m ES (GeV/c ) B(Kπ) (10-6 ) (GeV/c ) m ES B(ππ) (10-6 ) (GeV/c ) Events / 0.02 GeV Events / 0.02 GeV Events / 0.02 GeV (f) (c) E (GeV)

Transcription

1 SLACíPUBí695 BABAR-PROCíè7 hep-exè November Charmless Hadronic B Decays at BABAR J. Olsen Physics Department, University of Maryland College Park, MD, 7- èrepresenting the BABAR Collaborationè Abstract We present preliminary results of several searches for rare charmless hadronic decays of the B meson using data collected by the BABAR detector at the Stanford Linear Accelerator Center's PEP-II storage ring. We search for the decays h + h,, h + h, h +, h + h, ç, X h +, and X KS, where h = ç or K, and X = ç or!. In a sample of : million BB decays we measure the branching fractions: BèB! ç + ç, è=è9: +:6 +:,:,:è æ,6, BèB! K + ç, è = è:5 +: +:,:6,:7è æ,6, BèB! ç, ç + è = è9 æ +6,5è æ,6, and BèB +! ç K + è = è6 æ æ è æ,6. We calculate upper limits for the modes without a signiæcant signal. Contributed to the Meeting of the Division of Particles and Fields of the American Physical Society Columbus, Ohio, USA August 9íAugust, Stanford Linear Accelerator Center, Stanford University, Stanford, CA 99 Work supported in part by Department of Energy contract DE-AC-76SF55.

2 Introduction Charmless hadronic B decays will play an important role in the study of CP violation. Indirect CP violation arises in B íb mixing due to interference between direct and mixed decays. The CKM angle æ can be measured by observing the resulting time-dependent asymmetry in decays to çç and çç ænal states. Direct CP violation results from interference between two or more weak amplitudes and can arise in any decay mode where both tree and penguin contributions are non-negligible. Several modes reported in this paper are ëself-tagging", providing eæcient samples for direct CP violation searches. Finally, accurate branching fraction measurements provide important tests of factorization models, which facilitate calculation of æ in the presence of signiæcant penguin amplitudes, and can also be used to constrain the CKM angle æ. ëë In this paper we summarize preliminary results of searches for the following charmless hadronic B decays: ëë æ ç + ç,, K + ç,, K + K,, æ K æ ç +, ç K +, ç ç +, ç, ç +, K + ç, ç +, ç + ç, ç +, æ ç K +, ç K S,!h +,!K S, where charge conjugate modes are assumed throughout. The dataset consists of : million BB decays collected by the BABAR detector ëë at the PEP-II storage ring between January and June. Candidate Selection and Analysis Method We use only good quality tracks with a minimum transverse momentum of MeV=c in the laboratory èlabè frame. Charged pions and kaons are identiæed by their energy loss ède=dxè in the tracking system and the angle ç c of çcerenkov photons produced while traversing quartz bars ëë. Neutral kaons are reconstructed in the mode KS! ç + ç,, requiring the KS æight length to exceed mm and the angle between the æight direction and momentum to be less than mr. Photon candidates are deæned as calorimeter energy deposits unassociated with a track and having a shower shape consistent with the photon hypothesis. Candidate ç and ç mesons are formed from pairs of photons with a minimum LAB energy of 5 MeV. Candidate ç mesons are reconstruced in the channel çç + ç,, where the ç mass is constrained to the world average value. The! meson is reconstructed in the dominant decay channel,!! ç + ç, ç,keeping all candidates within 5 MeV=c of the known! mass. The ç and K æ resonances are reconstructed in the corresponding çç and Kç channels. We select candidate B mesons based on the energy-substituted mass m ES, where p s= is substituted for the candidate's energy, and the diæerence æe between the B-candidate energy and p s=. The dominant background for all modes is continuum qçq production, which exhibits a jet-like structure that distinguishes it from the more spherically symmetric BB events. To suppress this background we use the cosine of the angle ç T èç S è between

3 B(Kπ) ( -6 ) BABAR (statistical only) σ σ σ σ 5σ B(ππ) ( -6 ) Events /.5 GeV/c Events /.5 GeV/c Events /.5 GeV/c 6 (a) (b) (c) m ES (GeV/c ) m ES (GeV/c ) m ES (GeV/c ) Events /. GeV Events /. GeV Events /. GeV 7 6 (d) E (GeV) 7 6 (e) E (GeV)..6.. (f) E (GeV) Figure : Left: The central value èælled circleè for BèB! ç + ç, è and BèB! K + ç, è along with the nç statistical contour curves for the global likelihood æt. Right: m ES and æe for èa,dè çç, èb,eè Kç, and èc,fè KK candidates in the cut-based analysis. the thrust èsphericityè axis of the B candidate and the rest of the event, and the cosine of the angle ç B between the candidate's æight direction and the beam axis. In some cases we include several event-shape variables into a single Fisher discriminant. Results for h + h, Modes We select B! h + h, candidates satisfying 5: ém ES é 5:GeV=c and jæej é : GeV. No explicit particle identiæcation is required and the pion mass hypothesis is assumed for both tracks. We require jcos ç S j é :9 and construct a Fisher discriminant F from nine variables describing the momentum æow of charged and neutral particles around the B candidate thrust axis. Signal yields in all three modes are determined simultaneously from an unbinned maximum likelihood æt incorporating m ES,æE,F, and the measured ç c for each track. A sample of D æ -tagged D! K + ç, decays is used to parameterize the ç c distributions for pion and kaon tracks as a function of momentum. The K=ç separation varies from to ç across the relevant momentum range. All candidates in the region,: é æe é : GeV are included in the æt. We ænd signal yields of Nèççè =9 +,7, NèKçè= +9,, and NèKKè=7 +5,. As a cross-check we perform a cut-based analysis requiring a tighter cut on cos ç S and addi-

4 tional cuts on cos ç B and F. Signal yields are determined by applying particle identiæcation criteria to isolate independent samples of candidates corresponding to each mode and then ætting the m ES distribution in each sample. The results are consistent with the global likelihood æt. Figure shows the global æt likelihood contour curves for the çç and Kç modes, and the m ES and æe distributions for the cut-based analysis. The results are summarized in the upper section of Table. For the KK mode we calculate the 9è conædence level upper limit. The dominant systematic errors are due to tracking eæciency and the shapes of the æe and F distributions. Results for Three-body Modes We search for resonant three-body decays by combining a ç or K æ resonance with a charged pion or kaon. Kaons are required to be positively identiæed using de=dx and ç c information, while tracks not identiæed as kaons are assumed to be pions. We veto any combination consistent with the decay D! K, ç +. The selection criteria consist of optimized cuts on cos ç T, resonance mass, and the angle between the resonance daughters and the B candidate momentum calculated in the rest frame of the vector meson. We also explicitly search for non-resonant K + ç, ç + and ç + ç, ç + decays by removing all Kç and çç combinations with invariant mass less than GeV=c, and all three-body combinations consistent with the decay B +! J=èK +. We deæne a signal region within 6 MeV=c of the B mass in m ES and æ7 MeVinæE. The signal yield is determined by direct background subtraction, where the background in the signal region is estimated from the number of events in the region 5: ém ES é 5:7 GeV=c. This method is cross-checked using oæ-resonance data. The results are summarized in the middle section of Table. The dominant systematic errors are due to tracking eæciency, ç eæciency, and the background subtraction technique. 5 Results for Modes with ç or! We search for the modes ç K +, ç KS,!h +, and!ks. For ç K the kaon is positively identiæed, while for!h + the charged hadron is assumed to be a pion and the æe signal window is increased è,: é æe é :7 GeVè to take into account the resulting shift in energy when the mass is mis-assigned. The angle between the decay plane of the! daughters and the B direction in the! rest frame is used to reduce combinatoric background. We require jcos ç T j é :9 and optimize with respect to F. Signal yields are determined by background subtraction, where the background is determined from oæ-resonance data. The results are summarized in the lower third of Table. The dominant systematic errors are the same as in the three-body analysis.

5 Table : Branching fraction results. Signal yields èn S è for the h + h, modes are determined from a likelihood æt, the rest are obtained by a direct background subtraction. Eæciencies èæè include intermediate branching fractions. Mode N S Stat. Sig. èçè æèèè B è,6 è B! ç + ç, 9 + +,7, : +:6 +:,:,: B! K + ç, +9 +,, :5 +: +:,:6,:7 B! K + K, 7 +5, èé 5è. 5 é 6:6 B +! K æ ç + : æ :. é B +! ç K + :7 æ 5:. é 9 B +! K + ç, ç + 6: æ 5:. 6 é 5 B +! ç ç + :9 æ :. é 9 B +! ç + ç, ç + 5: æ 5:7.7 é B! ç, ç + 5:5 æ 9:.5 9 æ +6,5 B +! ç K + : æ : æ æ B! ç K : æ :. :6 é B +!!h + 5:9 æ :6.7 7:5 é B!!K,: æ :. é 6 Summary We have presented preliminary results of searches for several charmless hadronic B decays. Table summarizes the results. In all cases, our results are consistent with recent measurements reported by the CLEO ëë and Belle ë5ë collaborations at this conference. Acknowledgments We are grateful for the contributions of our PEP-II colleagues in achieving the excellent luminosity and machine conditions that have made this work possible. We acknowledge support from the Natural Sciences and Engineering Research Council ècanadaè, Institute of High Energy Physics èchinaè, Commissariat ça l'energie Atomique and Institut National de Physique Nuclçeaire et de Physique des Particules èfranceè, Bundesministerium fíur Bildung und Forschung ègermanyè, Istituto Nazionale di Fisica Nucleare èitalyè, The Research Council of Norway, Ministry of Science and Technology of the Russian Federation, Particle Physics and Astronomy Research Council èunited Kingdomè, the Department of Energy èusè, and the National Science Foundation èusè. In addition, individual support has been received from the Swiss National Foundation, the A. P. Sloan Foundation, the Research Corporation, and the Alexander von Humboldt Foundation. The visiting groups wish to thank SLAC for the support and kind hospitality extended to them. 5

6 References ëë M. Neubert, ëqcd Factorization and CP Violation in Hadronic B Decays", contributed to this conference. ëë For more detailed descriptions of these results see: BABAR Collaboration, B. Aubert et al., BABAR-CONF-è and BABAR-CONF-è5, submitted to the XXX th International Conference on High Energy Physics, Osaka, Japan, July. ëë BABAR Collaboration, B. Aubert et al., BABAR-CONF-è7, submitted to the XXX th International Conference on High Energy Physics, Osaka, Japan, July. ëë D. Urner, ërare B Decays at CLEO", contributed to this conference. ë5ë B. Casey, ërare B Decays without Charm from BELLE", contributed to this conference. 6