Mesons with Beauty and Charm: New Horizons in Spectroscopy

Estia J. Eichten∗ and Chris Quigg† Fermi National Accelerator Laboratory P.O. Box 500, Batavia, Illinois 60510 USA (Dated: June 18, 2019) + ¯ The Bc family of (cb) mesons with beauty and charm is of special interest among heavy quarko- + ¯ nium systems. The Bc mesons are intermediate between (cc¯) and (bb) states both in mass and size, so many features of the (c¯b) spectrum can be inferred from what we know of the charmonium and bottomonium systems. The unequal quark masses mean that the dynamics may be richer than a simple interpolation would imply, in part because the charmed quark moves faster in Bc than in the + J/ψ. Close examination of the Bc spectrum can test our understanding of the interactions between heavy quarks and antiquarks and may reveal where approximations break down. Whereas the J/ψ and Υ levels that lie below flavor threshold are metastable with respect to strong decays, the Bc ground state is absolutely stable against strong or electromagnetic decays. Its dominant weak decays arise from ¯b → cW¯ ?+, c → sW ?+, and c¯b → W ?+ transitions, where W ? designates a virtual weak boson. Prominent examples of the first category are quarkonium + + + + transmutations such as Bc → J/ψπ and Bc → J/ψ` ν`, where J/ψ designates the (cc¯) 1S level. The high data rates and extraordinarily capable detectors at the Large Hadron Collider give renewed impetus to the study of mesons with beauty and charm. Motivated by the recent exper- imental searches for the radially excited Bc states, we update the expectations for the low-lying spectrum of the Bc system. We make use of lattice QCD results, a novel treatment of spin split- tings, and an improved quarkonium potential to obtain detailed predictions for masses and decays. We suggest promising modes in which to observe excited states at the LHC. The 3P and 3S states, which lie close to or just above the threshold for strong decays, may provide new insights into the mixing between quarkonium bound states and nearby two-body open-flavor channels. Searches in the B(∗)D(∗) final states could well reveal narrow resonances in the J P = 0−, 1−, and 2+ channels and possibly in the J P = 0+ and 1+ channels at threshold. Looking further ahead, the prospect of very-high-luminosity e+e− colliders capable of producing tera-Z samples raises the possibility of investigating Bc spectroscopy and rare decays in a controlled environment.

PACS numbers: 14.40.Lb, 14.40.Nd, 14.40.Pq FERMILAB–PUB–19/075–T

I. INTRODUCTION the copious production of b-quarks at Fermilab’s Teva- tron Collider and CERN’s Large Electron–Positron Col- Although the lowest-lying (c¯b) meson has long been es- lider (LEP), we presented a comprehensive portrait of tablished, the spectrum of excited states is little explored. the spectroscopy of the Bc meson and its long-lived ex- The ATLAS experiment at CERN’s Large Hadron Col- cited states [6], based on then-current knowledge of the lider reported the observation of a radially excited B interaction between heavy quarks derived from (cc¯) and c ¯ state [1], but this sighting was not confirmed by the LHCb (bb) bound states, within the framework of nonrelativis- experiment [2]. The unsettled experimental situation and tic quantum mechanics [7]. Surveying four representative P − the large data sets now available for analysis make it potentials, we characterized the mass of the J = 0 timely for us to provide up-to-date theoretical expecta- ground state as M(Bc) ≈ 6258 ± 20 MeV. A small 0 arXiv:1902.09735v2 [hep-ph] 15 Jun 2019 tions for the spectrum and decay patterns of narrow (c¯b) number of Bc candidates appeared in hadronic Z de- states, and for their production in hadron colliders [3]. cays at LEP. The CDF Collaboration observed the de- ± ± New work from the CMS Collaboration [4] shows the cay Bc → J/ψ` ν in 1.8-TeVpp ¯ collisions at the Fer- way toward exploiting the potential of (c¯b) spectroscopy. milab Tevatron [8], estimating the mass as M(Bc) ≈ 6400 ± 411 MeV. (The generic lepton ` represents an electron or muon.) Subsequent work by the CDF [9],

A. What we know of the Bc mesons D0 [10], and LHCb [11] Collaborations has refined the mass to M(Bc) = 6274.9 ± 0.8 MeV [12], with the most precise determinations coming from fully reconstructed The possibility of a spectrum of narrow Bc states was + first suggested by Eichten and Feinberg [5]. Anticipating final states such as J/ψπ . Investigations based on the spacetime lattice formu- lation of QCD aim to provide ab initio calculations ∗ Electronic mail: [email protected]; ORCID: 0000-0003-0532-2300 that incorporate the full dynamical content of the the- † Electronic mail: [email protected]; ORCID: 0000-0002-2728-2445 ory of strong interactions. Before the nonleptonic Bc 2 decays had been observed, a first unquenched lattice B. Analyzing the (c¯b) bound states QCD prediction, incorporating 2 + 1 dynamical quark +18 flavors (u/d, s) found M(Bc) = 6304 ± 12−0 MeV [13], The nonrelativistic potential picture, motivated by the where the first error bar represents statistical and sys- asymptotic freedom of QCD [20], gave early insight into tematic uncertainties and the second characterizes heavy- the nature of charmonium and generated a template for quark discretization effects. Calculations incorporating the spectrum of excited states [21]. For more than four 2 + 1 + 1 dynamical quark flavors (u/d, s, c) [14] yield decades, it has served as a reliable guide to quarkonium 1 M(1 S0) = 6278 ± 9 MeV, in impressive agreement with spectroscopy, including the states lying near or just above 1 the measured Bc mass, and predict M(2 S0) = 6894 ± flavor threshold for fission into two heavy-light mesons 19 ± 8 MeV [15]. that are significantly influenced by coupled-channel ef- fects [22, 23]. Three distinct elementary processes contribute to the ¯ ?+ We view the nonrelativistic potential-model treatment decay of Bc: the individual decays b → cW¯ and c → sW ?+ of the two heavy constituents, and the an- as a steppingstone, not a final answer, however impressive nihilation c¯b → W ?+ through a virtual W -boson. Sev- its record of utility. Potential theory does not capture eral examples of the ¯b → c¯ transition have been ob- the full dynamics of the strong interaction, and while the + + + standard coupled-channel treatment is built on a plausi- served, including the final states J/ψ` ν`, J/ψπ , J/ψK , + + − + + − + + + − − + ble physical picture, it is not derived from first principles. J/ψπ π π , J/ψπ K K , J/ψπ π π π π , J/ψDs , ?+ + + Moreover, relativistic effects may be more important for J/ψDs , and J/ψπ pp¯ ; and ψ(2S)π . A single channel, ¯ 0 + (cb) than for (cc¯). The c-quark moves faster in the Bc me- Bs π , representing the c → s transition is known. The annihilation mechanism, which would lead to final states son than in the J/ψ, because it must balance the momen- such as τ +ν andppπ ¯ +, has not yet been established. tum of a more massive b-quark. One developing area of τ theoretical research has been to explore methods more ro- The observed lifetime, τ(Bc) = (0.507 ± 0.009) ps [12], is consistent with theoretical expectations [16, 17]. Predic- bust than nonrelativistic quantum mechanics [17, 24, 25]. tions for partial decay rates (or relative branching frac- Nonperturbative calculations on a spacetime lattice in tions) await experimental tests. Some recent theoretical principle embody the full content of QCD. This approach works explore the potential of rare B decays [18]. is yielding increasingly precise predictions for the masses c ¯ ∗0 3 of (cb) levels up through Bc (2 S1) state. It is not yet Until recently, the only evidence reported for a (c¯b) possible to extract reliable signals for higher-lying states excited state was presented by the ATLAS Collabo- from the lattice, so we rely on potential-model methods ration [1] in pp collisions at 7 and 8 TeV, in samples to construct a template for the Bc spectrum through the −1 3 of 4.9 and 19.2 fb . They observed a new state at 4 S1 level. If experiments should uncover systematic de- ± + − ± ± 6842 ± 7 MeV in the M(Bc π π ) − M(Bc ) − 2M(π ) viations from the expectations we present, they may be ± ± mass difference, with Bc detected in the J/ψ π mode. taken as evidence of dynamical features absent from the The mass (527 ± 7 MeV above hM(1S)i) and decay of nonrelativistic potential-model paradigm, including—of this state are broadly in line with expectations for the course—coupling to states above flavor threshold, which ± second s-wave state, Bc (2S). In addition to the non- we neglect our calculations of the spectrum. relativistic potential-model calculations cited above, the In the following §II, we develop the theoretical tools re- HPQCD Collaboration has presented preliminary results quired to compute the (c¯b) spectrum. In earlier work [6], from a lattice calculation using 2 + 1 + 1 dynamical we examined the Cornell Coulomb-plus-linear poten- fermion flavors and highly improved staggered quark cor- tial [22], a power-law potential [26], Richardson’s QCD- 1 relators [19]. They report M(2 S0) = 6892 ± 41 MeV, inspired potential [27], and a second QCD-inspired po- which is 576.5 ± 41 MeV above hM(1S)i). This result tential due to Buchmueller and Tye [28], which we took and the NRQCD prediction [14] lie above the ATLAS as our reference model. We used a perturbation-theory report by one and two standard deviations, respectively. treatment of spin splittings. Using insights from lattice The significance of the discrepancy is limited for the mo- QCD and higher-order perturbative calculations, we con- ment by lattice uncertainties. A plausible interpretation struct a new potential that differs in detail from those has been that ATLAS might have observed the transition explored in earlier work. We also use lattice results and ∗ ∗ + − ¯ Bc (2S) → Bc (1S)π π , missing the low-energy photon rich experimental information on the (cc¯) and (bb) spec- ∗ from the subsequent Bc → Bcγ decay, and that the sig- tra to refine the treatment of spin splittings. We present 3 1 nal is an unresolved combination of 2 S1 and 2 S0 peaks. our expectations for the spectrum of narrow states in A search by the LHCb collaboration in 2 fb−1 of 8-TeV pp Section III. We consider decays of the narrow states in data yielded no evidence for either Bc(2S) state [2]. As section IV, updating the results we gave in Ref. [6]. We we prepared this article for publication, the CMS Collab- compute differential and integrated cross sections for the oration provided striking evidence for both Bc(2S) lev- narrow Bc levels in proton–proton collisions at the Large + − els, in the form of well-separated peaks in the Bcπ π Hadron Collider in §V. Putting all these elements to- invariant mass distribution, closely matching the theoret- gether, we show how to unravel the 2S levels and explore ical template [4]. We incorporate these new observations how higher levels might be observed. Prospects for a fu- into the discussion that follows in §V A. ture e+e− → Tera-Z machine appear in §VI. We draw 3 some conclusions and look ahead in Section VII.

II. THEORETICAL PRELIMINARIES

We take as our starting point a Coulomb-plus-linear potential (the “Cornell potential”[22]), ( r ) s

κ r α V (r) = − + , (1) r a2

−1 where κ ≡ 4αs/3 = 0.52 and a = 2.34 GeV were cho- sen to fit the quarkonium spectra. Analysis of the J/ψ and Υ families led to the choices

m = 1.84 GeV m = 5.18 GeV. (2) c b r [fm] This simple form has been modified to incorporate run- FIG. 1. Dependence of the running coupling αs(r) on the ning of the strong coupling constant in Refs. [27, 28], interquark separation r. The strong coupling for our chosen among others, using the perturbative-QCD evolution potential is shown in the solid red curve. Those corresponding equation at leading order and beyond. At distances rel- to the Cornell potential (green dots) [22], Richardson poten- evant for confinement, perturbation theory ceases to be tial (blue dash-dotted) [27] and an alternative version of the a reliable guide. It is now widely held, following Gri- new potential with µ = 0.8 GeV (gold dashes) are shown for bov [29], that as a result of quantum screening αs ap- comparison. proaches a critical, or frozen, value at long distances (low energy scales). In a light (qq¯) system, Gribov estimated

3π   α → α = 1 − p2/3 ≈ 0.14π = 0.44. (3) s s 4 We incorporate the spirit of this insight into a new version of the Coulomb-plus-linear form that we call the frozen- αs potential. The long-range part is the standard Cornell linear V ( r ) [GeV] term. To obtain the Coulomb piece, we convert the four- loop running of αs(q) in momentum space [30] to the be- havior in position space using the method of [31], with an important modification. We set αs(q = 1.6 GeV) = 0.338 and evolve with three active quark flavors. To enforce saturation of αs(r) at long distances, we alter the recipe of Ref. [31], replacing the identification q = 1/r exp(γE), where γE = 0.57721 ... is Euler’s constant, with the r [fm] 2 2 1/2 damped form q = 1/[(r exp(γE) + µ ] . For our ref- erence potential, we have chosen the damping parameter FIG. 2. Dependence of quarkonium potentials V (r) on the µ = 1.2 GeV. The consequent evolution of αs(r) is plot- interquark separation r. Our frozen-αs potential is shown in ted as the solid red curve in Figure 1, where we also show the solid red curve. The Cornell potential (green dots) [22], an alternative choice of µ = 0.8 GeV (dashed gold curve), Richardson potential (blue dash-dotted) [27], and an alter- native version of the new potential with µ = 0.8 GeV (gold the constant αs of the original Cornell potential (dotted dashes) are shown for comparison. green curve) and αs(r) corresponding to the Richardson potential (dot-dashed blue curve). We plot in Figure 2 the frozen-αs potential for both our chosen example, µ = 1.2 GeV, and the alternative, We presented the general formalism for spin-dependent µ = 0.8 GeV. There we also show the Richardson and interactions as laid out by Eichten & Feinberg [5] and Cornell potentials. All coincide at large distances. The Gromes [32] in § II B of Ref. [6], where we took a pertur- Cornell potential is deeper at short distances than any bative approach to the spin–orbit and tensor interactions. of the potentials that take account of the evolution of In the intervening time, the charmonium and bottomo- αs. For the convenience of others who may wish to apply nium spectra have been mapped in detail, as summarized the new potential, we present values of αs(r) suitable for in Table I. This wealth of information leads us now to interpolation in an Appendix. choose a more phenomenological approach. 4

TABLE I. P -state masses [12] and splittings, in MeV. TABLE II. Values of Te2 and Te4 extracted from data State 2P (cc¯) 2P (b¯b) 3P (b¯b) (underlined) and the inferred values of the phenomenologi- 3 cal coefficientsc ˜2 andc ˜4 for the J/ψ and Υ families, from χ0( P0) 3414.71 ± 0.30 9859.44 ± 0.52 10 232.5 ± 0.64 3 which the coefficients for the (c¯b) system are derived. χ1( P1) 3510.67 ± 0.05 9892.78 ± 0.4 10 255.46 ± 0.55 1 h( P1) 3525.38 ± 0.11 9899.3 ± 0.8 10 259.8 ± 1.12 D αs(r) E 3 1 dV 3 3 System Te2 Te4 3 [GeV ]c ˜2 c˜4 r dr [GeV ] χ2( P2) 3556.17 ± 0.07 9912.21 ± 0.4 10 268.85 ± 0.55 r 3 1P(cc¯) 0.088 0.0308 0.0527 1.25 1.77 0.141 PJ centroid, hχi 3525.29 ± 0.01 9899.87 ± 0.17 10 260.35 ± 0.31 1P(b¯b) 0.258 0.0835 0.220 0.82 0.99 0.383 h − hχi 0.09 ± 0.11 −0.57 ± 0.82 −0.55 ± 1.25 2P(b¯b) 0.181 0.0547 0.166 0.82 0.99 0.278 3 3 χ1( P1) − χ0( P0) 95.96 ± 0.30 33.34 ± 0.66 22.96 ± 0.84 1P(c¯b) 0.119 0.0388 0.0885 1.012 1.316 0.198 3 3 χ1( P2) − χ0( P1) 45.5 ± 0.09 19.43 ± 0.57 13.39 ± 0.78

For the J/ψ and Υ families, composed of equal-mass We write the spin-dependent contributions to the (cb) heavy quarks, the familiar LS coupling scheme, in which masses as 2S+1 states are labeled by n LJ , is apt. When the quark 4 masses are unequal, as in the case at hand, spin- X ∆ = Tk , (4) dependent terms in the Hamiltonian mix the spin-singlet k=1 and spin-triplet J = L states. We define 0 1 3 where the individual terms are |nLLi = cos θ|n LLi + sin θ|n LLi (7) hL~ · ~s i hL~ · ~s i i j |nL i = − sin θ|n1L i + cos θ|n3L i , T1 = 2 Te1(mi, mj) + 2 Te1(mj, mi) L L L 2mi 2mj where hL~ · ~sii hL~ · ~sji T2 = Te2(mi, mj) + Te2(mj, mi) (5) 2A mimj mimj tan θ = √ , (8) B + B2 + 4A2 h~si · ~sji T3 = Te3(mi, mj) mimj with   hSiji 1 p 1 1 T4 = Te4(mi, mj) , A = L(L + 1) − T (9) m m 4 2 2 e1 i j mc mb

~si and ~sj are the heavy-quark spins, S~ = ~si + ~sj and is the total spin, L~ is the orbital angular momentum   1 1 1 1 1 of quark and antiquark in the bound state, Sij = B = 4 2 + 2 Te1 + Te2 − 2 Te4. (10) mc mb mcmb mcmb 4 [3(~si · nˆ)(~sj · nˆ) − ~si · ~sj] is the tensor operator, andn ˆ is an arbitrary unit vector. Then our calculations of the Tek defined in Eq. (6) lead to We will deal with the hyperfine interaction T3 momen- ◦ these values for the mixing angle: θ2P = 18.7 , θ3D = tarily. We express the other T as ◦ ◦ ◦ ◦ ek −49.2 , θ3P = 21.2 , θ4F = −49.5 , θ4D = −40.3 .A 1 dV  Lattice calculation in quenched QCD [33] gave θ2P = ◦ Te1(mi, mj) = − + 2Te2(mi, mj) 33 ± 2 . r dR   The masses of the mixed states are 4c ˜2 αs(r) p Te2(mi, mj) = (6) 0 1 2 2 3 r3 M(nLL) = hM(nL)i − 2 (B − 4A + B ) (11)   1 p 2 2 c˜4 αs(r) M(nLL) = hM(nL)i − 2 (B + 4A + B ), Te4(mi, mj) = 3 , 3 r wherehM(nL)i is the nL centroid. where we have introduced the phenomenological coeffi- At lowest order, the hyperfine splitting between s-wave cientsc ˜2 andc ˜4, which take the value unity in the per- states, arising from T3, is given by turbative approach. 2 (n) 3 1 8αs|Rn0(0)| We extract values of Te2 and Te4 for the observed levels ∆HFS = M(n S1) − M(n S0) = , (12) that appear in Table I. These are shown as the underlined 9mcmb entries in Table II. Then, we combine the definitions in which is susceptible to significant quantum corrections. 3 Eq. (6) with our calculated values of hαs/r i to determine Rather than make a priori calculations of the hyperfine ¯ c˜2 andc ˜4 in the (cc¯) and (bb) families. The geometric splitting, we adopt the lattice QCD result for the ground mean of these values is our estimate for the coefficients state and scale the splittings of excited states according in the c¯b system. We insert these back into Eq. (6) to to estimate the values of Te2 and Te4 for the Bc family. For ∆(n) |R (0)|2 completeness, we include our evaluations of h(1/r)dV/dri HFS = n0 . (13) (1) 2 |R10(0)| in the Table. ∆HFS 5

TABLE III. Calculated excitation energies (in MeV) for (c¯b) 7600 levels with respect to the Bc(1S) centroid according to po- tential models and Lattice QCD simulations. The potential models have been aligned with the 1S doublet centroid at 7400 6315.5 MeV. Communication with states above flavor thresh- 7200 old is neglected.

Level EQ94 [6] Lattice QCD This Work 7000 1

1 S0 −54.8 −40.5 [14, 34, 35] −40.5 Mass [MeV] 3 6800 1 S1 18.2 13.5 [14, 34, 35] 13.5 3 2 P0 381 393(17)(7) [35] 377 6600 2P1 411 417(18)(7) [35] 415 0 2P1 417 446(30) [33] 423 3 2 P2 428 464(30) [33] 435 6400 1 2 S0 537 561(18)(1) [14] 551 3 2 S1 580 601(19)(1) [14] 582 6200 3 3 D1 693 - 691 3D2 693 - 690 0 3D2 686 - 700 3 3 D3 690 - 695 3 FIG. 3. Calculated c¯b spectrum, with (spin-singlet, spin- 3 P0 789 - 789 3P1 823 - 820 triplet) states shown on the (left [red], right [blue]) for each 0 3P1 816 - 828 orbital-angular-momentum family S, P, D, F. Dashed lines 3 3 P2 834 - 839 1 indicate thresholds for decay into two-body open-flavor chan- 3 S0 925 - 938 3 nels given in Table IV. 3 S1 961 - 964 3 4 F2 - - 918 4F3 - - 906 0 4F3 - - 922 3 4 F4 - - 908 TABLE IV. Open-flavor (c¯b) thresholds and excitations above 3 4 D1 - - 1031 1S centroid, 6315.5 MeV, for Bc levels, in MeV. 4D2 - - 1033 0 4D2 - - 1040 State Flavor threshold Excitation energy 3 4 D3 - - 1038 3 B+D0 7144.15 ± 0.15 829 4 P0 - - 1121 0 + 4P1 - - 1149 B D 7149.28 ± 0.16 834 0 ∗+ 0 4P1 - - 1158 B D 7189.48 ± 0.26 874 3 ∗0 + 4 P2 - - 1167 B D 7194.30 ± 0.26 879 1 4 S0 1243 - 1257 B+D∗0 7286.17 ± 0.15 971 3 4 S1 1276 - 1280 B0D∗+ 7289.89 ± 0.16 974 B∗+D∗0 7331.50 ± 0.26 1016 B∗0D∗+ 7334.91 ± 0.26 1019 0 + Bs Ds 7335.23 ± 0.20 1020 0 ∗+ III. THE Bc SPECTRUM Bs Ds 7479.09 ± 0.44 1164 ∗0 + +1.80 Bs Ds 7383.74−1.50 1068 ∗0 ∗+ +1.84 ∗ 3 Bs Ds 7527.60−1.55 1212 The vector meson Bc , the 1 S1 hyperfine partner of Bc and analogue of J/ψ and Υ, has not yet been observed. Modern lattice calculations [14, 34, 35] give consistent values for the hyperfine splitting M(B∗) − M(B ) = c c levels will lie below the beauty+charm flavor threshold, (53 ± 7, 54 ± 7, 55 ± 3 MeV), so we take the mass of in agreement with general arguments [36]. All of the po- the vector state to be M(B∗) = 6329 MeV and fix the c tential models cited in Ref. [7] predict 33S masses well centroid hM(1S)i of the ground-state s-wave doublet at 1 above the 829-MeV BD threshold. For the 31S level, 6315.5 MeV for the lattice. 0 only the Ebert et al. prediction does not lie significantly We summarize in Table III predictions for the spec- above B∗D threshold. Lattice QCD calculations do not trum of mesons with beauty and charm from our 1994 ar- yet exist for states beyond the 2S levels. ticle [6], lattice QCD calculations, and the present work, expressed as excitations with respect to the 1S centroid. Other potential-model calculations, some incorporating relativistic effects, may be found in the works cited in Ref. [7]. IV. DECAYS OF NARROW (c¯b) LEVELS Our expectations for the spectrum of states are shown in the Grotrian diagram, Figure 3, along with several A. Electromagnetic transitions of the lowest-lying open-flavor thresholds. The thresh- olds for strong decays of excited (c¯b) levels are known 3 ∗ experimentally to high accuracy, as shown in Table IV. The only significant decay mode for the 1 S1 (Bc ) state Comparing with the model calculations summarized in is the magnetic dipole (spin-flip) transition to the ground ¯ Table III, we conclude that two sets of narrow s-wave (cb) state, Bc. The M1 rate for transitions between s-wave 6 levels is given by k is the photon energy, and the statistical factor Sif = Sfi is as defined by Eichten and Gottfried [37]. Sif = 1 16α 2 3 2 for 3S →3 P transitions and S = 3 for allowed E1 ΓM1(i → f + γ) = µ k (2Jf + 1)|hf|j0(kr/2)|ii| , 1 J if 3 transitions between spin-singlet states. The statistical (14) factors for d-wave to p-wave transitions are reproduced where the magnetic dipole moment is in Table V. m e − m e The significant M1 and E1 electromagnetic transition µ = b c c b (15) 4m m rates and the ππ cascade rates are given in Table VI, c b along with the total widths in the absence of strong de- and k is the photon energy. cays. Apart from that M1 transition, only the electric dipole transitions are important for mapping the (c¯b) spectrum. The strength of the electric-dipole transitions is governed 3 3 TABLE V. Statistical factor Sif = Sfi for PJ → DJ0 + γ 3 3 by the size of the radiator and the charges of the con- and DJ → P J0 + γ transitions. stituent quarks. The E1 transition rate is given by 0 JJ Sif 4αhe i2 0 1 2 Γ (i → f + γ) = Q k3(2J + 1)|hf|r|ii|2S , E1 27 f if 1 1 1/2 (16) 1 2 9/10 where the mean charge is 2 1 1/50 2 2 9/50 mbec − mceb 2 3 18/25 heQi = , (17) mb + mc

B. Hadronic transitions C. Properties of (cb) wave functions at the origin

For quarks bound in a central potential, it is conve- We evaluate the rates for hadronic transitions between nient to separate the Schr¨odingerwave function into ra- (c¯b) levels according to the prescription we detailed in dial and angular pieces, as Ψ (~r) = R (r)Y (θ, φ), §IIIB of Ref. [6]. The results are included in Table VI. n`m n` `m where n is the principal quantum number, ` and m are Dipion cascades to the ground-state doublet are the dom- the orbital angular momentum and its projection, R (r) inant decay modes of 23S and 21S , and will be key to n` 1 0 is the radial wave function, and Y (θ, φ) is a spherical characterizing those states, as we shall discuss in §V A. `m harmonic [40]. The Schr¨odingerwave function is normal- R 3 2 R ∞ 2 As observed long ago by Brown and Cahn [38], an ized, d ~r |Ψn`m(~r)| = 1, so that 0 r dr|Rn`(r)| = 1. amplitude zero imposed by chiral symmetry pushes the The value of the radial wave function, or its first nonva- π+π− invariant mass distribution to higher invariant nishing derivative, at the origin, masses than phase-space alone would predict. In its sim- ` plest form, this analysis yields a universal form for the (`) d Rn`(r) Rn` (0) ≡ ` , (19) normalized dipion invariant mass distribution in quarko- dr r=0 nium cascades Φ0 → Φ π+π−, is required to evaluate pseudoscalar decay constants and production rates through heavy-quark fragmentation. ~ Our calculated values of |R(`)(0)|2 are given in Table VII. 1 dΓ |K| 2 2p 2 n` = Constant × 2 (2x − 1) x − 1 , (18) Γ dM MΦ0 The pseudoscalar decay constant fBc , which enters the calculations of annihilation decays such as c¯b → W + → + τ + ντ , is defined by where x = M/2mπ and K~ is the three-momentum car- ried by the pion pair. The soft-pion expression (18) h0|Aµ(0)|Bc(q)i = ifB Vcb qµ , (20) describes the depletion of the dipion spectrum at low c invariant masses observed in the transitions ψ(2S) → where Aµ is the axial-vector part of the charged weak ψ(1S)ππ, Υ(2S) → Υ(1S)ππ, and Υ(3S) → Υ(2S)ππ, current, Vcb is an element of the Cabibbo-Kobayashi- but fails to account for structures in the Υ(3S) → Maskawa quark-mixing matrix, and qµ is the four- Υ(1S)ππ spectrum [39]. We expect the 3S levels to lie momentum of the Bc. Its counterpart for the vector state above flavor threshold in the (c¯b) system, and so to have is very small branching fractions for cascade decays (but ∗ ∗ see the final paragraph of §V A. h0|V (0)|B (q)i = if ∗ V ε , (21) µ c Bc cb µ 7

TABLE VI. Total widths Γ and branching fractions B for principal decay modes of (c¯b) states below threshold, updating Table IX of Ref. [6]. Dissociation into BD, etc., will dominate over the tabulated decay modes for states above threshold.

Decay Mode kγ [keV] Branching Fraction (%) Decay Mode kγ [keV] Branching Fraction (%) 1 3 1 S0 (6275) : weak decays 3 P0 (7104) : Γ = 60.9 keV 3 13S (6329) : Γ = 0.144 keV 1 S1 (6329) 733 46.4 1 3 11S + γ 54 100 2 S1 (6897) 204 45.0 0 3 3 3 D1 (7006) 97 8.44 2 P0 (6692) : Γ = 53.1 keV 3 3P1 (7135) : Γ = 87.1 keV 1 S1 (6329) 354 100 13S (6329) 761 32.6 2P (6730) : Γ = 72.5 keV 1 1 11S (6275) 809 6.24 13S (6329) 389 86.2 0 1 23S (6897) 234 40.4 11S (6275) 440 13.7 1 0 21S (6866) 264 8.39 0 0 2P1 (6738) : Γ = 99.9 keV 3D2 (7005) 129 4.74 11S (6275) 448 92.4 0 0 3D2 (7015) 119 4.55 3 3 1 S1 (6329) 397 7.51 3 D1 (7006) 128 2.88 23P (6750) : Γ = 79.7 keV 0 2 3P1 (7143) : Γ = 113 keV 3 1 1 S1 (6329) 409 100 1 S0 (6275) 816 32.9 1 3 2 S0 (6866) : Γ = 73.1 keV 1 S1 (6329) 768 3.88 1 1 1 S0 + ππ 81.1 2 S0 (6866) 272 47.0 0 3 2P1 (6738) 126 16.5 2 S1 (6897) 242 5.15 0 2P1 (6730) 134 2.24 3D2 (7015) 127 4.22 3 3D2 (7005) 137 6.72 2 S1 (6897) : Γ = 76.8 keV 3 3 1 S1 + ππ 65.0 3 P2 (7154) : Γ = 100 keV 3 3 2 P0 (6692) 201 7.66 1 S1 (6329) 777 35.2 3 2P1 (6730) 165 11.5 2 S1 (6897) 252 49.2 0 2P1 (6738) 157 1.13 3D2 (7005) 147 1.09 3 0 2 P2 (6750) 145 14.6 3D2 (7015) 137 1.18 3 3 D1 (7006) 146 0.16 3D2 (7005) : Γ = 93.7 keV 3 1 3 D3 (7010) 142 13.0 1 S0 + ππ 12.2 3 1 S1 + ππ 9.0 4F3 (7221) : Γ = 77.6 keV 2P1 (6730) 270 29.1 3D2 (7005) 213 52.4 0 3D0 (7015) 203 42.8 2P1 (6738) 262 42.3 2 3 3 2 P2 (6750) 250 7.24 3 D3 (7010) 208 4.71 3 3 3 D1 (7006) : Γ = 117 keV 4 F4 (7223) : Γ = 79.9 keV 3 3 1 S1 + ππ 17.0 3 D3 (7010) 210 100 3 3 2 P0 (6692) 306 53.4 4 F2 (7233) : Γ = 95.3 keV 2P1 (6730) 270 25.3 3D2 (7005) 225 6.73 2P 0 (6738) 262 2.62 0 1 3D2 (7015) 215 7.96 3 3 2 P2 (6750) 251 1.51 3 D1 (7006) 224 84.8 3 3 3 D3 (7010) : Γ = 87.2 keV 3 D3 (7010) 220 0.42 13S + ππ 22.9 0 1 4F3 (7237) : Γ = 89.9 keV 3 0 2 P2 (6750) 255 77.0 3D2 (7015) 218 47.4 0 3D2 (7005) 228 52.5 3D2 (7015) : Γ = 92.1 keV 1 1 S0 + ππ 9.2 3 1 S1 + ππ 12.4 0 2P1 (6738) 272 37.2 2P1 (6730) 279 41.0

∗ ∗ and εµ is the polarization vector of the Bc . The ground- TABLE VII. Squares of radial wave functions at the origin state pseudoscalar and vector decay constants are given and related quantities (cf. Eq. (19)) for (c¯b) mesons. in terms of the wave function at the origin by the Van (`) 2 Level |Rn` (0)| Royen–Weisskopf formula [41], generically 1S 1.994 GeV3 2 2P 0.3083 GeV5 2 3|R10(0)| 2 f (∗) = C (αs), (22) 2S 1.144 GeV3 Bc πM 3D 0.0986 GeV7 3P 0.3939 GeV5 where the leading-order QCD correction is given by [42] 3S 0.9440 GeV3   4F 0.0493 GeV9 2 αs P,V mc − mb mc 7 C (αs) = 1 − δ − ln , (23) 4D 0.1989 GeV π mc + mb mb 4P 0.4540 GeV5 4S 0.8504 GeV3 and

δP = 2; δV = 8/3. (24) where Vµ is the vector part of the charged weak current Choosing the representative value αs = 0.38, and using 8 the quark masses given in Eq. (2), we find 40

0.904, P 35 C(α ) = (25) s 0.858, V 30 Consequently, we estimate the ground-state meson decay constants as 25

f = 498 MeV; f ∗ = 471 MeV, (26) 20 Bc Bc d σ /dy [nb/ Δ y ] so that f ∗ /f = 0.945. The compact size of the (c¯b) 15 Bc Bc system enhances the pseudoscalar decay constant relative to fπ and fK . 10 This is to be compared to a state-or-the-art lattice 5 evaluation [43], fBc = 434 ± 15 MeV, which entails im- proved NonRelativistic QCD for the valence b quark and 0 the Highly Improved Staggered Quark (HISQ) action for -4 -2 0 2 4 the lighter quarks on gluon field configurations that in- y clude the effect of u/d, s and c quarks in the sea with ∗ the u/d quark masses going down to physical values. FIG. 4. Rapidity√ distribution for production√ of Bc in pp colli- sions at s = 8 TeV (dotted blue curve), s = 13 TeV (solid The same calculation yields fB∗ /fB = 0.988 ± 0.027. √ c c black curve), and s = 14 TeV (dashed red curve), calcu- A calculation in the framework of QCD sum rules gives lated using BCVEGPY2.2 [45]. The bin width is ∆y = 0.5. fBc = 528 ± 19 MeV [44]. The mild asymmetries are statistical fluctuations.

V. PRODUCTION OF (c¯b) STATES AT THE 10 LARGE HADRON COLLIDER 5 We present in Table VIII cross sections for the produc- tion of B states at the Large Hadron Collider, calculated c ] using the framework of the BCVEGPY2.2 generator [45], 2 2 which we have extended to include the production of 3P (0) states. Cross sections for the physical (2, 3)P1 states are 1 appropriately weighted mixtures of the 3P and 1P cross

1 1 [nb/GeV 2 sections. T 0.5

TABLE VIII. Production rates (in nb) for (c¯b) states in pp d σ /d p collisions at the LHC. The production rates were calculated 0.2 using the BCVEGPY2.2 generator of Ref. [45], extended to include the production of 3P states. Color-octet contributions 0.1 to s-wave production are small; we show them (following |) 0 2 4 6 8 10 only for the 1S states. pT [GeV] √ √ √ (c¯b) level σ( s = 8 TeV) σ( s = 13 TeV) σ( s = 14 TeV) FIG. 5. Transverse momentum distribution of Bc produced in 1 √ √ 1 S0 46.8 | 1.01 80.3 | 1.75 88.0 | 1.90 pp collisions at s = 8 TeV√ (dotted blue curve), s = 13 TeV 3 1 S1 123.0 | 4.08 219.1 | 6.97 237.0 | 7.55 (solid black curve), and s = 14 TeV (dashed red curve), 3 2 P0 1.113 1.959 2.108 calculated using BCVEGPY2.2 [45] Small shape variations 3 2 P1 2.676 4.783 5.214 are statistical fluctuations. 1 2 P1 3.185 5.702 6.166 3 2 P2 6.570 11.57 12.64 1 2 S0 9.58 16.94 18.45 The rapidity distributions (for B∗ production, Fig- 3 c 2 S1 23.46 41.72 45.53 ure 4) and transverse-momentum distributions (shown 3 3 P0 0.915 1.642 1.806 for B production, Figure 5) are similar in character for 3 √ c 3 P1 2.263 4.082 4.478 1 s = 8, 13, and 14 TeV. The rapidity distributions for 3 P1 2.695 4.817 5.287 ¯ 3 low-lying (cb) states are shown in Figure 6. The accep- 3 P2 5.53 9.98 10.90 1 tance of the CMS and ATLAS detectors covers central 3 S0 4.23 7.53 8.08 33S 10.16 18.21 19.83 pseudorapidity |η| ≤ 2.5, whereas the geometrical accep- 1 tance of the LHCb detector is characterized by 2 ≤ η ≤ 5. 9

102 2.0

1.5

1 10 1.0 d σ /dM [nb/MeV] 0.5 d σ /d y [nb/ Δ ]

0 10 0 6820 6830 6840 6850 6860 6870 6880 6890 π+π– M(Bc ) [MeV]

FIG. 7. Calculated positions and relative strengths of the + − two-pion cascades Bc(2S) → Bc(1S)π π , represented as 10–1 Gaussian line shapes with standard deviation of 4 MeV. Pro- -4 -2 0 2 4 y duction rates are given in Table VIII and branching fractions in Table VI. We assume that the photon in the transition ∗ Bc → Bc + /γ is not included in the reconstruction. Rates FIG. 6. Rapidity distributions√ for the production of low-lying (c¯b) states in pp collisions at s = 13 TeV, calculated using confined to rapidity |y| ≤ 2.5 are 0.68× those shown. BCVEGPY2.2 [45]. From highest to lowest, the histograms 3 1 3 1 3 refer to production of the 1 S1, 1 S0, 2 S1, 2 S0, 2 P2, 2P1(0) , 3 √ 2 P0 levels. At s = 13 TeV, the ratio of cross sections is nearly 2.5. Taking account of the branching fractions, we estimate ∗ ∗ R ≈ 2. If Bc and Bc were produced with equal frequency, For comparison, approximately 68% of the Bc cross sec- we would find R ≈ 0.8. tion lies within |y| ≤ 2.5, and approximately 22% is pro- Now the CMS Collaboration [4] at the Large Hadron duced at forward rapidities y > 2. Similar fractions hold −1 √ for all the (c¯b) levels. Collider, analyzing 140 fb of pp collisions at s = 13 TeV, has observed a pattern that closely resem- bles the template of Figure 7. In the distribution of + − obs A. Dipion cascades M(Bcπ π ) − M(Bc) + M(Bc), they reconstruct a peak at 6871.0±1.2 MeV, which they identify as Bc(2S), and a second peak 29.0±1.5 MeV lower in mass (statisti- The path to establishing excited states will proceed by cal errors only). [The observed B mass is replaced, event resolving two separate peaks in the invariant mass distri- c 0 + − by event, with the world-average value to sharpen reso- butions associated with the cascades Bc → Bcπ π and ∗0 ∗ + − ∗ lution.] The putative Bc(2S) lies within 5 MeV of our B → B + π π , B → Bc + /γ (gamma unobserved). 1 c c c expectation for the 2 S0 level, and the separation is to The splitting between the peaks is set by the difference be compared with our expectation of 23 MeV. If we im- of mass differences, pose the scaling relation Eq. (13) for the hyperfine split- ∗0 0 ∗ tings, we reproduce the observed 29-MeV separation with ∆21 ≡ [M(Bc ) − M(Bc)] − [M(Bc ) − M(Bc)], (27) ∗ ∗0 0 M(Bc )−M(Bc) = 68 MeV, M(Bc )−M(Bc) = 39 MeV. ∗ generically expected to be negative [46]. The correspond- The Bc → Bc + γ photon momentum would be 68 MeV. ing quantity is approximately −64 MeV in the (cc¯) family An unbinned extended maximum-likelihood fit to the and −37 MeV in the (b¯b) family [12]. For the (c¯b) system, CMS data returns 66 ± 10 events for the lower peak and 51 ± 10 for the upper. These yields are not yet corrected a modern lattice simulation [14] gives ∆21 = −15 MeV, whereas the result of our potential-model calculation is for detection efficiencies and acceptances, so they cannot −23 MeV. In these circumstances, the undetected four- be used to infer ratios of production cross sections times momentum of the photon means that the reconstructed branching fractions. We look forward to the final result ∗ and to studies of the π+π− invariant mass distribution “Bc ” mass should correspond to the lower peak. We show an example of what is to be expected in Fig- as next steps in Bc spectroscopy. ure 7, taking the direct production cross sections (with Our calculations indicate that the 3S levels will lie no rapidity cuts) from Table VIII and the branching frac- above flavor threshold (see §V C, especially the discussion tions from Table VI. The (relative heights of, relative surrounding Figures 9 and 10), but it is conceivable that number of events in) the peaks measures the ratio coupled-channel effects might push one or both states lower in mass. For that reason, it is worth examining the ∗0 ∗0 ∗ + − + − σ(Bc + X) B(Bc → Bc + π π ) Bcπ π mass spectrum up through 7200 MeV for indi- R ≡ 0 0 + − . (28) 1 + − 3 ∗ + − σ(Bc + X) B(Bc → Bc + π π ) cations of 3 S0 → Bcπ π and 3 S1 → Bc π π lines. 10

3.5

3.0

2.5

2.0 σ B [nb/MeV] 1.5

1.0

0.5

0.0 100 150 200 250 300 350 400 450 500 1 ¯ k [MeV] FIG. 9. Strong decay widths of the 3 S0 (cb) level near open- flavor threshold. The shaded band on the mass axis indicates FIG. 8. Photon energies k and relative strengths of E1 tran- ±20 MeV around our nominal value for the mass of this state, sitions from 2S → 2P (left group, blue curves) and 2P → 1S 7253 MeV. (right group, black curves) (c¯b) states. Production rates are taken from Table VIII and branching fractions from Table VI. ∗ ∗ We suppose that the photon transition Bc → Bc + /γ goes transitions to Bc will be shifted downward because of unobserved in the cascade transitions. We assume Gaussian the unobserved M1 photon. It is not possible to produce lineshapes with standard deviation 2 MeV. enriched samples of the 2S levels by tuning the energy of e+e− collisions, as is done for J/ψ and Υ, so reconstruc- tion of the left-hand group of 2S → 2P transitions (blue According to our estimate of the 3S hyperfine splitting, lines in Figure 14) will be problematic. 3 1 the 3 S1 line would lie about 28 MeV below the 3 S0 In the far future, combining the photon transition en- line (36 MeV if we reset the 1S splitting to 68 MeV). ergies and relative rates with expectations for production For orientation, note that B(Υ(3S) → Υ(1S)π+π−) = and decay may eventually make it possible to disentangle 4.37±0.08%, while 36% of Υ(3S) decays proceed through mixing of the spin-singlet and spin-triplet J = L states. the ggg channel, which is not available to the (c¯b) states. According to Table VIII, the 3S states are produced at approximately 44% of the rate for their 2S counterparts. C. States above open-flavor threshold

We estimate the strong decay rates for (c¯b) states B. Electromagnetic transitions that lie above flavor threshold using the Cornell coupled- channel formalism [22] that we elaborated and applied to It may in time become possible for experiments to de- charmonium states in [23]. 1 3 tect some of the more energetic E1-transition photons We expect both the 3 S0 and 3 S1 states to lie above 1 that appear in Table VI. As an incentive for the search, threshold for strong decays. The 3 S0 state can decay ∗ 3 we show in Figure 14 the spectrum of E1 photons in de- into the final state B D and the 3 S1 level has decays 3 1 into both the BD and B∗D final states. The open decay cays of the 2 S1 and 2 S0 levels as well as the 2P → 2S ∗ channels as a function of the masses of these states is transitions, assuming as always a missing Bc → Bc/γ pho- ton in the reconstruction. Here we include direct produc- shown in Figures 9 and 10. 3 tion of the 2P states as well as feed-down from 2S → 2P The 3 P2 state might be observed as a very narrow ∗ (d-wave) BD line near open-flavor threshold. Its decay transitions. The strong Bc → Bc line arising from direct ∗ width as a function of mass for the 2P states are given production√ of Bc , for which we calculate σ · B ≈ 225 nb at s = 13 TeV, is probably too low in energy to be ob- in Figure 11. served. More promising are the 2P levels, which might In the phenomenological models the remaining 3P show themselves in Bc + γ invariant mass distributions. states lie just below the thresholds for strong decays. These lines make up the right-hand group (black lines) However they are near enough to these thresholds that 3 ∗ there might be interesting behavior at the threshold for in Figure 14. The 2 P2(6750) → Bc γ line is a particu- ∗ (0) larly attractive target for experiment, because of the fa- B D in the 3P1 cases and for the BD threshold in the 3 3 vorable production cross section, branching fraction, and case of the 3 P0 state. Figure 12 shows that the 3 P0 409-MeV photon energy. The 2P masses inferred from width grows rapidly just above threshold. The strong 11

3 3 FIG. 10. Strong decay widths of the 3 S1 (c¯b) level near open- FIG. 12. Estimated strong decay widths of the 3 P0 (c¯b) level flavor threshold. The shaded band on the mass axis indicates near open-flavor threshold. The shaded band on the mass axis ±20 MeV around our nominal value for the mass of this state, indicates ±20 MeV around our nominal value for the mass of 7279 MeV. this state, 7104 MeV.

3 (0) FIG. 11. Strong decay widths of the 3 P2 (c¯b) level near open- FIG. 13. Strong decay widths of the 3P1 or 3P1 . The shaded flavor threshold. The shaded band on the mass axis indicates band on the mass axis indicates ±20 MeV around our nominal ±20 MeV around our nominal value for the mass of this state, values for the masses of these state, 7135 and 7143 MeV. 7154 MeV.

√ run at c.m. energy s ≈ 240 GeV. As now envi- decay widths as a function of mass for the 3P and 3P 0 1 1 sioned, these machines would have the added capabil- states have a common behavior, displayed in Figure 13. √ ity of high-luminosity running at s = MZ that would It is worth keeping in mind that while narrow BD 12 + − 0 ¯ ¯ accumulate 10 examples of the reaction e e → Z . peaks may signal excited (cb) levels, narrow BD peaks With the observed branching fraction, B(Z0 → b¯b) = could indicate nearly bound bcq¯kq¯l tetraquark states [47]. (15.12 ± 0.05)% [12], the tera-Z mode would produce some 3 × 1011 boosted b-quarks, which would enable high-sensitivity searches for (c¯b) states in a variety of VI. TERA-Z PROSPECTS decay channels. A recent computation suggests that B(Z0 → (c¯b) + X) ≈ 6 × 10−4 [51]. In response to the discovery of the 125-GeV Higgs The largest existing e+e− → Z0 → hadrons data boson, H(125) [48], plans for large circular electron– sets were recorded by experiments at CERN’s Large positron colliders (FCC-ee [49] and CEPC [50]) are be- Electron–Positron collider (LEP) during the 1990s. In ing developed as e+e− → HZ0 “Higgs factories” to samples of (3.02, 3.9, and 4.2) million hadronic Z0 12

∗ decays, the DELPHI, ALEPH, and OPAL Collabora- from the Bc –Bc splitting. tions [52] found a small number of candidates for the + + decays Bc → J/ψπ , J/ψ` ν, and J/ψ3π. Those few spec- imens were not sufficient to establish a discovery, but the Appendix: Strong coupling evolution experiments were able to bound combinations of branch- ing fractions B as To make calculations with the frozen-αs potential, one must combine a linear term with a Coulomb term,  +  0 J/ψπ B(Z → B + X)   −4αs(r)/3r, for which αs(r) is characterized by the solid c B(B → J/ψ`+ν ) < few × 10−4, B(Z0 → hadrons) c ∼ red curve of Figure 1. We present in Table IX numerical  J/ψ3π  values of the strong coupling over the relevant range of (29) distances, 0 ≤ r ≤ 0.8 fm. The entries advance in steps at 90% confidence level, where X denotes anything. The of δ ln r = 0.1. relative simplicity of e+e− → Z0 events and the boosted kinematics of resulting Bc mesons suggest that a Tera-Z factory might be a felicitous choice to investigate 2P → TABLE IX. Evolution of the strong coupling. 1S + γ lines. r [fm] αs(r) 0.0080 0.1706 0.0088 0.1742 0.0097 0.1780 VII. CONCLUSIONS AND OUTLOOK 0.0108 0.1819 0.0119 0.1862 0.0132 0.1908 0.0145 0.1957 In this article, we have presented a new analysis of 0.0161 0.2007 the spectrum of mesons with beauty and charm. First, 0.0178 0.2061 0.0196 0.2116 we modified the traditional Coulomb-plus-linear form of 0.0217 0.2174 the quarkonium potential to incorporate running of the 0.0240 0.2235 strong coupling constant α that saturates at a fixed 0.0265 0.2299 s 0.0293 0.2365 value at long distances. The new frozen-αs potential 0.0323 0.2434 incorporates both perturbative and nonperturbative as- 0.0358 0.2505 0.0395 0.2579 pects of quantum chromodynamics. Second, we have set 0.0437 0.2659 aside the perturbative treatment of spin splittings, in- 0.0483 0.2743 0.0533 0.2829 stead incorporating lessons from Lattice QCD and ob- 0.0589 0.2915 servations of the (cc¯) and (b¯b) spectra. 0.0651 0.3001 0.0720 0.3087 We look forward to additional experimental progress, 0.0796 0.3171 first by confirming and elaborating the characteristics of 0.0879 0.3252 the 2S levels reported by the CMS Collaboration [4]. Key 0.0972 0.3330 0.1074 0.3403 1 observables are the mass of the 2 S0 state, the splitting 0.1187 0.3471 between the two lines, and the ratio of peak heights cor- 0.1312 0.3533 0.1450 0.3590 rected for efficiencies and acceptance. It is also of inter- 0.1602 0.3640 est to test whether the dipion mass spectra in the cas- 0.1771 0.3685 0 + − ∗0 ∗ + − 0.1957 0.3723 cade decays Bc → Bcπ π and Bc → Bc π π follow 0.2163 0.3757 the pattern seen in ψ(2S) → J/ψπ+π− and Υ(2S) → 0.2390 0.3786 + − 0.2642 0.3811 Υ(1S)π π decays. Although we expect the 3S levels to 0.2920 0.3832 + − lie above flavor threshold, exploring the Bcπ π mass 0.3227 0.3849 spectrum up through 7200 MeV might yield indications 0.3566 0.3864 1 + − 3 ∗ + − 0.3941 0.3876 of 3 S0 → Bcπ π and 3 S1 → Bc π π lines. The pres- 0.4355 0.3886 ence of one or the other of these could signal interactions 0.4813 0.3895 0.5320 0.3902 of bound states with open channels. Prospecting for nar- 0.5879 0.3908 row B(∗)D(∗) peaks near threshold could yield evidence 0.6497 0.3913 0.7181 0.3917 of Bc states beyond the 2S levels. 0.7936 0.3920 The next frontier is the search for radiative transitions among (c¯b) levels. The most promising candidate for 3 ∗ first light is the 2 P2(6750) → Bc γ transition. Determin- ∗ ∗ ing the Bc mass, perhaps by reconstructing Bc → Bcγ, would provide an important check on lattice QCD calcu- ADDENDUM lations and a key input to future calculations. + Detecting the Bc → τντ and Bc → ppπ¯ decays would In the discussion surrounding Figure 8 of the pub- be impressive experimental feats, and would provide an- lished version of this Article [53], we highlighted the other test of the short-distance behavior of the ground- possibility that E1 electric-dipole transitions from the state wave function, complementing what will be learned 2P → 1S levels might offer an imminent opportunity 13 to establish orbitally excited levels We pointed to the U.S. Department of Energy, Office of Science, Office of 3 ∗ 2 P2(6750) → Bc γ line as an especially promising tar- High Energy Physics. get for experiment, because of the favorable production cross section and 409-MeV photon energy. We did not specifically comment of prospects for establishing the 3P 1.0 states. This Addendum repairs that omission. We show in a new Figure 14 cross sections × branching fractions for the spectrum of E1 photons in decays of the 0.8 3PJ to 1S levels. (Since the 3S levels should lie above flavor threshold, we neglect feed-down from 3S → 3P (0) 0.6 transitions. Cross sections for the physical 3P1 states 3 1 are appropriately weighted mixtures of the 3 P1 and 3 P1 cross sections.) Although the yields are approximately σ B [nb/MeV] four times smaller than those for the 2P → 1S lines, the 0.4 higher photon energies may be a decisive advantage for 3 ∗ detection. The 3 P2(7154) → Bc γ(777 MeV) line is a particularly attractive target for experiment. 0.2 Experiments at the Large Hadron Collider have demonstrated the feasibility of E1 spectroscopy in the ¯ 00 0 (bb) family, discovering and characterizing χb1 and 720 740 760 780 800 820 840 00 ¯ k [MeV] χb2 [54]. Observation of some (cb) P -wave states should be possible with the data sets now in hand. FIG. 14. Photon energies k and predicted yields of E1 tran- sitions from 3P → 1S (c¯b) states. Photon momenta and E1 branching fractions are taken from from Table VI; production ACKNOWLEDGMENTS rates are taken from Table VIII. The 3P masses inferred from ∗ transitions to Bc will be shifted downward because of the ∗ This work was supported by Fermi Research Alliance, missing Bc → Bc/γ photon in the reconstruction. We model LLC under Contract No. DE-AC02-07CH11359 with the Gaussian lineshapes with standard deviation 2 MeV.

[1] Georges Aad et al. (ATLAS Collaboration), “Obser- arXiv:1902.00571 [hep-ex]. ± vation of an Excited Bc Meson State with the AT- [5] E. Eichten and F. Feinberg, “Spin Dependent Forces in LAS Detector,” Phys. Rev. Lett. 113, 212004 (2014), QCD,” Phys. Rev. D23, 2724 (1981). arXiv:1407.1032 [hep-ex]. [6] Estia J. Eichten and Chris Quigg, “Mesons with beauty [2] Roel Aaij et al. (LHCb Collaboration), “Search for ex- and charm: Spectroscopy,” Phys. Rev. D49, 5845–5856 + cited Bc states,” JHEP 01, 138 (2018), arXiv:1712.04094 (1994), arXiv:hep-ph/9402210 [hep-ph]. [hep-ex]. [7] For other work in a similar spirit, see V. V. Kise- [3] For a recent assessment, see contributions to the lev, A. K. Likhoded, and A. V. Tkabladze, “Bc spec- + Micro Workshop on Bc physics at LHCb, 13 July troscopy,” Phys. Rev. D51, 3613–3627 (1995), arXiv:hep- 2016, https://indico.cern.ch/event/549155/, es- ph/9406339 [hep-ph]; Lewis P. Fulcher, “Phenomeno- pecially Zhenwei Yang, “Experiment – Recent his- logical predictions of the properties of the Bc system,” + tory of the Bc meson,” https://indico.cern.ch/ Phys. Rev. D60, 074006 (1999), arXiv:hep-ph/9806444 event/549155/contributions/2226387/attachments/ [hep-ph]; S. Godfrey and Nathan Isgur, “Mesons in a 1308369/1957602/Bc_results20160713.pdf; Alexey Relativized Quark Model with Chromodynamics,” Phys. Luchinsky, “Theory – Status of the theoretical de- Rev. D32, 189–231 (1985); Stephen Godfrey, “Spec- + scription of the Bc meson,” https://indico. troscopy of Bc mesons in the relativized quark model,” cern.ch/event/549155/contributions/2226440/ Phys. Rev. D70, 054017 (2004), arXiv:hep-ph/0406228 attachments/1308367/1956579/bc_Luchinsky.pdf; [hep-ph]; D. Ebert, R. N. Faustov, and V. O. Galkin, Rolf Oldeman, “Experiment – Perspectives for fu- “Properties of heavy quarkonia and Bc mesons in the rel- ture studies,” https://indico.cern.ch/event/ ativistic quark model,” Phys. Rev. D67, 014027 (2003), 549155/contributions/2226443/attachments/ arXiv:hep-ph/0210381 [hep-ph]; A. V. Berezhnoy, V. V. 1308858/1957460/20160713_Bc.pdf; Aleksandr V. Kiselev, A. K. Likhoded, and A. I. Onishchenko, “Bc me- + Berezhnoy, “Theory – Bc production and spec- son at LHC,” Phys. Atom. Nucl. 60, 1729–1740 (1997), troscopy,” https://indico.cern.ch/event/549155/ [Yad. Fiz. 60N10, 1889 (1997)], arXiv:hep-ph/9703341 contributions/2226438/attachments/1309000/ [hep-ph]; N. R. Soni, B. R. Joshi, R. P. Shah, H. R. 1957694/Berezhnoy_Bc.pdf. Chauhan, and J. N. Pandya, “QQ¯ ( Q ∈ {b, c} ) spec- [4] Albert M. Sirunyan et al. (CMS Collaboration), “Obser- troscopy using the Cornell potential,” Eur. Phys. J. C78, + vation of two excited Bc states and measurement of the 592 (2018), arXiv:1707.07144 [hep-ph]. + √ Bc (2S) mass in pp collisions at s = 13 TeV,” (2019), [8] F. Abe et al. (CDF Collaboration), “Observation of the 14

√ Bc meson in pp¯ collisions at s = 1.8 TeV,” Phys. Rev. Jonna Koponen, “Bc spectroscopy using highly improved Lett. 81, 2432–2437 (1998), arXiv:hep-ex/9805034 [hep- staggered quarks,” in 36th International Symposium on ex]. Lattice Field Theory (Lattice 2018) East Lansing, MI, [9] T. Aaltonen et al. (CDF Collaboration), “Observation United States, July 22-28, 2018 (2018) arXiv:1811.09448 + ± of the Decay Bc → J/ψπ and Measurement of [hep-lat]. + the Bc Mass,” Phys. Rev. Lett. 100, 182002 (2008), [20] Thomas Appelquist and H. David Politzer, “Heavy arXiv:0712.1506 [hep-ex]. Quarks and e+e− Annihilation,” Phys. Rev. Lett. 34, [10] V. M. Abazov et al. (D0 Collaboration), “Observation of 43 (1975). the Bc Meson in the Exclusive Decay Bc → J/ψπ,” Phys. [21] Thomas Appelquist, A. De Rujula, H. David Politzer, Rev. Lett. 101, 012001 (2008), arXiv:0802.4258 [hep-ex]. and S. L. Glashow, “Spectroscopy of the New Mesons,” [11] R. Aaij et al. (LHCb Collaboration), “Measurements of Phys. Rev. Lett. 34, 365 (1975); E. Eichten, K. Got- + + + Bc production and mass with the Bc → J/ψπ decay,” tfried, T. Kinoshita, John B. Kogut, K. D. Lane, Phys. Rev. Lett. 109, 232001 (2012), arXiv:1209.5634 and Tung-Mow Yan, “Spectrum of Charmed Quark- + + [hep-ex]; “Observation of Bc → J/ψDs and Antiquark Bound States,” Phys. Rev. Lett. 34, 369–372 + ∗+ Bc → J/ψDs decays,” Phys. Rev. D87, 112012 (1975), [Erratum: Phys. Rev. Lett. 36, 1276 (1976)]. (2013), [Addendum: Phys. Rev. D89, 019901(2014)], [22] E. Eichten, K. Gottfried, T. Kinoshita, K. D. Lane, arXiv:1304.4530 [hep-ex]; “First observation of a bary- and Tung-Mow Yan, “Charmonium: The Model,” Phys. + onic Bc decay,” Phys. Rev. Lett. 113, 152003 (2014), Rev. D17, 3090 (1978), [Erratum: Phys. Rev. D21, 313 arXiv:1408.0971 [hep-ex]. (1980)]; “Charmonium: Comparison with Experiment,” [12] M. Tanabashi et al. (Particle Data Group), “Review of Phys. Rev. D21, 203 (1980). Particle Physics,” Phys. Rev. D98, 030001 (2018). [23] Estia J. Eichten, Kenneth Lane, and Chris Quigg, [13] Ian F. Allison, Christine T. H. Davies, Alan Gray, An- “Charmonium levels near threshold and the narrow dreas S. Kronfeld, Paul B. Mackenzie, and James N. state X(3872) → π+π−J/ψ,” Phys. Rev. D69, 094019 Simone (Fermilab Lattice, HPQCD, UKQCD Collabo- (2004), arXiv:hep-ph/0401210 [hep-ph]; “New states rations), “Mass of the Bc meson in three-flavor lattice above charm threshold,” Phys. Rev. D73, 014014 (2006), QCD,” Phys. Rev. Lett. 94, 172001 (2005), arXiv:hep- [Erratum: Phys. Rev. D73, 079903(2006)], arXiv:hep- lat/0411027 [hep-lat]. ph/0511179 [hep-ph]. [14] R. J. Dowdall, C. T. H. Davies, T. C. Hammant, and [24] N. Brambilla et al., “Heavy quarkonium: progress, puz- R. R. Horgan, “Precise heavy-light meson masses and zles, and opportunities,” Eur. Phys. J. C71, 1534 (2011), hyperfine splittings from lattice QCD including charm arXiv:1010.5827 [hep-ph]. quarks in the sea,” Phys. Rev. D86, 094510 (2012), [25] Yukinari Sumino, “Computation of Heavy Quarkonium arXiv:1207.5149 [hep-lat]. Spectrum in Perturbative QCD,” in 13th DESY Work- 2S+1 [15] We use spectroscopic notation n LJ , where n is the shop on Elementary Particle Physics: Loops and Legs principal quantum number, S is total spin, and L = in Quantum Field Theory (LL2016) Leipzig, Germany, S, P, D, . . . represents the angular momentum 0, 1, 2,.... April 24-29, 2016 (2016) arXiv:1607.03469 [hep-ph]. [16] Martin Beneke and Gerhard Buchalla, “The Bc Meson [26] Andr´eMartin, “A Fit of Upsilon and Charmonium Spec- Lifetime,” Phys. Rev. D53, 4991–5000 (1996), arXiv:hep- tra,” Phys. Lett. 93B, 338–342 (1980). ph/9601249 [hep-ph]; A. Yu. Anisimov, I. M. Narodet- [27] John L. Richardson, “The Heavy Quark Potential and sky, C. Semay, and B. Silvestre-Brac, “The Bc meson the Υ, J/ψ Systems,” Phys. Lett. 82B, 272–274 (1979). lifetime in the light front constituent quark model,” Phys. [28] W. Buchm¨ullerand S. H. H. Tye, “Quarkonia and Quan- Lett. B452, 129–136 (1999), arXiv:hep-ph/9812514 [hep- tum Chromodynamics,” Phys. Rev. D24, 132 (1981). ph]; V. V. Kiselev, A. E. Kovalsky, and A. K. Likhoded, [29] V. N. Gribov, “The theory of quark confinement,” Eur. “Bc decays and lifetime in QCD sum rules,” Nucl. Phys. Phys. J. C10, 91–105 (1999), arXiv:hep-ph/9902279 B585, 353–382 (2000), arXiv:hep-ph/0002127 [hep-ph]; [hep-ph]. Chao-Hsi Chang, Shao-Long Chen, Tai-Fu Feng, and [30] K. G. Chetyrkin, “Four-loop renormalization of QCD: Xue-Qian Li, “The Lifetime of Bc meson and some Full set of renormalization constants and anoma- relevant problems,” Phys. Rev. D64, 014003 (2001), lous dimensions,” Nucl. Phys. B710, 499–510 (2005), arXiv:hep-ph/0007162 [hep-ph]. arXiv:hep-ph/0405193 [hep-ph]; M. Czakon, “The [17] N. Brambilla et al. (Quarkonium Working Group), Four-loop QCD beta-function and anomalous dimen- “Heavy quarkonium physics,” (2004), arXiv:hep- sions,” Nucl. Phys. B710, 485–498 (2005), arXiv:hep- ph/0412158 [hep-ph]. ph/0411261 [hep-ph]. ± [18] A. Ali, L. Maiani, A. D. Polosa, and V. Riquer, “Bc de- [31] M. Je˙zabek, M. Peter, and Y. Sumino, “On the rela- cays into tetraquarks,” (2016), arXiv:1604.01731 [hep- tion between QCD potentials in momentum and position ph]; A. Esposito, M. Papinutto, A. Pilloni, A. D. space,” Phys. Lett. B428, 352–358 (1998), arXiv:hep- Polosa, and N. Tantalo, “Doubly charmed tetraquarks ph/9803337 [hep-ph]. in Bc and Ξbc decays,” Phys. Rev. D88, 054029 (2013), [32] Dieter Gromes, “Spin Dependent Potentials in QCD and arXiv:1307.2873 [hep-ph]; Wei Wang, Yue-Long Shen, the Correct Long Range Spin Orbit Term,” Z. Phys. C26, − − − and Cai-Dian Lu, “The Study of Bc → X(3872)π (K ) 401 (1984). decays in the covariant light-front approach,” Eur. Phys. [33] C. T. H. Davies, K. Hornbostel, G. P. Lepage, A. J. Lid- J. C51, 841–847 (2007), arXiv:0704.2493 [hep-ph]; Wei sey, J. Shigemitsu, and J. H. Sloan, “Bc spectroscopy Wang and Qiang Zhao, “Decipher the short-distance from lattice QCD,” Phys. Lett. B382, 131–137 (1996), component of X(3872) in Bc decays,” Phys. Lett. B755, arXiv:hep-lat/9602020 [hep-lat]. 261–264 (2016), arXiv:1512.03123 [hep-ph]. [34] E. B. Gregory, C. T. H. Davies, E. Follana, E. Gamiz, [19] Andrew Lytle, Brian Colquhoun, Christine Davies, and I. D. Kendall, G. P. Lepage, H. Na, J. Shigemitsu, 15

∗ and K. Y. Wong, “A Prediction of the Bc mass in gram vary with the produced state, to reproduce its mass. full lattice QCD,” Phys. Rev. Lett. 104, 022001 (2010), 1S: mb = 5.000, mc = 1.275; 2S: mb = 5.234, mc = arXiv:0909.4462 [hep-lat]. 1.633; 2P : mb = 5.184, mc = 1.573; 3S: mb = [35] Nilmani Mathur, M. Padmanath, and Sourav Mondal, 5.447, mc = 1.825; 3P : mb = 5.502, mc = 1.633, all in “Precise predictions of charmed-bottom hadrons from GeV. ν lattice QCD,” Phys. Rev. Lett. 121, 202002 (2018), [46] In an effective power-law potential V (r) = λr , ∆21 < 0 arXiv:1806.04151 [hep-lat]. so long as ν < 1. See §4.1.1 and §5.3.2 of C. Quigg and [36] C. Quigg and Jonathan L. Rosner, “Counting Nar- Jonathan L. Rosner, “Quantum Mechanics with Appli- row Levels of Quarkonium,” Phys. Lett. B72, 462–464 cations to Quarkonium,” Phys. Rept. 56, 167–235 (1979) (1978). particularly Eqns. (4.21, 4.22). [37] E. Eichten and K. Gottfried, “Heavy Quarks in e+e− [47] Estia J. Eichten and Chris Quigg, “Heavy-quark symme- Annihilation,” Phys. Lett. 66B, 286 (1977). try implies stable heavy tetraquark mesons QiQj q¯kq¯l,” [38] Lowell S. Brown and Robert N. Cahn, “Chiral Symmetry Phys. Rev. Lett. 119, 202002 (2017), arXiv:1707.09575 and ψ0 → ψππ Decay,” Phys. Rev. Lett. 35, 1 (1975). [hep-ph] and references cited therein. [39] See §7 of Ref. [17] and §3.3 of Ref. [24] for surveys of cas- [48] Georges Aad et al. (ATLAS Collaboration), “Observation cade decays. of a new particle in the search for the Standard Model [40] We adopt the conventional normalization, Higgs boson with the ATLAS detector at the LHC,” R ∗ dΩ Y`m(θ, φ)Y`0m0 (θ, φ) = δ``0 δmm0 . See, e.g., the Phys. Lett. B716, 1–29 (2012), arXiv:1207.7214 [hep-ex]; Appendix of Hans A. Bethe and Edwin E. Salpeter, Serguei Chatrchyan et al. (CMS Collaboration), “Obser- Quantum Mechanics of One- and Two-Electron Atoms vation of a new boson at a mass of 125 GeV with the (Springer-Verlag, Berlin, 1957). CMS experiment at the LHC,” Phys. Lett. B716, 30–61 [41] R. Van Royen and V. F. Weisskopf, “Hadron Decay Pro- (2012), arXiv:1207.7235 [hep-ex]. cesses and the Quark Model,” Nuovo Cim. A50, 617–645 [49] “FCC-ee, the electron–positron option of the Future Cir- (1967), [Erratum: Nuovo Cim. A51, 583 (1967)] The fac- cular Colliders design study,” http://tlep.web.cern. tor of 3 accounts for quark color. ch. [42] Eric Braaten and Sean Fleming, “QCD radiative cor- [50] “CEPC, Circular Electron–Positron Collider,” http:// rections to the leptonic decay rate of the Bc meson,” cepc.ihep.ac.cn. Phys. Rev. D52, 181–185 (1995), arXiv:hep-ph/9501296 [51] Qi-Li Liao, Yan Yu, Ya Deng, Guo-Ya Xie, and Guang- [hep-ph]; A. V. Berezhnoy, V. V. Kiselev, and A. K. Chuan Wang, “Excited heavy quarkonium production via 0 Likhoded, “Photonic production of S- and P wave Bc Z decays at a high luminosity collider,” Phys. Rev. D91, states and doubly heavy baryons,” Z. Phys. A356, 89– 114030 (2015), arXiv:1505.03275 [hep-ph]. 97 (1996). [52] P. Abreu et al. (DELPHI Collaboration), “Search for the [43] B. Colquhoun, C. T. H. Davies, R. J. Dowdall, J. Ket- Bc Meson,” Phys. Lett. B398, 207–222 (1997); R. Barate tle, J. Koponen, G. P. Lepage, and A. T. Lytle et al. (ALEPH Collaboration), “Search for the Bc me- (HPQCD), “B-meson decay constants: a more com- son in hadronic Z decays,” Phys. Lett. B402, 213– plete picture from full lattice QCD,” Phys. Rev. D91, 226 (1997); K. Ackerstaff et al. (OPAL Collaboration), 0 114509 (2015), arXiv:1503.05762 [hep-lat]. For further “Search for the Bc meson in hadronic Z decays,” Phys. work on semileptonic decays, see Andrew Lytle, Brian Lett. B420, 157–168 (1998), arXiv:hep-ex/9801026 [hep- Colquhoun, Christine Davies, Jonna Koponen, and ex]. Craig McNeile, “Semileptonic Bc decays from full lattice [53] Estia J. Eichten and Chris Quigg, “Mesons with Beauty QCD,” in 16th International Conference on B-Physics and Charm: New Horizons in Spectroscopy,” Phys. Rev. at Frontier Machines (Beauty 2016) Marseille, France, D 99, 054025 (2019), arXiv:1902.09735 [hep-ph]. May 2-6, 2016 (2016) arXiv:1605.05645 [hep-lat]; An- [54] Georges Aad et al. (ATLAS Collaboration), “Observa- + drew Lytle, “Theory – Semileptonic Bc decays and tion of a new χb state in radiative transitions to Υ(1S) Lattice QCD,” In Ref. [3], https://indico.cern.ch/ and Υ(2S) at ATLAS,” Phys. Rev. Lett. 108, 152001 event/549155/contributions/2226442/attachments/ (2012), arXiv:1112.5154 [hep-ex]; Roel Aaij et al. (LHCb 1308725/1957203/Bc_lattice.pdf. Collaboration),√ “Study of χb meson production in p p [44] Michael J. Baker, Jose Bordes, Cesareo A. Dominguez, collisions at s = 7 and 8 TeV and observation of the de- Jose Penarrocha, and Karl Schilcher, “B Meson Decay cay χb(3P) → Υ(3S)γ,” Eur. Phys. J. C74, 3092 (2014),

Constants fBc , fBs and fB from QCD Sum Rules,” JHEP arXiv:1407.7734 [hep-ex]; “Measurement of the χb(3P ) 07, 032 (2014), arXiv:1310.0941 [hep-ph]. mass and of the relative rate of χb1(1P ) and χb2(1P ) pro- [45] Chao-Hsi Chang, Jian-Xiong Wang, and Xing-Gang Wu, duction,” JHEP 10, 088 (2014), arXiv:1409.1408 [hep- “BCVEGPY2.0: An upgraded version of the genera- ex]; A. M. Sirunyan et al. (CMS Collaboration), “Ob- tor BCVEGPY with the addition of hadroproduction of servation of the χb1(3P) and χb2(3P) and measurement the P -wave Bc states,” Comput. Phys. Commun. 174, of their masses,” Phys. Rev. Lett. 121, 092002 (2018), 241–251 (2006), arXiv:hep-ph/0504017 [hep-ph]. We use arXiv:1805.11192 [hep-ex]. Note that these articles label (derivatives of) wave functions at the origin derived from states by the radial quantum number, n − L. our current work. The quark mass parameters in this pro-