{"id":"cc88c591-0e19-4fe5-b310-d0076e6e2976","arxiv_id":"2501.00550","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"The paper computes leading-order electroproduction cross sections for Bc, Bs, and Ds mesons and predicts that the EIC could detect Bc with tens of events per year, whereas HERA could not.","lead":"Scientists calculated how often electron-proton collisions at colliders like HERA and the future EIC would create Bc, Bs, and Ds mesons. They found Bc production would be measurable at the EIC but not at HERA, while Bs and Ds rates are large and depend strongly on quark mass choices.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The B_c observability claim rests on a LO amplitude whose Dirac algebra was simplified with manual contracting rules because FeynCalc 'gives wrong results'; without an independent amplitude check, the 5–28 events/year conclusion is numerically insecure.","rationale":"The paper is a straightforward leading-order NRQCD calculation of electroproduction of B_c, B_s, and D_s, with the headline result that B_c is observable at the EIC through 5–28 reconstructed events per year. For that claim to hold, the LO color-singlet cross section must be numerically correct. The manuscript itself contains a load-bearing disclosure: FeynCalc produced wrong results for long Dirac chains and the authors manually added contraction rules. Because no code or independent calculation is provided, the numerical output has no external check. This is a sharper and more specific concern than the generic NLO caveat: if the LO amplitude is miscomputed, the NLO discussion is moot, whereas if the amplitude is correct, the LO-only limitation still leaves the yield uncertain at the factor level. The reader identified the LO-only approximation as the weakest assumption but also noted the FeynCalc issue and the absence of code; my concern overlaps only partially because I place the primary weight on the unverified amplitude algebra rather than on missing higher orders. The appropriate disposition remains conditional: the physics case is plausible and interesting, but the quantitative event-rate claim should not be relied upon until an independent numerical check of the LO amplitude is performed.","tokens_in":930,"tokens_out":1117,"duration_ms":109301,"concrete_test":"Rebuild the e g → e B_c μc b squared amplitude in a second independent algebraic framework (e.g., a FORM-based trace with a different Dirac-contraction scheme, or FormCalc) and, with identical inputs (m_c = 1.5 GeV, m_b = 4.9 GeV, CT10 PDF, and the HERA and EIC cuts specified in Sec. III), recompute σ(B_c) and σ(B_c*). If the independent result reproduces Eqs. (9) and (10) within about 10%, the manual-contraction concern is resolved; if it does not, the EIC observability claim rests on an unverified amplitude.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central, falsifiable claim is in Sec. III, Eq. (10): σ(B_c) = 5.44 fb and σ(B_c*) = 22.12 fb at the EIC, leading to 3922–19209 produced B_c per year and 5–28 reconstructed events per year. Every digit of this estimate depends on the LO squared amplitude for γ* + g → B_c + μc + b. In Sec. II the authors state that FeynCalc 'seems to give wrong results when contracting long chains of Dirac gamma matrix, and can be solved by manually add contracting rules.' No code, ancillary data, or independent cross-check is provided, and the quoted uncertainty bands are generated by mass/scale variations of this same unchecked amplitude. A single sign or factor error in the manual contraction rules changes σ(B_c) by an O(1) factor; a factor-of-three downward shift would reduce the central reconstructed yield from roughly 10 events/year to about 3, and a larger error would remove the observability claim entirely. The NLO/color-octet limitation is real but secondary: unless the LO amplitude itself is verified, higher-order corrections cannot rescue or condemn the numerical prediction.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents leading-order (LO) NRQCD predictions for the electroproduction of Bc, Bc*, Bs, Bs*, Ds, and Ds* in electron-proton collisions at HERA and at the EIC. The calculation uses standard color-singlet factorization with a gluon PDF, covariant projection operators, and long-distance matrix elements taken from potential-model wave functions for Bc and from decay constants for Bs and Ds. The central phenomenological claim is that Bc production is observable at the EIC, with a predicted cross section of 5.44 fb plus a Bc* cross section of 22.12 fb (Section III, Eq. 10), yielding roughly 5 to 28 reconstructed Bc -> J/psi pi -> l+l- events per year. The paper also reports much larger but strongly parameter-sensitive cross sections for Bs and Ds at HERA.","tokens_in":9214,"tokens_out":2611,"duration_ms":29148,"significance":"If the numerical predictions are correct, the paper provides a concrete, falsifiable observable for the EIC and fills a gap in LO electroproduction calculations for mesons containing one or two heavy quarks. The work has several strengths: the formalism is standard NRQCD; the input parameters are taken from external sources rather than fitted to the target observables; the paper explicitly separates quark-mass and scale uncertainties; and it gives differential distributions in cos(theta) and pT^2. The central event-rate estimate, however, rests on an amplitude calculation that the authors report required manual patching of FeynCalc's Dirac contractions, and no independent verification or code is provided. Since the EIC observability claim scales linearly with the LO cross section, the numerical security of that claim is presently unverified.","major_comments":[{"comment":"The central Bc cross-section prediction, and hence the 5 to 28 reconstructed events per year at the EIC, depends entirely on the LO squared amplitude for gamma* + g -> Bc + cbar + b. In Section II the authors state that 'Feyncalc seems to give wrong results when contracting long chains of Dirac gamma matrix, and can be solved by manually add contracting rules.' No code, ancillary file, or independent amplitude check is provided, so the reader cannot verify that the manual contraction rules do not introduce a sign or factor error. An O(1) error would change the quoted cross sections in Eq. (10) by a factor that could remove or greatly inflate the observability claim. I therefore request an independent verification, for example a comparison of the squared amplitude in a known kinematic limit, a helicity-amplitude check, or release of the calculation code, before the central claim can be assessed.","section":"Section II and Eq. (10)"},{"comment":"The quoted Bc and Bc* cross sections are LO color-singlet predictions only; NLO QCD corrections and color-octet contributions are omitted. Because the event-rate estimate is directly proportional to the LO cross section, the absence of any estimate of these corrections leaves the central claim with an unquantified systematic uncertainty. The large scale variation in Eq. (10) (the +6.36/-1.91 fb band on the Bc cross section) already spans a factor of several, and NLO corrections in comparable quarkonium leptoproduction calculations are known to be substantial. The authors should either provide an estimate of the expected size of NLO/octet contributions or explicitly restrict the conclusion to 'an LO estimate suggests observability.'","section":"Section III, Eq. (10)"},{"comment":"For Bs and Ds, the NRQCD velocity expansion is applied to mesons containing a strange quark with ms = 0.5 GeV, for which v is not parametrically small. The paper itself demonstrates the fragility of this assumption: Eq. (11) shows that a 0.1 GeV change in quark mass can alter the Bs cross section by more than 200%, and the total uncertainty bands are very wide. The abstract and conclusion describe these Bs and Ds cross sections as 'notably significant,' but without a discussion of the convergence of the NRQCD expansion for light strange quarks, the quoted event numbers are difficult to interpret as more than order-of-magnitude estimates. I recommend either adding a quantitative discussion of the v-expansion validity for cs and bs systems or softening the claims accordingly.","section":"Section III, Eq. (11) and Bs/Ds discussion"}],"minor_comments":[{"comment":"There are several typos: 'chromodyn amics' should be 'chromodynamics', and the affiliation 'University of Chinese Academy of Science s' has an extra 's'.","section":"Abstract and affiliations"},{"comment":"The lower uncertainty for sigma_Ds is printed as '-3,03 pb' and should presumably be '-3.03 pb'.","section":"Eq. (11)"},{"comment":"The figure captions for Figs. 2-4 describe the upper/lower bounds in terms of specific quark mass combinations, but it would be clearer to state that the upper/lower curves correspond to the full mass-and-scale envelope rather than to a single combination.","section":"Section III, figure captions"},{"comment":"The statement that the EicC cross section is 'less than or similar to 0.03 fb' is given without a formula or reference; a brief derivation or citation would improve transparency.","section":"Section III, EicC discussion"},{"comment":"The phase-space decomposition notation in Eq. (3) is hard to follow; in particular the meaning of the integration limits 'E2/(4m^2)' and 'M^2/m^2' is not explained. A sentence defining M and the limits would help.","section":"Eq. (3)"}],"recommendation":"major_revision","confidential_remarks":"The paper is potentially interesting for the EIC community, but the manual FeynCalc patch is a serious reproducibility concern. I would encourage the editors to require an independent numerical check or code release before publication. The Bs/Ds part is more exploratory and should be framed accordingly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a legitimate first pass at electroproduction cross sections for Bc, Bs, and Ds at ep colliders, and the EIC feasibility number for Bc is new and worth noticing. The calculation is standard LO NRQCD, imported from the J/psi leptoproduction formalism, and the authors are appropriately modest about its limits. The central caveat is the one they disclose themselves: FeynCalc gave wrong Dirac contractions and they patched it by hand. Without an independent check or released code, I would not put much weight on the exact 5–28 events/year; a factor-of-three error would make the signal marginal, and a factor-of-ten error would kill it. That said, the stress-test headline overstates it a bit: the LO amplitude is not obviously wrong, the authors are transparent, and the conclusion is not razor-thin—even with a factor-three downward shift you still have a few events per year to start with at the EIC.\n\nWhat the paper does well: clear setup, explicit cuts and parameter choices, differential distributions in cos(theta) and pT^2, and uncertainty bands from mass and scale variation. The Bs/Ds results are presented honestly, with their large s-quark mass sensitivity flagged. No circularity: the LDMEs and decay constants are external inputs, not fit to the predicted rates.\n\nSoft spots: beyond the amplitude verification issue, the LO-only treatment is a real limitation; NLO and color-octet contributions could shift rates by factors. The Bs/Ds application of NRQCD to a 0.5 GeV strange quark is questionable, and the paper's own mass-scan shows >200% sensitivity, so those numbers are qualitative. No code or data files are provided, so independent reproduction is impossible.\n\nWho this is for: people planning heavy-quark measurements at the EIC and NRQCD practitioners. It is not a conceptual breakthrough, but it fills a real gap. I would send it to a serious referee. The right outcome is conditional acceptance: require either an independent amplitude check or release of the calculation scripts, and make sure the EIC rate claim is framed with the LO caveat and the FeynCalc patching.","headline":"First NRQCD electroproduction estimates for Bc, Bs, Ds at ep colliders; the Bc EIC rate is new but rests on an unverified LO amplitude, so treat the event numbers as indicative.","tokens_in":9968,"tokens_out":3727,"would_cite":true,"duration_ms":39241,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["13.60.-r","12.38.Bx"],"model":"deepseek-v4-flash","headline":"The paper claims that electroproduced $B_c$ mesons should be observable at the EIC, predicting 5–28 reconstructed events per year, while HERA cannot accumulate a reconstructable sample.","keywords":["B_c meson","electroproduction","NRQCD","Electron-Ion Collider","HERA","B_s meson","D_s meson","leading-order cross section"],"falsifier":"Run the EIC for one nominal year (about 315 fb$^{-1}$) and search for $e+p\\to e+B_c+\\bar c+b$ with $B_c^\\pm\\to J/\\psi\\pi^\\pm$ and $J/\\psi\\to l^+l^-$: fewer than about five reconstructed events, or a central cross-section value more than a factor of three below 5.44 fb, would falsify the central claim. Before any collider data exist, a lattice-QCD determination of the $B_c$ decay constant that contradicts the potential-model input $|R_{B_c}(0)|^2 = 1.642$ GeV$^3$ would undercut the same prediction.","tokens_in":8767,"feed_emoji":"⚛️","tokens_out":8237,"duration_ms":72592,"temperature":0.7,"pith_summary":"This paper predicts that the $B_c$ meson — a bound state of a bottom and a charm quark, and the only meson whose decay is purely weak — should be producible in observable numbers at the future Electron-Ion Collider (EIC), but not at the earlier HERA collider. Working at leading order in nonrelativistic QCD, the authors compute electroproduction cross sections of $B_c$ and $B_c^*$ along with the single-heavy-flavour mesons $B_s$ and $D_s$ at $ep$ colliders. At EIC energies the $B_c$ cross section is $\\sigma(B_c)=5.44$ fb, which through the $B_c\\to J/\\psi\\pi$, $J/\\psi\\to l^+l^-$ chain yields an estimated 5 to 28 reconstructed events per year — enough to make the first electroproduced $B_c$ observable. At HERA the $B_c$ rate is too small for reconstruction, while $B_s$ and $D_s$ are produced abundantly but with leading-order predictions that vary by more than a factor of two with quark mass and renormalisation scale.","feed_headline":"EIC should catch 5 to 28 B_c meson events per year","feed_subtitle":"A leading-order NRQCD prediction makes the weakly-decaying B_c observable at the EIC; HERA cannot.","key_machinery":"The calculation rests on the NRQCD factorisation formula $d\\sigma = \\int dx\\, f_{g/p}(x,\\mu^2)\\, d\\hat\\sigma\\, \\langle O_M(n)\\rangle$, which splits the process into a gluon PDF, a perturbative hard part, and a long-distance matrix element $\\langle O_M\\rangle = 2(2J+1)N_c |R(0)|^2/4\\pi$. The hard part is built from 24 tree-level $\\gamma^* g \\to q\\bar q' + q + \\bar q'$ diagrams evaluated with covariant projection operators ($\\gamma_5$ for pseudoscalar, $\\not\\epsilon$ for vector), with the leptonic tensor folded in and the four-body phase space decomposed into sequential two-body phase spaces for Monte Carlo efficiency. The long-distance matrix elements are fixed at leading order by $|R(0)|^2$ from a potential-model wave function for $B_c$ and from the decay constants $f_{B_s}$, $f_{D_s}$ for the single-heavy mesons; heavy-quark spin symmetry gives vector and pseudoscalar mesons the same $R(0)$. The comparison numbers come from the CT10 gluon PDF, HERA cuts ($p_T>1$ GeV, $0.3<z<0.9$, $2<Q^2<100$ GeV$^2$) and EIC Yellow Report cuts ($Q^2>1$ GeV$^2$, $20<W<80$ GeV, $0.05<z<0.9$).","core_discovery":"The central claim is that leading-order NRQCD electroproduction of the $B_c$ meson at the EIC is observable: with all $B_c^*$ decays feeding $B_c$, the predicted cross sections $\\sigma(B_c)=5.44$ fb and $\\sigma(B_c^*)=22.12$ fb translate into 3922–19209 $B_c$ mesons per year and, after the $B_c\\to J/\\psi\\pi^\\pm$ (0.5%) and $J/\\psi\\to l^+l^-$ (12%) branching cuts and counting both charge states, about 5–28 reconstructed events per year. The same calculation at HERA gives a larger cross section (42.13 fb) but insufficient accumulated events for reconstruction, while at the proposed EicC the cross section is below 0.03 fb. For $B_s$ and $D_s$, the paper reports that HERA produces them abundantly ($\\sigma(B_s)=0.29$ pb, $\\sigma(D_s)=9.35$ pb at leading order) but that these predictions are highly sensitive to the strange-quark mass and renormalisation scale, a sensitivity the authors read as a possible experimental window rather than as a failure.","pith_inferences":["The event-number claim is effectively a leading-order, colour-singlet estimate; if next-to-leading-order or colour-octet corrections shift the cross section down by roughly a factor of three, the 5–28 event prediction would essentially vanish, so the estimate is best read as an order-of-magnitude target awaiting NLO confirmation.","Because the $B_c^*$ cross section is about four times the $B_c$ one and feed-down is assumed complete, the same calculation predicts the EIC sample to be dominated by $B_c$ from vector-meson decay; a polarisation or momentum-distribution measurement of reconstructed $B_c$ would test that feed-down assumption directly.","The dramatic quark-mass sensitivity of the $B_s$ and $D_s$ cross sections (greater than 200% for a 0.1 GeV mass shift) could be inverted: precision $D_s$ electroproduction data at a future $ep$ collider would constrain the strange-quark mass or delimit where NRQCD's velocity expansion breaks down for single-heavy mesons."],"forward_implications":["At the EIC, the predicted 3922–19209 $B_c$ mesons per year translate into roughly 5–28 reconstructed $B_c^\\pm$ events per year through $B_c\\to J/\\psi\\pi^\\pm$ with $J/\\psi\\to l^+l^-$, making the $B_c$ a reachable target for the first time in electroproduction.","At HERA the $B_c$ cross section is larger in absolute terms (42.13 fb) but the integrated luminosity is insufficient for reconstruction, so HERA cannot test the prediction.","The proposed EicC collider is ruled out for this purpose: its $B_c$ cross section is below 0.03 fb.","At HERA, $B_s$ and $D_s$ production is plentiful (roughly 48–189 $B_s^0$, 119–476 $B_s^*$, 1575–5100 $D_s^+$, and 3465–11440 $D_s^{*+}$ events per 315 pb$^{-1}$), enough for the main decay channels.","An order-of-magnitude luminosity upgrade (HL-EIC) would raise the expected $B_c$ yield correspondingly, strengthening the case for observation."],"supporting_citations":[{"why":"Bodwin–Braaten–Lepage NRQCD factorisation: supplies the decomposition of the cross section into short-distance coefficients and long-distance matrix elements on which the whole calculation rests.","marker":"[4]"},{"why":"EIC Yellow Report: provides the EIC beam energies, luminosity of $10^{34}$ cm$^{-2}$s$^{-1}$, and kinematic cuts used for the $σ_{\\rm EIC}$ numbers and the yearly event estimate.","marker":"[25]"},{"why":"Sun and Zhang leptoproduction study: gives the leptonic-tensor treatment and phase-space decomposition used to make the four-body Monte Carlo integral tractable.","marker":"[28]"},{"why":"PDG review: supplies the decay constants $f_{B_s}$ and $f_{D_s}$, the quark masses and $Λ_{\\rm QCD}$ inputs, and the $J/\\psi\\to l^+l^-$ branching ratio used in the event estimate.","marker":"[32]"},{"why":"Eichten and Quigg spectroscopy: computes $B_c$ bound-state properties with the potential model, source of $|R_{B_c}(0)|^2 = 1.642$ GeV$^3$.","marker":"[33]"},{"why":"Buchmüller and Tye potential model: the potential whose wave function at the origin sets the $B_c$ long-distance matrix element.","marker":"[34]"},{"why":"CT10 parton distribution functions: supplies the gluon density $f_{g/p}$ used in the factorisation formula for both HERA and EIC kinematics.","marker":"[35]"},{"why":"Chang and Chen $B_c$ decays: source of the $B_c^\\pm \\to J/\\psi\\pi^\\pm$ branching ratio of about 0.5% used to convert produced $B_c$ into reconstructed events.","marker":"[36]"}],"fun_headline_variants":["EIC could log 5–28 B_c events yearly, HERA can't","NRQCD: B_c visible at EIC, not HERA; B_s, D_s suited to HERA","5–28 B_c events per year predicted at EIC","B_c observable at EIC, HERA too sparse; B_s, D_s abundant","Leading-order NRQCD yields 5–28 B_c events at EIC"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole rate estimate stands or falls on the assumption that a leading-order, colour-singlet NRQCD calculation, seeded by a potential-model wave function at the origin with $|R_{B_c}(0)|^2 = 1.642$ GeV$^3$ and assuming every $B_c^*$ decays to $B_c$, is accurate to within a factor of about three.","fun_headline_variants_meta":{"raw":{"variants":["EIC could log 5–28 B_c events yearly, HERA can't","NRQCD: B_c visible at EIC, not HERA; B_s, D_s suited to HERA","5–28 B_c events per year predicted at EIC","B_c observable at EIC, HERA too sparse; B_s, D_s abundant","Leading-order NRQCD yields 5–28 B_c events at EIC"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000262,"raw_usage":{"total_tokens":1592,"prompt_tokens":934,"completion_tokens":658,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":550,"completion_tokens_details":{"reasoning_tokens":545}},"tokens_in":550,"tokens_out":658,"duration_ms":5234,"temperature":1.0,"reasoning_tokens":545,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:48:07.179721+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the EIC for one nominal year (about 315 fb$^{-1}$) and search for $e+p\\to e+B_c+\\bar c+b$ with $B_c^\\pm\\to J/\\psi\\pi^\\pm$ and $J/\\psi\\to l^+l^-$: fewer than about five reconstructed events, or a central cross-section value more than a factor of three below 5.44 fb, would falsify the central claim. Before any collider data exist, a lattice-QCD determination of the $B_c$ decay constant that contradicts the potential-model input $|R_{B_c}(0)|^2 = 1.642$ GeV$^3$ would undercut the same prediction.","supporting_citations":[{"cited_title":"Quarkonia and Quantum Ch romodynamics,","cited_arxiv_id":null,"evidence_quote":"CT10 parton distribution functions: supplies the gluon density $f_{g/p}$ used in the factorisation formula for both HERA and EIC kinematics."}],"review_version":1}