{"id":"db3ee803-507e-42cd-8932-045bbfe42db6","arxiv_id":"hep-ph/9607366","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":7,"one_line_summary":"At large recoil, heavy-meson form factors are governed by hard gluon exchange and the meson light-cone wave functions, forcing an exact zero in the longitudinally polarized B* form factor and an approximate zero in the B form factor.","lead":"Physicists calculated how the form factors of heavy mesons behave when the mesons scatter at very high recoil, using the hard-gluon methods developed for pions. The result forces certain form factors through zero, a sharp qualitative signature that the authors say is unlikely to be measurable because the cross sections are tiny.","discovery_kind":"unclear","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"The ln²(v·v') enhancement of the Isgur–Wise function rests on an endpoint assumption and an incomplete wave-function set; the qualitative scaling and the zero survive, but the quantitative coefficient is soft.","rationale":"The paper's central qualitative claims — the two scaling regimes and the sign-based zero in hL — are robust, and the authors flag the quantitative softness themselves (rough estimate, v·v' ~ 100, cross sections likely irrelevant). The reader's endpoint concern is real, but the paper supports φ+ ~ ω, φ- ~ 1 via sum rules (Eq. 4.18) and the QCD analogue, so it is not the weakest link. The more load-bearing soft spot is the incompleteness of Eq. (5.4) (deferred q qbar g contribution, footnote 2) and the cutoff dependence of the ln² coefficient (Eq. 5.5, unknown O(ϑ) term). Neither overturns the central claim, so the verdict stands. The reader's note that the 'exact zero' is stronger than the hard-scattering approximation strictly guarantees is fair and consistent with my read.","tokens_in":59406,"tokens_out":39507,"duration_ms":844835,"concrete_test":"Recompute Eq. (5.4) including the deferred quark–antiquark–gluon subleading-twist contribution (footnote 2) and with an explicit IR cutoff in Eq. (5.5); if the coefficient of ln²(v·v')/(v·v')² changes by more than a factor of 2, the quantitative ξ prediction and the zero location in Eq. (5.16) shift materially, though the qualitative scaling survives.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative result is Eq. (5.4)/(5.10): ξ ~ αs f²/Λ³ · ln²(v·v')/(v·v')². Two things make its coefficient soft. First, footnote 2 defers the quark–antiquark–gluon subleading-twist wave function contribution to ξ, so Eq. (5.4) includes only the quark–antiquark interference; since subleading twist in the HQET includes both q qbar and q qbar g components (cf. Sec. II, where ⟨ω²⟩± are tied to q qbar g wave functions), the omitted piece is plausibly the same order. Second, Eq. (5.5) shows the ln² arises from an IR-divergent integral cut off by Λ²e^-ϑ, and the O(ϑ) term is explicitly stated to be uncalculable without the IR cutoff details, so the coefficient of ln² is scheme-dependent at subleading-log level. The reader's endpoint concern (φ+ ~ ω, φ- ~ 1) is the proximate cause, but the paper does support it via sum rules (Eq. 4.18) and the QCD analogue, so the endpoint behaviour itself is not the weakest link; the incompleteness and cutoff dependence are. Importantly, the zero in hL (Eq. 5.16) is a sign change driven by ξ > 0 versus ξ3 > 0 with different scalings, so it survives even if the ξ coefficient shifts; only its location moves.","agreement_with_reader":"partial"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"Grozin and Neubert compute the asymptotic behaviour of heavy-meson form factors at large recoil using the hard-scattering machinery of Brodsky–Lepage. The meson is described by two light-cone wave functions, φ+(ω) (leading twist) and φ−(ω) (subleading twist), defined as Fourier transforms of bilocal Q∗q operators (Sec. II, Eqs. 2.9–2.12), with moments fixed by the equations of motion (Eq. 2.19) and by the gluon-field matrix elements λ²E, λ²H (Eq. 2.24). Section III derives one-loop evolution kernels Γ±(ω,ω′) (Eq. 3.13) and identifies an extra ultraviolet divergence in the ω-integrated operators, absorbed into Z′± (Eq. 3.19). Section IV obtains QCD sum rules for λ²E, λ²H (Eq. 4.6, with estimates in Eq. 4.10) and for the wave functions themselves (Eq. 4.17), leading to the model shapes of Eq. (4.18). Section V computes the hard-gluon exchange diagrams: the Isgur–Wise function from leading–subleading interference (Eq. 5.4), and the order-1/mQ functions ξ3 and χ2 from leading twist alone (Eqs. 5.7, 5.9), with the relation χ2 = −ξ3 e^{−ϑ}. The relative sign and different scaling of ξ and ξ3 then force an exact zero in the longitudinally polarized B* form factor hL at v·v′ ~ mb/Λ, and an approximate zero in the timelike B form factor (Eq. 5.16). Section VI applies this to e+e− → B(*)B̄(*) and D(*)D̄(*), predicting a vanishing of the B̄B rate at x0 ≈ 25 (Eq. 6.4). Four appendices give the covariant trace formalism, non-local condensates, a comparison with two-loop sum rules, and a静态","tokens_in":60236,"tokens_out":37599,"duration_ms":687867,"significance":"If correct, the result gives a model-independent statement about the large-recoil behaviour of heavy-meson form factors that goes beyond the HQET's own expansion: it identifies which HQET functions dominate in which recoil window and predicts sign-forced zeros in physical form factors. The zero in hL is a falsifiable, sign-based prediction that does not depend on the detailed normalization of the wave functions, and the paper supports it with two independent cross-checks: the large-ϑ limit of the existing two-loop QCD sum rules (Appendix C, Eq. C1) and the static heavy-quark model (Appendix D). The explicit one-loop anomalous dimensions (Eq. 3.13) and the sum rules for the wave functions (Eq. 4.17) are of independent value. The quantitative coefficients, by contrast, rest on model wave functions and on a leading-log truncation, and the paper is candid about this.","major_comments":[{"comment":"The asymptotic Isgur–Wise function in Eq. (5.4) retains only the quark–antiquark interference term; footnote 2 defers the quark–antiquark–gluon subleading-twist contribution to ξ to a separate publication. This matters for the central claim of the abstract (s2), that ξ is 'determined by the interference between the wave functions of leading and subleading twist'. Section II itself ties the second moments ⟨ω²⟩± to quark–antiquark–gluon wave functions (Eq. 2.23 and the sentence following Eq. 2.24), so the omitted piece is plausibly of the same order in αs and in 1/Λ as the retained one. The authors acknowledge the deferral; what remains is either a bound or an argument that the omitted term is higher order, or a softening of the abstract's 'determined by' to 'receives its leading contribution from'.","section":"§V, Eq. (5.4) and footnote 2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The abstract states that 'the sum of these contributions describes the form factors in the whole region |v·v′| ≫ 1' (s4). Section V is more careful: it notes that QCD sum rules suggest soft contributions falling like 1/(v·v′)², that the asymptotic form (5.4)–(5.5) would then dominate only if ln(v·v′) ≫ 1 and αs ln²(v·v′) ≫ 1, and that reaching the asymptotic regime for ξ requires v·v′ = O(100). The abstract should carry this qualification, since as written it asserts a matching across the whole region that the body does not demonstrate.","section":"Abstract (s4) versus §V","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Section III derives the evolution equations (Eq. 3.13) but then states explicitly that renormalization effects are ignored in the rest of the paper. The hard-scattering scale in Section V is of order Λ v·v′, far above the sum-rule scale at which the model wave functions (4.18) are defined, so evolution would change the normalization of ξ and ξ3 and hence the zero location x0. The abstract lists the evolution equations as a result, yet they are never applied to the quantities that the paper predicts. Either estimate the effect of evolution on the ξ and ξ3 normalizations, or state plainly that the predictions are quoted at a fixed low scale and that the evolution equations are a separate structural result.","section":"§III versus §V","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Equation (5.5) shows that the ln²(v·v′) enhancement arises from an infrared-divergent integral cut off by Λ²e^{−ϑ}, and the text states that the O(ϑ) term cannot be computed without details of the infrared cutoff. The coefficient of ln² is therefore a leading-log result, scheme-dependent at subleading-log level, and the numerical values x0 ≈ 25 and x1 ≈ 3500 inherit the model dependence of Eq. (4.18) and of αs f²/Λ̄³ = 0.06. The paper says of x0 that 'the accuracy of this prediction is not high'; that caveat should be extended explicitly to x1 and to the ξ normalization quoted in Eq. (5.10), so that the reader does not take the numbers in Eqs. (6.4)–(6.5) as more than order-of-magnitude estimates.","section":"§V, Eq. (5.5) and Eqs. (5.10), (6.4), (6.5)","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The endpoint behaviour φ+(ω) ~ ω, φ−(ω) ~ 1 is motivated by the QCD analogue and later by the sum-rule model (4.18), but it is not derived. Since it drives the logarithmic enhancement in Eq. (5.5), a sentence in Section II recording its status as an assumption would help the reader separate what is derived from what is assumed.","section":"§II, Eq. (2.33)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The note that the light-cone-gauge check of Γ−(ω,ω′) could not be completed is honest but leaves the reader without an independent confirmation of the minus-projection kernel. A sentence on what check does exist, or on why the Feynman-gauge result is trustworthy, would strengthen this section.","section":"§III, footnote 5","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The quoted errors on λ²E and λ²H are described as 'a rough estimate' from averaging over the stability window, but the procedure that turns the τ-spread in Fig. 5 into the stated uncertainties is not given. A short explanation would make the numbers reproducible.","section":"§IV, Eq. (4.10) and Fig. 5","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The figures are computed with C = 1, i.e. with bremsstrahlung neglected, although the text estimates η ≈ 0.2–0.3. Stating this in the captions, or showing the curves with the C factor included, would make the figures consistent with the discussion in the text.","section":"§VI, Figs. 9 and 10","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The label R1 ≡ R_{BB*+B*_T B*_L} mixes a sum of final states with subscripts that elsewhere denote polarizations; the notation should be made uniform with Eq. (5.15).","section":"§VI, Eq. (6.2)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The comparison between the two-loop sum-rule normalization and the factorized square-model result is a valuable check, and the paper correctly notes that the triangle model is not fully consistent at large recoil. A sentence stating which of (C1) or (C3) is used for the numerical estimates in Section VI would remove the ambiguity.","section":"Appendix C, Eqs. (C1) and (C3)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The affiliation line reads 'Ru ssia' and several words in the running text are broken by stray spaces; the typeset version should be checked before publication.","section":"Front matter","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"minor_revision","confidential_remarks":"The physics is sound and the sign-based zero prediction is the kind of falsifiable statement the journal should publish. My reservation is presentational: the abstract (s2, s4) claims more than Section V delivers, and the evolution equations of Section III are advertised but never applied. Both are fixable with wording changes plus one short estimate, so I do not regard this as a major-revision case."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"Grozin and Neubert point the Brodsky–Lepage machinery at heavy-meson form factors at large recoil. The results worth having are the scaling laws: ξ falls as αs f²/Λ³ · ln²(v·v')/(v·v')², while the 1/mQ functions ξ3 and χ2 fall only as 1/(v·v'), so they take over past v·v' ~ mQ/Λ. The evolution equations for φ±(ω) (Eq. 3.13) and the sum rules for the wave functions are a useful byproduct, and the sign-based zero in hL is a sharp, checkable statement that survives even if the ξ coefficient shifts — only its location moves. Appendix C is the part I trust most: the asymptotic limit of the two-loop sum rules reproduces Eqs. (5.4), (5.7), (5.9), including the χ2 = −ξ3 e^{−ϑ} relation. That is a real cross-check, not decoration, and it is the right way to use prior published work as a consistency test rather than a crutch. The derivations are standard and carefully laid out. Where it is soft: the coefficient of ln²(v·v') in ξ. Footnote 2 defers the quark–antiquark–gluon subleading-twist contribution, and since subleading twist in HQET includes both q qbar and q qbar g pieces — the paper itself ties ⟨ω²⟩± to q qbar g wave functions in Sec. II — the omitted term is plausibly the same order. Eq. (5.5) also makes clear the O(ϑ) piece is not calculable without IR cutoff details, so the subleading-log coefficient is scheme-dependent. I would push back on the reader's framing here: the endpoint behaviour φ+ ~ ω, φ− ~ 1 is not the weak link, since the paper supports it from sum rules (4.18) and the QCD analogue. The incompleteness and cutoff dependence are. The 'exact zero' language is a touch strong: it is exact within the truncated hard-scattering result, and the neglected corrections move its location. The authors do flag that the quantitative predictions are rough — v·v' ~ 100 needed, cross sections likely irrelevant — which is the right posture. For anyone working on large-recoil HQET or exclusive heavy-meson production, this is a reference worth having. Send it out.","headline":"A solid hard-exclusive calculation that gets the large-recoil scaling laws and a sharp zero prediction right, with a soft ln² coefficient the authors themselves leave open.","tokens_in":30752,"tokens_out":1666,"would_cite":true,"duration_ms":27260,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["12.38.Bx","12.39.Hg","13.40.Gp","14.40.Nd"],"model":"selfhosted/deepseek-v4.1-flash","headline":"Heavy-meson form factors are argued to change sign at large recoil, with an exact zero in one $B^*$ channel.","keywords":["heavy quark effective theory","Isgur-Wise function","light-cone wave functions","hard exclusive processes","QCD sum rules","heavy-meson form factors","large recoil asymptotics","Brodsky-Lepage evolution"],"falsifier":"A lattice determination of the $B^* to B^*$ matrix element of the heavy-quark vector current for longitudinally polarized mesons at $vcdot v'$ between about 2 and 10 that stays positive throughout would kill the exact-zero claim; conversely, observing the sign change and its approximate location near $m_b/barLambda$ would confirm it.","tokens_in":31106,"feed_emoji":"⚛️","tokens_out":3146,"duration_ms":58573,"temperature":0.7,"pith_summary":"The paper asks what a heavy meson's transition form factor looks like when the two mesons fly apart with a very large relative velocity, a regime far outside ordinary semileptonic decays but reachable in principle in $e^+e^-$ pair production. Using the hard-scattering machinery developed for exclusive processes such as the pion form factor, the authors argue that the answer is governed by the meson's light-cone wave functions rather than by any single universal function. Two different gluon-exchange mechanisms compete: a longitudinal gluon that flips the meson helicity and is sensitive to the interference of leading- and subleading-twist wave functions, and a transverse gluon that conserves helicity and is sensitive only to the leading-twist wave function. Because the two mechanisms scale differently with $vcdot v'$ and with the heavy-quark mass, their relative weight shifts as recoil grows, and the interference between them is what the authors claim produces sign changes in specific form factors. If the argument holds, the form factor for two longitudinally polarized $B^*$ mesons must pass through zero at some $vcdot v'$ of order $m_b/Lambda$, and the $B$ meson form factor in the timelike region must do the same near $vcdot v' sim -m_b/Lambda$.","feed_headline":"Heavy-meson form factors must flip sign at high recoil","feed_subtitle":"Two competing gluon-exchange mechanisms force an exact zero in one B* channel and an approximate one in B.","key_machinery":"The load-bearing objects are the two light-cone wave functions $phi_pm(omega)$ of a heavy meson, defined as Fourier transforms of the matrix element of the bilocal operator $Q^*(0)E(0,z)q(z)$ between the vacuum and the meson state, where $omega$ is the light-quark light-cone momentum in the meson rest frame. Their first moments are fixed exactly by the equations of motion, $langleomegarangle_+ = tfrac{4}{3}barLambda$ and $langleomegarangle_- = tfrac{2}{3}barLambda$, and their second moments are tied to the chromoelectric and chromomagnetic parameters $lambda_E^2$ and $lambda_H^2$. The calculational engine is the hard-gluon-exchange factorization familiar from exclusive QCD: in the Breit frame the gluon virtuality is $k_g^2 simeq -omegaomega' e^{vartheta}$, so for large rapidity the amplitude factorizes into a perturbative kernel convolved with the wave functions. The paper also derives the Brodsky-Lepage-type evolution equations for $phi_pm(omega)$, with kernels given explicitly at one loop, and shows that the renormalization of these non-local operators does not by itself renormalize the tower of local moments, so an extra divergence must be subtracted separately. QCD sum rules, including the non-locality of the quark condensate, supply numerical estimates of $lambda_E^2$, $lambda_H^2$, and the shapes of $phi_pm$.","core_discovery":"The central claim is that the large-recoil asymptotics of heavy-meson form factors is not exhausted by the Isgur-Wise function but is a sum of two contributions with different origins and different scaling. For $1 ll |vcdot v'| ll m_Q/Lambda$, the Isgur-Wise function $xi$ dominates and is itself generated by interference between the leading-twist wave function $phi_+$ on one side of the hard-gluon exchange and the subleading-twist wave function $phi_-$ on the other, giving $xi sim alpha_s f^2/Lambda^3,(vcdot v')^{-2}$ up to logarithmic enhancement. For $|vcdot v'| gg m_Q/Lambda$, two order-$1/m_Q$ functions, $xi_3$ and $chi_2$, take over; both are built from the leading-twist wave function alone, both conserve meson helicity, and both fall only as $1/(vcdot v')$ with an extra factor $m_Q$. The authors show that $chi_2 = -xi_3 e^{-vartheta}$, so that in the physical form factors the combination $xi pm 2m_b xi_3$ appears with a relative minus sign between the two terms. Since $xi$ falls faster than $m_b xi_3$ at large recoil, the combination must cross zero. The paper states this as an exact statement for the longitudinally polarized $B^*$ form factor at spacelike momentum transfer, and as an approximate one for the $B$ form factor in the timelike region where the form factors acquire a small imaginary part.","pith_inferences":["The predicted zero in $h_L$ is a clean lattice target: a direct lattice computation of the $B^* to B^*$ matrix element of the heavy-quark vector current at several values of $vcdot v'$ between 1 and roughly 10 would either show the sign change or fail to, without needing any wave-function model.","Because the zero location scales as $m_Q/Lambda$, comparing the $B$ and $D$ systems gives a further check: the charm zero should sit at a proportionally smaller $vcdot v'$, so a joint fit of both would test the $m_Q$ scaling rather than just the sign.","The whole chain leans on the endpoint behaviour of $phi_pm$ at small $omega$; a sum-rule or lattice determination of the low-$omega$ slope of $phi_+$ and the low-$omega$ constant of $phi_-$ would test the $ln^2(vcdot v')$ enhancement directly, independently of the form-factor zero.","The relative sign $chi_2 = -xi_3 e^{-vartheta}$ follows from helicity conservation in the transverse-gluon channel; any process that breaks the identification of helicity with the light-quark spin projection, such as a non-zero light-quark mass correction, would shift the zero and could be looked for as a small displacement."],"forward_implications":["The form factor $h_L$ for two longitudinally polarized $B^*$ mesons, positive at small recoil, must become negative beyond $vcdot v' sim m_b/Lambda$, giving an exact zero in the spacelike region.","The $B$ meson form factor $h_+$ acquires an approximate zero in the timelike region inside the physical phase space for $e^+e^- to Bbar B$, which would show up as a dip in the $Bbar B$ production rate near $sqrt{s}/m_B sim 25$.","Cross-section ratios for $e^+e^- to B^{(*)}bar B^{(*)}$ change their scaling pattern across the turnover: below the zero they scale as $1/x^8$, and above it the $xi_3$ term makes the $Bbar B$ and $B^*_Lbar B^*_L$ channels scale as $1/x^4$.","The same structure predicts a zero in the longitudinally polarized vector-meson channel at spacelike momentum transfer for light mesons such as the $rho$, matching what is already known for $pi$-$rho$ form factors.","The moments of the wave functions are fixed by the equations of motion and by the sum-rule estimates $lambda_E^2 = (0.11 pm 0.06)$ GeV$^2$ and $lambda_H^2 = (0.18 pm 0.07)$ GeV$^2$, giving a concrete normalization for the asymptotic formulae."],"supporting_citations":[{"why":"Supplies the covariant trace formalism and the leading-order Isgur-Wise parametrization of heavy-meson matrix elements that the whole calculation is built on.","marker":"[15]"},{"why":"Provides the order-$1/m_Q$ current and Lagrangian corrections and the definitions of $xi_3$ and the $chi_i$ functions whose asymptotics the paper computes.","marker":"[16]"},{"why":"The Brodsky-Lepage evolution equation that the paper's wave-function evolution equations are modelled on.","marker":"[28]"},{"why":"The hard exclusive QCD treatment of $pi$-$rho$ form factors, which supplies the helicity-flip versus helicity-conserving logic and the precedent for a zero in a longitudinally polarized vector form factor.","marker":"[29]"},{"why":"The QCD sum rule for the Isgur-Wise function, used as the baseline whose large-recoil limit is checked against the new asymptotic result and as the source of the continuum threshold.","marker":"[18]"},{"why":"Two-loop sum-rule expressions for the Isgur-Wise function whose large-$vartheta$ limit is shown to reproduce the paper's asymptotic behaviour.","marker":"[21]"},{"why":"Two-loop sum-rule expressions for $xi_3$ and $chi_2$, used to confirm both the asymptotic scaling and the relation $chi_2 = -xi_3 e^{-vartheta}$.","marker":"[24]"},{"why":"Supplies the non-local quark condensate and its virtuality distribution, needed for the sum rules for the wave functions themselves rather than just their moments.","marker":"[43]"}],"fun_headline_variants":["Two mechanisms, not just Isgur-Wise, set heavy-meson form factors","Isgur-Wise alone does not exhaust heavy-meson form factors at large recoil","An exact zero appears in the longitudinally polarized B* form factor","The B form factor has an approximate zero in the timelike region","Leading- and subleading-twist waves interfere to generate the Isgur-Wise function"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The whole scaling story rests on the assumed behaviour of the wave functions at vanishing light-quark momentum, $phi_+(omega) sim omega$ and $phi_-(omega) sim 1$ as $omega to 0$, which makes the relevant integrals diverge logarithmically and produces the $ln^2(vcdot v')$ enhancement; the paper motivates this from QCD analogies and from its own sum-rule model but does not derive it from first principles.","fun_headline_variants_meta":{"raw":{"variants":["Two mechanisms, not just Isgur-Wise, set heavy-meson form factors","Isgur-Wise alone does not exhaust heavy-meson form factors at large recoil","An exact zero appears in the longitudinally polarized B* form factor","The B form factor has an approximate zero in the timelike region","Leading- and subleading-twist waves interfere to generate the Isgur-Wise function"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":1673,"prompt_tokens":1123,"completion_tokens":550,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":458}},"tokens_in":1123,"tokens_out":550,"duration_ms":8690,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":458,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-29T19:15:00.321269+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"A lattice determination of the $B^* to B^*$ matrix element of the heavy-quark vector current for longitudinally polarized mesons at $vcdot v'$ between about 2 and 10 that stays positive throughout would kill the exact-zero claim; conversely, observing the sign change and its approximate location near $m_b/barLambda$ would confirm it.","supporting_citations":[{"cited_title":"Nucl. Phys. B 343, 1","cited_arxiv_id":null,"evidence_quote":"Supplies the covariant trace formalism and the leading-order Isgur-Wise parametrization of heavy-meson matrix elements that the whole calculation is built on."},{"cited_title":"Phys. Lett. B 252, 447","cited_arxiv_id":null,"evidence_quote":"Provides the order-$1/m_Q$ current and Lagrangian corrections and the definitions of $xi_3$ and the $chi_i$ functions whose asymptotics the paper computes."},{"cited_title":"Phys. Lett. B 87, 359 (1979); Phys. Rev. D 22, 2157","cited_arxiv_id":null,"evidence_quote":"The Brodsky-Lepage evolution equation that the paper's wave-function evolution equations are modelled on."},{"cited_title":"Phys. Rep. 112, 173","cited_arxiv_id":null,"evidence_quote":"The hard exclusive QCD treatment of $pi$-$rho$ form factors, which supplies the helicity-flip versus helicity-conserving logic and the precedent for a zero in a longitudinally polarized vector form factor."},{"cited_title":"Phys. Rev. D 45, 2451","cited_arxiv_id":null,"evidence_quote":"The QCD sum rule for the Isgur-Wise function, used as the baseline whose large-recoil limit is checked against the new asymptotic result and as the source of the continuum threshold."},{"cited_title":"Phys. Rev. D 47, 4063","cited_arxiv_id":null,"evidence_quote":"Two-loop sum-rule expressions for the Isgur-Wise function whose large-$vartheta$ limit is shown to reproduce the paper's asymptotic behaviour."},{"cited_title":"Phys. Lett. B 301, 101 (1993); Phys. Rev. D 47, 5060 (1993); 49, 1302","cited_arxiv_id":null,"evidence_quote":"Two-loop sum-rule expressions for $xi_3$ and $chi_2$, used to confirm both the asymptotic scaling and the relation $chi_2 = -xi_3 e^{-vartheta}$."},{"cited_title":"JETP Lett. 43, 712 (1986); Yad. Fiz. 49, 794 (1989) [Sov. J. Nucl. Phys. 49, 494 (1989)]; Phys. Rev. D 45, 1754","cited_arxiv_id":null,"evidence_quote":"Supplies the non-local quark condensate and its virtuality distribution, needed for the sum rules for the wave functions themselves rather than just their moments."}],"review_version":1}