: if it is nonzero at any order, the claimed one-loop vanishing is false. A complementary check is to evaluate the two-loop Barr-Zee diagrams and confirm they are the first non-vanishing EDM contributions, as the argument predicts.","tokens_in":18038,"feed_emoji":"🧲","tokens_out":12386,"duration_ms":110253,"temperature":0.7,"pith_summary":"This paper examines a class of new-physics models motivated by neutrino masses: heavy, lepton-flavour-changing extensions with complex couplings in which the heavy particles talk to the left-handed lepton doublets but not to the right-handed charged-lepton singlets. In such models, one-loop diagrams generate complex, flavour-changing dipole amplitudes such as $\\mu\\to e\\gamma$, but the one-loop electric dipole moments (EDMs) are claimed to vanish. The worry is that new-physics corrections to the charged-lepton mass matrix could misalign the mass and Yukawa bases, rotating the complex off-diagonal dipole phases into diagonal, CP-violating entries. The paper shows this does not happen: the one-loop dipole coefficient matrix always has the form $H m_l$ with $H$ hermitian in flavour space and $m_l$ diagonal, so the diagonal entries are real and EDMs first appear at two loops.","feed_headline":"One-loop EDMs vanish in left-handed new-physics models","feed_subtitle":"Complex flavour-changing couplings still give μ→eγ at one loop, but no EDM until two loops.","key_machinery":"The load-bearing object is the factorisation $C_D = (1/(16\\pi^2 v))\\, H D_m$ together with its analogue for the mass matrix, $C_{EH} = H' Y_l$. $H$ is hermitian in flavour space because it is built from the combination $y^J_{i\\alpha} y^{J*}_{j\\alpha}$ of complex couplings (or the analogous vector couplings), and the mass matrix appears on the right because the chirality flip comes from the external lepton line. The second ingredient is the SMEFT Green-basis reduction of the one-loop matching, which uses the lepton and Higgs equations of motion to express the dimension-six mass correction in terms of hermitian Green coefficients acting on the Yukawa matrix. The renormalisation-group running of $C_{EH}$ preserves this form because the Higgs minimisation condition cancels the potentially dangerous terms, so the misalignment between mass and Yukawa bases never creates a one-loop EDM.","core_discovery":"The central claim is that the one-loop dipole coefficient matrix in these models factorises as $C_D = (1/(16\\pi^2 v))\\, H D_m$, where $H$ is hermitian in lepton flavour space and $D_m$ is the diagonal charged-lepton mass matrix. Since the diagonal entries of a hermitian matrix are real, the imaginary parts of the diagonal dipole coefficients --- the electric dipole moments --- vanish at one loop, in both the mass eigenstate basis and the Yukawa eigenstate basis. The argument has two steps: one-loop matching diagrams with heavy particles produce a hermitian combination of couplings multiplied by an external mass insertion, and the new-physics correction to the charged-lepton mass matrix, encoded in the dimension-six SMEFT coefficient $C_{EH}$, is itself $H' Y_l$ up to corrections of order $m_l^2/v^2$, so any basis misalignment preserves the hermitian-times-mass structure. The residual EDM contributions are therefore two-loop effects, roughly the size quoted in eq. (1).","pith_inferences":["A decisive check of the structural argument would be to compute the full one-loop $C_D$ in a concrete model of this class (for example the type-II seesaw) without dropping the $m_l^2/v^2$ and $Y_l Y_l^\\dagger C_{EH}$ terms; a nonzero diagonal imaginary part at that order would mean the vanishing is approximate rather than exact.","The same hermitian-times-mass logic may extend to other chirality-flipping CP-violating observables, such as quark chromo-electric dipole moments, in models where new physics couples only to left-handed doublets; if so, the no-go would be a general feature of 'left-handed only' new physics rather than a lepton-specific accident.","A future positive muon EDM measurement would be difficult to accommodate in this model class, since the two-loop prediction lies roughly two orders of magnitude below the projected sensitivity; such a signal would point to new interactions with right-handed leptons or to CP violation in operators beyond the dipole."],"forward_implications":["The flavour-changing radiative decays $l_j \\to l_i \\gamma$ remain one-loop effects with complex off-diagonal coefficients, so $\\mu\\to e\\gamma$ and $\\tau\\to l\\gamma$ searches remain the most direct probes of this new-physics class.","The leading EDMs are two-loop and scale as $d_i \\sim \\frac{m_i}{(16\\pi^2)^2 \\Lambda_{\\rm NP}^2}\\,\\mathrm{Im}\\{\\Pi_{ii}\\}$, placing the muon and tau EDMs below foreseeable experimental sensitivity unless the CP-violating coupling products are large or the new-physics scale is low.","Because $H$ is hermitian, $|H_{ij}| = |H_{ji}|$, so the outgoing charged lepton in $l_j \\to l_i \\gamma$ is strongly left-handed; the right-handed polarisation rate is suppressed by $m_i^2/m_j^2$.","The electron EDM can probe a lower new-physics scale than $\\mu\\to e\\gamma$ in this model class, because the EDM carries an extra loop factor and an $m_e/m_\\mu$ suppression relative to the radiative decay; a future electron-EDM signal without $\\mu\\to e\\gamma$ would require a hierarchy in the hermitian matrix $H$.","Accidental cancellations or imposed flavour and CP symmetries can modify the expected patterns, but the absence of one-loop EDMs is structural and independent of the size of the complex off-diagonal couplings."],"supporting_citations":[{"why":"supplies the reduction-formula analysis of vanishing electric dipole moment diagrams that the paper adapts to the e_R-decoupled model class.","marker":"[16]"},{"why":"provides the original proof that two-loop sunset diagrams do not generate EDMs, the template for the chirality-flip argument.","marker":"[44, 45]"},{"why":"gives the Green-basis matching with equations of motion that turns the dimension-six mass correction into the $H Y_l$ form.","marker":"[47]"},{"why":"provides the one-loop SMEFT renormalisation-group equations for $C_{EH}$ that preserve the hermitian-times-Yukawa structure.","marker":"[50]"},{"why":"supplies the Yukawa-dependent RGE terms used to check that running does not spoil the mass-matrix structure.","marker":"[51]"},{"why":"defines the SMEFT operator basis and conventions used for the dipole, mass-matrix, and RGE expressions.","marker":"[21, 22]"}],"fun_headline_variants":["No one-loop EDMs, but μ→eγ still glows","Hermiticity kills one-loop EDMs","Left-handed NP: EDMs wait for two loops","Complex couplings, zero one-loop EDMs","μ→eγ at one loop, EDMs at two"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the new heavy particles genuinely have no interactions with the right-handed singlet leptons $e_R$ in the physical mass-eigenstate basis at the new-physics scale, with all Standard-Model particles treated as massless; if that condition is basis-dependent or acquires loop corrections, or if the dropped order-$m_l^2/v^2$ terms matter numerically, the one-loop $e_R$-Higgs diagrams and their EDM contributions need not vanish.","fun_headline_variants_meta":{"raw":{"variants":["No one-loop EDMs, but μ→eγ still glows","Hermiticity kills one-loop EDMs","Left-handed NP: EDMs wait for two loops","Complex couplings, zero one-loop EDMs","μ→eγ at one loop, EDMs at two"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000896,"raw_usage":{"total_tokens":3839,"prompt_tokens":901,"completion_tokens":2938,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":517,"completion_tokens_details":{"reasoning_tokens":2860}},"tokens_in":517,"tokens_out":2938,"duration_ms":23018,"temperature":1.0,"reasoning_tokens":2860,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T17:52:52.769720+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take any concrete model in this class (for instance the type-II seesaw, where the triplet does not couple to $e_R$), compute the complete one-loop dipole coefficient matrix in the mass basis without dropping the $m_l^2/v^2$ and $Y_l Y_l^\\dagger C_{EH}$ terms, and check the diagonal imaginary part: if it is nonzero at any order, the claimed one-loop vanishing is false. A complementary check is to evaluate the two-loop Barr-Zee diagrams and confirm they are the first non-vanishing EDM contributions, as the argument predicts.","supporting_citations":[{"cited_title":"Vanishing or non-vanishing rainbow? Reduction formulas of electric dipole moment","cited_arxiv_id":"2106.03384","evidence_quote":"supplies the reduction-formula analysis of vanishing electric dipole moment diagrams that the paper adapts to the e_R-decoupled model class."}],"review_version":2}