BB s B s B s Decays at Belle
Remi Louvot ∗ (On behalf of the Belle collaboration)Laboratoire de Physique des Hautes Énergies,École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, SwitzerlandE-mail: [email protected] The large data sample recorded with the Belle detector at the ϒ ( S ) energy provides a uniqueopportunity to study the poorly-known B s meson. Several analyses, made with a data samplerepresenting an integrated luminosity of 23.6 fb − , are presented. We report the study of thelarge-signal B s → D ( ∗ ) − s h + ( h + = π + , ρ + ) decays including the first observations of B s → D ∗− s π + and B s → D ( ∗ ) − s ρ + . In addition, several results on CP -eigenstate B s decays are described. Theseinclude the study of the B s → J / ψ η ( (cid:48) ) and B s → J / ψ f ( ) decays, the charmless B s → K + K − , B s → π + π − and B s → K S K S decays and the simultaneous fit of the three B s → D ( ∗ )+ s D ( ∗ ) − s modesfrom which ∆Γ CP s / Γ s is extracted. The preliminary measurement of B ( B s → J / ψ f ( )) < . × − (at 90% C.L.) is presented for the first time.1 November 2010LPHE Note 2010-06 Flavor Physics and CP Violation - FPCP 2010May 25-29, 2010Turin, Italy ∗ Speaker. © Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike Licence. http://pos.sissa.it/ a r X i v : . [ h e p - e x ] J a n s Decays at Belle
Remi Louvot
Introduction
The Belle experiment [1], located at the interaction point of the KEKB asymmetric-energy e + e − collider [2], was designed for the study of B mesons produced in e + e − annihilation at acenter-of-mass (CM) energy corresponding to the mass of the ϒ ( S ) resonance ( √ s ≈ .
58 GeV).After having recorded an unprecedented sample of ∼
700 millions of B ¯ B pairs, the Belle collabora-tion started to record collisions at higher energies, opening the possibility to study other particles,like the B s meson. Up to now, a data sample of integrated luminosity of L int = ( . ± . ) fb − (outof a total of 120 fb − ) has been analyzed at the energy of the ϒ ( S ) resonance ( √ s ≈ .
87 GeV).Since the ϒ ( S ) resonance is just above the B s ¯ B s threshold, it was naturally expected that the B s meson could be studied with ϒ ( S ) data as well as the B mesons are with ϒ ( S ) data. Thelarge potential of such ϒ ( S ) data was quickly confirmed [3, 4] with the 2005 engineering runrepresenting 1.86 fb − . The main advantage with respect to the hadronic colliders is the possibilityof measurements of absolute branching fractions. However, the abundance of B s mesons in ϒ ( S ) hadronic events has to be precisely determined. Above the e + e − → u ¯ u , d ¯ d , s ¯ s , c ¯ c continuum events,the e + e − → b ¯ b process can produce different kinds of final states involving a pair of non-strange B mesons [5] ( B ∗ ¯ B ∗ , B ∗ ¯ B , B ¯ B , B ∗ ¯ B ∗ π , B ∗ ¯ B π , B ¯ B π , B ¯ B ππ and B ¯ B γ ), a pair of B s mesons ( B ∗ s ¯ B ∗ s , B ∗ s ¯ B s and B s ¯ B s ), or final states involving a lighter bottomonium resonance below the open-beautythreshold [6]. The B ∗ and B ∗ s mesons always decay by emission of a photon. The total e + e − → b ¯ b cross section at the ϒ ( S ) energy was measured to be σ b ¯ b = ( ± ) pb [3, 7] and the fraction of B s events to be f s = σ ( e + e − → B ( ∗ ) s ¯ B ( ∗ ) s ) / σ b ¯ b = ( . ± . ) % [8]. The dominant B s productionmode, b ¯ b → B ∗ s ¯ B ∗ s , represents f B ∗ s ¯ B ∗ s = (cid:0) . + . − . ± . (cid:1) % of the b ¯ b → B ( ∗ ) s ¯ B ( ∗ ) s events, as measuredwith B s → D − s π + events (next Section).For all the exclusive modes presented here, the B s candidates are fully reconstructed fromthe final-state particles. From the reconstructed four-momentum in the CM, ( E ∗ B s , ppp ∗ B s ) , two vari-ables are formed: the energy difference ∆ E = E ∗ B s − √ s / M bc = (cid:113) s / − ppp ∗ B s . The signal coming from the dominant e + e − → B ∗ s ¯ B ∗ s production mode is extractedfrom a two-dimensional fit performed on the distribution of these two variables. The correspond-ing branching fraction is then extracted using the total efficiency (including sub-decay branchingfractions) determined with Monte-Carlo (MC) simulations, ∑ ε B , and the number of B s mesonsproduced via the e + e − → B ∗ s ¯ B ∗ s process, N B s = × L int × σ b ¯ b × f s × f B ∗ s ¯ B ∗ s = ( . ± . ) × .
1. Dominant CKM-favored B s B s B s Decays
We report the measurement of exclusive B s → D ( ∗ ) − s h + ( h + = π + or ρ + ) decays [9, 10] whichis an important milestone in the study of the poorly-known decay processes of the B s meson.These modes are expected to produce an abundant signal because of their relatively large predictedbranching fractions [11, 12] and their clean signatures: four charged tracks and up to two pho-tons. The leading amplitude for the four B s → D ( ∗ ) − s π + and B s → D ( ∗ ) − s ρ + modes is a b → c The notation “ B ” refers either to a B or a B + . Moreover, charge-conjugated states are implied everywhere. The branching-fraction values for B s → D − s π + and those in Sections 2 and 3 are calculated with f s = ( . + . − . ) %,also provided in Ref. [8]. s Decays at Belle
Remi Louvot tree diagram of order λ (in the Wolfenstein parametrization [13] of the CKM quark-mixing ma-trix [14, 15]) with a spectator s quark. Besides being interesting in their own right, such measure-ments, if precise enough, can be of high importance for the current and forthcoming hadron colliderexperiments. It was for example recently pointed out [16] that the search for the very rare decay B s → µ + µ − , which has a branching fraction very sensitive to New Physics contributions, will besystematically limited at LHCb by the poor knowledge of B s production, in case New Physics willenhance the decay probability by no more than a factor 3 above the Standard Model expectation.In addition, polarization measurements of B decays have become of high interest since the ob-servation of a surprisingly large transverse polarization in B → φ K ∗ decays by Belle and BaBar [17,18]. The relative strengths of the longitudinal and transverse states can be measured with an an-gular analysis of the decay products. In the helicity basis, the expected B s → D ∗− s ρ + differentialdecay width is proportional tod Γ ( B s → D ∗− s ρ + ) d cos θ D ∗− s d cos θ ρ + ∝ f L sin θ D ∗− s cos θ ρ + + ( − f L )( + cos θ D ∗− s ) sin θ ρ + , where f L = | H | / ∑ λ | H λ | is the longitudinal polarization fraction, H λ ( λ = ± ,
0) are the helicity
Figure 1:
Left: M bc distributions for the B s → D − s π + (top) B s → D ∗− s π + (middle) and B s → D − s ρ + (bot-tom) candidates with ∆ E restricted to the B ∗ s ¯ B ∗ s signal region. Right: ∆ E distributions with M bc restrictedto the B ∗ s ¯ B ∗ s signal region. The black- (green-) dotted line represents the continuum (peaking) background,while the red-dashed curves are the signal shapes. The larger one is the signal in the B ∗ s ¯ B ∗ s kinematic regionand the two others, which are very close to 0, are the signals in the two other B s production modes ( B ∗ s ¯ B s and B s ¯ B s ). s Decays at Belle
Remi Louvot -0.8 -0.4 0 0.4 0.8 -0.8 -0.4 0 0.4 0.8
Figure 2:
Fit of the B s → D ∗− s ρ + candidates. Top: M bc and ∆ E distributions, similarly to Fig. 1. Bottom:helicity distributions of the D ∗− s (left) and ρ + (right) with M bc and ∆ E restricted to the B ∗ s ¯ B ∗ s kinematicregion. The black-dotted line represents the background, while the two red-dashed curves are the signal.The large (small) signal shape corresponds to the longitudinal (transverse) component. amplitudes, and θ D ∗− s ( θ ρ + ) is the helicity angle of the D ∗− s ( ρ + ) defined as the supplement of theangle between the B s and the D − s ( π + ) momenta in the D ∗− s ( ρ + ) frame.The D − s mesons are reconstructed via three modes : D − s → φ ( → K + K − ) π − , D − s → K ∗ ( → K + π − ) K − and D − s → K S ( → π + π − ) K − . Based on the ratio of the second and the zeroth Fox-Wolfram moments [19], R , the continuum events are efficiently rejected by taking advantage ofthe difference between their event geometry (jet like, high R ) and the signal event shape (spherical,low R ). The B s → D − s π + and B s → D ∗− s π + ( B s → D − s ρ + and B s → D ∗− s ρ + ) candidates with R smaller than 0.5 (0.35) are kept for further analysis. A best candidate selection, based onthe intermediate-particle reconstructed masses, is then implemented in order to keep only one B s candidate per event. The M bc and ∆ E distributions of the selected B s candidates for the three D − s modes are shown in Figs. 1 and 2, where the various components of the probability density function(PDF) used for the fit are described. The B s → D ∗− s ρ + candidates are observed with two additionalvariables, cos θ D ∗− s and cos θ ρ + , which are the cosines of the helicity angles defined above. Theyare needed for the measurement of the longitudinal polarization fraction, f L .Table 1 presents a summary of the numerical results obtained for the B s → D ( ∗ ) − s π + and B s → D ( ∗ ) − s ρ + modes. The different sources of systematic uncertainties affecting the measurements areidentified and quoted as a second error. Our results on the B s decays are consistent with theoreticalpredictions [11, 12] and with existing measurements (Table 1).4 s Decays at Belle
Remi Louvot
Mode N B ∗ s ¯ B ∗ s S ε (10 − ) B (10 − ) B World average (10 − ) B s → D − s π + + − σ . . + . − . ± . ± . . ± . B s → D ∗− s π + . + . − . . σ .
13 2 . + . − . ± . ± . B s → D − s ρ + . + . − . . σ .
40 8 . + . − . ± . ± . B s → D ∗− s ρ + . + . − . . σ .
67 11 . + . − . ± . ± . m ( B s ) ( . ± . ± . ) MeV / c ( . ± . ) MeV / c [20] m ( B ∗ s ) ( . ± . ± . ) MeV / c ( . ± . ) MeV / c [21] f B ∗ s ¯ B ∗ s (cid:0) . + . − . ± . (cid:1) % ( + − ) % [4] f B ∗ s ¯ B s (cid:0) . + . − . ± . (cid:1) % First measurement f B s ¯ B s (cid:0) . + . − . (cid:1) % First measurement f L ( B s → D ∗− s ρ + ) . + . − . + . − . First measurement
Table 1:
Summary of the results for the four B s → D ( ∗ ) − s π + and B s → D ( ∗ ) − s ρ + modes [9, 10]. Top: signalyields in the B ∗ s ¯ B ∗ s production mode, N B ∗ s ¯ B ∗ s , significances, S , including systematics, total signal efficiencies, ε (including all sub-decay branching fractions), and branching fractions, B , where the uncertainty due to f s (third error) is separated from the others systematics (second error). The first error represents the statis-tical uncertainties. Bottom: other measurements obtained with the B s → D − s π + analysis and B s → D ∗− s ρ + longitudinal polarization fraction. The world averages (made without the measurements presented here) areshown for comparison in the last column of the tables.
2. Study of B s → J / B s → J / B s → J / ψ η ( (cid:48) ) and Search for B s → J / B s → J / B s → J / ψ f f f (980) B s decays to CP eigenstates are important for CP -violation parameter measurements [22].Results about the first observation of B s → J / ψ η and the first evidence for B s → J / ψ η (cid:48) are re-ported [23]. The J / ψ candidates are formed with oppositely-charged electron or muon pairs, while η candidates are reconstructed via the η → γγ and η → π + π − π modes. A mass (mass and vertex)constrained fit is then applied to the η ( J / ψ ) candidates. The η (cid:48) candidates are reconstructed viathe η (cid:48) → ηπ + π − and η (cid:48) → ρ γ modes, while the ρ candidates are selected from π + π − pairs.If more than one candidate per event satisfies all the selection criteria, the one with the smallestfit residual is selected. The main background is the continuum, which is reduced by requiring R < .
4. The combined M bc and ∆ E distributions are presented in Figs. 3 ( B s → J / ψ η ) and 4( B s → J / ψ η (cid:48) ). We obtain B ( B s → J / ψ η ) = ( . ± . ( stat . ) + . − . ( syst . ) ± . ( f s )) × − and B ( B s → J / ψ η (cid:48) ) = ( . ± . ( stat . ) + . − . ( syst . ) ± . ( f s )) × − . This is, respectively, the first ob-servation (7 . σ ) and the first evidence (3.8 σ ) for these modes.The B s → J / ψ f ( ) mode is especially interesting for the hadron-collider experiments be-cause it has only four charged tracks in its final state. Recent calculations predict the ratio R f / φ = B ( B s → J / ψ f ( )) × B ( f ( ) → π + π − ) B ( B s → J / ψ φ ) × B ( φ → K + K − ) s Decays at Belle
Remi Louvot proj.
Figure 3: M bc (left) and ∆ E (right) distributions, similarly to Fig. 1, of the B s → J / ψ η candidates (pointswith error bars) and the fitted PDF (solid line). The sub-modes η → γγ and η → π + π − π , which are fittedseparately, are summed in these plots. The green-dotted line (red region) represents the continuum (signal)component of the PDF. The small peak in the M bc plot is the B ∗ s ¯ B s contribution, as the B ∗ s ¯ B ∗ s signal range in ∆ E overlaps with that of the B ∗ s ¯ B s signal. Figure 4: M bc (left) and ∆ E (right) distributions, similarly to Fig. 1, of the B s → J / ψ η (cid:48) candidates (pointswith error bars) and the fitted PDF (solid line). The green-dotted line represents the continuum componentof the PDF. The red region represents the signal component of the PDF. to be ≈ . D + s → f ( ) e + ν e , R f / φ is estimated to be 0 . ± .
11 [25]. From QCD estimates [26] and BES result of B ( f ( ) → π + π − ) , R f / φ ≈ .
24. Otherpredictions from generalized QCD factorization [27] are compatible with these estimates.With the same selection for the J / ψ as described above, and the reconstruction of f ( ) → π + π − candidates, the B s → J / ψ f ( ) signal is fitted using the energy difference, ∆ E , and the f ( ) mass, M π + π − , distributions (Fig. 5). No significant signal (6 . ± . σ ) is seenand we set the upper limit B ( B s → J / ψ f ( )) × B ( f ( ) → π + π − ) < . × − (at 90% C.L.) , or, similarly, R f / φ < .
275 (at 90% C.L.)6 s Decays at Belle
Remi Louvot
Figure 5: f ( ) mass (left) and ∆ E (right) distributions of the B s → J / ψ f ( ) candidates. The solid-black line is the total fitted PDF. The green region represents the contribution of the non-resonant B s → J / ψ π + π − , while the red region is the signal. The dotted-black curve is the contribution of the other B s → J / ψ X modes. using our preliminary result of B ( B s → J / ψ φ ) [28]. These limits are clearly in the region of interestand an update using our full data sample (120 fb − ) is very important.
3. Observation of B s → K + K − B s → K + K − B s → K + K − and Searches for B s → π + π − B s → π + π − B s → π + π − , B s → K − π + B s → K − π + B s → K − π + and B s → K S K S B s → K S K S B s → K S K S We present our results for the B s → K + K − , B s → K − π + , B s → π + π − and B s → K S K S charm-less decays [29]. The B s → K + K − mode is particularly interesting because it can be used for thedetermination of the CKM angle γ [30] and may be sensitive to New Physics [31]. The chargedpion and kaon candidates are selected using charged tracks and identified with energy deposi-tion, momentum and time-of-flight measurements. The K S candidates are reconstructed via the K S → π + π − decay, by selecting two oppositely-charged tracks matching various geometrical re-quirements [32]. A likelihood based on a Fisher discriminant using 16 modified Fox-Wolframmoments [33] is implemented to reduce the continuum, which is the main source of background.We do observe a 5.8 σ excess of 24 ± B ∗ s ¯ B ∗ s region for the B s → K + K − mode(Fig. 6). The branching fraction B ( B s → K + K − ) = ( . + . − . ( stat . ) ± . ( syst . ) ± . ( f s )) × − is derived. However, no significant signal is seen for the other modes. Including the systemat-ics uncertainties, we set the following upper limits at 90% confidence level: B ( B s → π + π − ) < . × − , B ( B s → K − π + ) < . × − and, assuming B ( B s → K ¯ K ) = × B ( B s → K S K S ) , B ( B s → K ¯ K ) < . × − . The later is the first limit set for the B s → K ¯ K mode. All the othervalues are compatible with the CDF results [34, 35]. Figure 6:
Distributions, similarly toFig. 1, of the B s → K + K − candi-dates and the fitted PDF (solid blueline). The solid-red and the dotted-grey curves represent the signal and thecontinuum component of the PDF, re-spectively. s Decays at Belle
Remi Louvot
4. Study of B s → D ( ∗ )+ s D ( ∗ ) − s B s → D ( ∗ )+ s D ( ∗ ) − s B s → D ( ∗ )+ s D ( ∗ ) − s and Measurement of ∆Γ CP s / Γ s ∆Γ CP s / Γ s ∆Γ CP s / Γ s Figure 7: ∆ E (left) and M bc (right) distributions,similarly to Fig. 1, ofthe B s → D + s D − s (top), B s → D ∗± s D ∓ s (middle)and B s → D ∗ + s D ∗− s (bot-tom) candidates, togetherwith the fitted PDF. Exceptthe continuum back-ground component, whichis shown by the blackdashed-dotted curve, allthe other contributions arepeaking in M bc . The cor-rect (wrong) combinationsignal, shown by the peak-ing (smooth) red dashedcurve and the cross-feedcomponents, shown by theblue dashed-dotted curveare well separated in ∆ E . We finally report the results from our analysis of the B s → D ( ∗ )+ s D ( ∗ ) − s decays [36]. Thesemodes are CP eigenstates and CKM favored ( b → c ¯ cs transition of order λ ). In the heavy-quarklimit, they are CP even and dominate ∆Γ [37]. The relative width difference of the B s − ¯ B s systemcan be obtained from the relation ∆Γ CP s Γ s = × B ( B s → D ( ∗ )+ s D ( ∗ ) − s ) − B ( B s → D ( ∗ )+ s D ( ∗ ) − s ) . (4.1)In order to reconstruct the B s → D ( ∗ )+ s D ( ∗ ) − s candidates, we form D − s candidates from 6 modes: D − s → φ π − , D − s → K ∗ K − , D − s → K S K − , D − s → φ ρ − , D − s → K ∗ K ∗− and D − s → K S K ∗− . Onlyone candidate per event is selected using M ( D − s ) and M ( D ∗− s ) − M ( D − s ) informations. The samelikelihood as in the previous Section, based on modified Fox-Wolfram moments [33], is used toreject 80% of the continuum events, while 95% of the signal is kept. The ∆ E and M bc distributionsfor each of the three B s → D ( ∗ )+ s D ( ∗ ) − s modes are fitted simultaneously. The signal PDF is madeof two components studied with signal MC simulations: the correctly reconstructed candidates andthe wrong combinations in which a non-signal track (photon) is included in place of a true daughtertrack (photon). In addition the so-called cross-feed contributions are included: a D ∗± s D ∓ s ( D ∗ + s D ∗− s )event can be selected as a D + s D − s ( D ∗± s D ∓ s ) candidate with a lower energy because one photon8 s Decays at Belle
Remi Louvot is missing; conversely, a D + s D − s ( D ∗± s D ∓ s ) candidate can be reconstructed as a D ∗± s D ∓ s ( D ∗ + s D ∗− s )candidate with an additional photon, hence its energy larger than expected.Mode N sig . S B B
World Average B s → D ∗ + s D ∗− s . + . − . σ ( . + . − . ( stat . ) ± . ( syst . )) % First evidence B s → D ∗± s D ∓ s . + . − . σ ( . + . − . ( stat . ) ± . ( syst . )) % First observation B s → D + s D − s . + . − . σ ( . + . − . ( stat . ) + . − . ( syst . )) % ( . + . − . ) % B s → D ( ∗ )+ s D ( ∗ ) − s . + . − . ( . + . − . ( stat . ) ± . ( syst . )) % ( . ± . ) % Table 2:
Signal event yields, N sig . , significances, S , including systematics and branching fractions, B ,for the three B s → D ( ∗ )+ s D ( ∗ ) − s modes and their sum. The world averages, performed from other existingmeasurements [38 – 40], are those reported in Ref. [41]. The fit results can be seen in Fig. 7 while the numerical values are reported in Table 2. WithEq. (4.1), we extract ∆Γ CP s Γ s = ( . + . − . ( stat . ) + . − . ( syst . )) × − . This value is in agreement with the SM expectations [42] and with the results from ALEPH, ( + − ) % [38], DØ, ( . ± . ) % [40], and CDF , ( + − ) % [43]. With only 23 fully-reconstructedsignal events, our measurement is already competitive with the Tevatron values. Conclusion
We presented new results on B s decays obtained from 23.6 fb − of ϒ ( S ) data recorded by theBelle detector. While modes with large statistics can provide precise measurements of branchingfractions and B ( ∗ ) s properties, first observations of several CP -eigenstate B s decays are a confirma-tion of the large potential of our 120 fb − e + e − → ϒ ( S ) data sample and advocate an ambitious B s program at super- B factories. References [1] A. Abashian et al. (Belle Collaboration)
Nucl. Instrum. Methods Phys. Res., Sect. A (2002) 117.[2] S. Kurokawa and E. Kikutani
Nucl. Instrum. Methods Phys. Res., Sect. A (2003) 1.[3] A. Drutskoy et al. (Belle Collaboration)
Phys. Rev. Lett. (2007) 052001.[4] A. Drutskoy et al. (Belle Collaboration) Phys. Rev. D (2007) 012002.[5] A. Drutskoy et al. (Belle Collaboration) Phys. Rev. D (2010) 112003.[6] K.F. Chen et al. (Belle Collaboration) Phys. Rev. Lett. (2008) 112001.[7] G.S. Huang et al. (CLEO Collaboration)
Phys. Rev. D (2007) 012002.[8] C. Amsler et al. (Particle Data Group) Phys. Lett. B (2008) 1.[9] R. Louvot et al. (Belle Collaboration)
Phys. Rev. Lett. (2009) 021801. This is a measurement of ∆Γ s / Γ s = ( ∆Γ CP s cos φ ) / Γ s , φ being the CP -violating phase (assumed to be negligible). s Decays at Belle
Remi Louvot[10] R. Louvot et al. (Belle Collaboration)
Phys. Rev. Lett. (2010) 231801.[11] A. Deandrea et al. Phys. Lett. B (1993) 549.[12] R.H. Li, C.D. Lü and H. Zou
Phys. Rev. D (2008) 014018.[13] L. Wolfenstein Phys. Rev. Lett. (1983) 1945.[14] N. Cabibbo Phys. Rev. Lett. (1963) 531.[15] M. Kobayashi and T. Maskawa Prog. Theor. Phys. (1973) 652.[16] B. Adeva et al. (LHCb Collaboration), LHCb-PUB-2009-029, arXiv:0912.4179v1 [hep-ex] (2009).[17] B. Aubert et al. (BaBar Collaboration) Phys. Rev. Lett. (2003) 171802.[18] K.F. Chen et al. (Belle Collaboration) Phys. Rev. Lett. (2003) 201801.[19] G.C. Fox and S. Wolfram Phys. Rev. Lett. (1978) 1581.[20] W.M. Yao et al. (Particle Data Group) J. Phys. G (2006) 1. and 2007 partial update for the 2008edition.[21] O. Aquines et al. (CLEO Collaboration) Phys. Rev. Lett. (2006) 152001.[22] I. Dunietz, R. Fleischer and U. Nierste Phys. Rev. D (2001) 114015.[23] I. Adachi et al. (Belle Collaboration). Belle-conf-0902, arXiv:0912.1434 [hep-ex] (2009).[24] S. Stone and L. Zhang Phys. Rev. D (2009) 074024.[25] K.M. Ecklund et al. (CLEO Collaboration) Phys. Rev. D (2009) 052009.[26] P. Colangelo, F. De Fazio and W. Wang Phys. Rev. D (2010) 074001.[27] O. Leitner et al. Phys. Rev. D (2010) 076006.[28] R. Louvot. Talk presented at the Lake Louise Winter Institute 2009, Alberta (Canada, February 2009);arXiv:0905.4345v2 [hep-ex] (2009).[29] C.C. Peng et al. (Belle Collaboration) Phys. Rev. D (2010) 072007.[30] R. Fleischer Phys. Lett. B (1999) 306.[31] D. London and J. Matias
Phys. Rev. D (2004) 031502.[32] F. Fang, “Measurement of Branching Fractions and CP Violation in B → η c K and Observation of B ± → p ¯ pK ± .” Ph.D. thesis, University of Hawaii (2003).[33] S.H. Lee et al. (Belle Collaboration) Phys. Rev. Lett. (2003) 261801.[34] A. Abulencia et al. (CDF collaboration) Phys. Rev. Lett. (2006) 211802.[35] T. Aaltonen et al. (CDF Collaboration) Phys. Rev. Lett. (2009) 031801.[36] S. Esen et al. (Belle Collaboration).
Phys. Rev. Lett. (in press); arXiv:1005.5177 [hep-ex] (2010).[37] R. Aleksan et al. Phys. Lett. B (1993) 567.[38] R. Barate et al. (ALEPH Collaboration)
Phys. Lett. B (2000) 286.[39] T. Aaltonen et al. (CDF Collaboration)
Phys. Rev. Lett. (2008) 021803.[40] V.M. Abazov et al. (D0 Collaboration)
Phys. Rev. Lett. (2009) 091801.[41] K. Nakamura et al. (Particle Data Group)
J. Phys. G (2010) 075021.[42] A. Lenz and U. Nierste J. High Energy Phys. (2007) JHEP06(2007)072.[43] T. Aaltonen et al. (CDF Collaboration)
Phys. Rev. Lett. (2008) 121803.(2008) 121803.