Event Reconstruction and Physics Performance of the LHCb Experiment HCP2005, 4-9 July 2005, les Diablerets Switzerland MC truth Reconstructed Event Reconstruction and Physics Performance of the LHCb Experiment Yuehong Xie University of Edinburgh representing the LHCb collaboration

Outline Introduction Event reconstruction capability Physics potential Tracking Vertexing and proper time determination Flavour tagging Particle identification Physics potential Physics programs Bs Mixing CKM angles Rare B decays Conclusions HCP2005, 4-9 July 2005 Yuehong Xie

Introduction Goal of LHCb: look for new physics in CP violation and rare B decays Time-dependant measurements: exclusive signal reconstruction, proper time determination and B flavour tagging Requirements to reconstruction: Exclusive B reconstruction requires good mass resolution → need precise momentum measurement up to 100 GeV Very fast Bs oscillation requires a good proper time resolution → need high precision of primary and secondary vertex position Flavour tagging and background rejection require efficient particle identification for 2-100 GeV HCP2005, 4-9 July 2005 Yuehong Xie

LHCb detector Tracking: Vertex Locator TT T1-T3 Magnet PID: Related LHCb talks: LHCb Status: Roger Forty Particle Identification: Ann Van Lysebetten PID: RICH1/RICH2 SPD/PS ECAL HCAL M1-M5 HCP2005, 4-9 July 2005 Yuehong Xie

Track finding strategy T track Upstream track VELO seeds Long track (forward) Long track (matched) VELO track T seeds Downstream track Multiple pass track finding: Velo seeds T seeds Long tracks  highest quality for physics Downstream tracks  needed for efficient KS finding Upstream tracks  lower p, worse p resolution, useful for RICH1 pattern recognition HCP2005, 4-9 July 2005 Yuehong Xie

Tracking performance Eff = 94% (p > 10 GeV) Long track efficiency vs p Ghost rate = 3% (for pT > 0.5 GeV) Ghost rate vs p : dp/p ~ 0.37% A typical reconstructed event: red = detected hits On average: 26 long tracks 11 upstream tracks 4 downstream tracks 5 T tracks 26 VELO tracks Blue = reconstructed tracks HCP2005, 4-9 July 2005 Yuehong Xie

Mass resolution Mass resolution based on the dp/p ~ 0.37% momentum resolution Mass of Ds-K+K-- Mass of BsDs (KK) K+ Ds mass [GeV/c2] HCP2005, 4-9 July 2005 Yuehong Xie

Vertex detection dIP = 14mm+35mm/pT <dIP> ~ 40mm R-sensor 21 silicon stations placed along the beam direction, 8mm to beam Two detector halves in each station Two sensor types to measure R and f Track impact parameter precision to origin vertex: dIP = 14mm+35mm/pT <dIP> ~ 40mm R-sensor f-sensor HCP2005, 4-9 July 2005 Yuehong Xie

Vertex reconstruction proper time t=lm/(pc) Proper time resolution is dominated by B vertex resolution Bs→Dsp sz=168 mm Proper time sz1=43.9 mm sz2=124 mm (21.8%) st1 = 33 fs st2 = 67 fs (31%) Primary Vertex HCP2005, 4-9 July 2005 Yuehong Xie

Particle identification performance K/p separation is provided by RICH Lepton identification is mainly provided by calorimeter and muon chambers Typical event in the RICH1 photon detectors: Hits from signal tracks and background tracks Reconstructed rings: large from aerogel, small from gas C4F10 RICH PID 2<p<100 GeV/c <e(KK)> = 88% <e(K)> = 3% Lepton ID <e(mm)> = 94% <e( m)> = 1% <e(ee )> = 81% <e( e)> = 1% 95% within calo acceptance HCP2005, 4-9 July 2005 Yuehong Xie

Flavour tagging Opposite side Same side Figure of merit: High Pt leptons K± from b → c → s Vertex charge Jet charge K+ Qvertex ,QJet Same side Opposite side PV e-, m- Bs0signal D K + K- B0opposite Same side Fragmentation K± accompanying Bs p± from B** → B(*)p± Tagging power in % Tag Bd Bs Muon 1.2 1.4 Electron 0.6 Kaon opp.side 2.1 2.4 Jet/ Vertex Charge 0.7 0.8 Same side p / K 0.7 (p) 3.1 (K) Combined ~ 4.4 ~7.5 Figure of merit: eD2=e(1-2w)2: tagging power e: tagging efficiency; w: wrong tagging fraction ( Same opposite side performance for Bd and Bs within errors) HCP2005, 4-9 July 2005 Yuehong Xie

Physics Programs 1012 bb pairs produced in a nominal year (107 s at L=21032 cm2s1) Bd : Bu: Bs: Bc: b-baryons ~ 40 : 40 : 10 : 0.1 : 10 Measuring Bs mixing to look for new physics effect in box diagrams Dms from Bs→Dsp fs and DGs from Bs→J/yf Measuring CKM angles from different processes to over-constrain unitary triangle sin(2b) from Bd → J/yKs g from Bs → DsK , Bd →D*p g from Bd → D0K* g from B → hh a from Bd → rp Looking for new physics effect in rare B decays leptonic rare decays: Bs → mm Semileptonic rare decays: Bd → K*mm Radiative penguin: Bs → fg, Bd → K*g and Bd → r/wg Hadronic penguin: Bs → ff and Bd → fKs Other physics CP violation in Bc decays b hadron spectrum, lifetimes (ratio) charm physics … Red: more details to come HCP2005, 4-9 July 2005 Yuehong Xie

Bs mixing: Dms from Bs→Dsp Statistical uncertainty on Bs oscillation as a function of Dms New physics probe 80k events/year B/S = 0.32 ± 0.10 st ~ 40 fs If Dms = 25 ps -1 Bs mixing can be clearly observed in Bs Ds with one year of data taking Once oscillation observed, Dms can be measured with very high statistical precision s(Dms)=0.01 ps-1 5s sensitivity in one year Dms< 68 ps-1 is well beyond SM preferred value Dms ~ 20 ps-1 HCP2005, 4-9 July 2005 Yuehong Xie

Bs mixing: DGs and fs from Bs→J/yf SM fs = -2l2h ~ -0.04, sensitive to new weak phases in box process DGs: input to other measurements; test of HQET Gold-plated mode: 120k events/year, B/S<1 Admixture of CP-even and CP-odd eigenstates Kinematic variables for angular analysis Fit proper time and angular distribution simultaneously use Dms from Bs→Dsp Can be combined with CP eigenstate modes ( Bs→J/yh, hcf ) to improve precision Using untagged samples can achieve similar sensitivity of DGs HCP2005, 4-9 July 2005 Yuehong Xie

g from Bs→DsK 5.4k events/year, B/S<1 Measure g+fs in tree diagrams fit four time dependant decay rates Bs ( Bs) → Ds- K+ + cc use fs measured in Bs→J/yf Free of new physics contribution to penguin diagrams Measured g is not affected by new physics effect in Bs mixing Discrepancy with indirect g from CKM fit will signify new physics in mixing U-spin combination with Bd→D(*)p to resolve discrete ambiguity 200k events/year s Ds- c b u K+ Bs s s s K+ u Bs b c Ds- s s sensitivity in one year: Dms 20ps-1 25ps-1 30ps-1 s(g) 14.2 16.2 18.3 (55 < g <105 ) HCP2005, 4-9 July 2005 Yuehong Xie

g from B→hh d /K Bd/s Bd/s /K Bd (95%CL) Measure time dependant asymmetries for Bd→pp and Bs→KK to determine Adir and Amix ACP(t) = Adir cos(Dm t) + Amix sin(Dm t) Adir and Amix depend on g Mixing phases fd or fs Penguin/Tree=deiq Use fd and fs from J/ and J/Ks U-spin symmetry: dpp=dKK, qpp=qKK 4 observables, 3 unknowns: solve for g BsKK (95%CL) 26k Bd→pp events/year, B/S<0.7 37k Bs→KK events/year, B/S=0.31±0.1 s(g) ~ 5+ uncertainty from U-spin symmetry breaking Sensitive to new physics in penguin HCP2005, 4-9 July 2005 Yuehong Xie

g from Bd→D0K*0 Gronau-Wyler-Dunietz method can be simply illustrated as Measure 6 time-integrated tree level rates Self-tagged through K*0 or K*0 Only relative rates needed Big ratio: |A2/A1| ~ 0.4 Need to resolve 8-fold ambiguity New physics in D0-D0 mixing could affect this measurement Mode Yield S/B Bd  D0 (K-+) K*0 3400 > 3.3 Bd  D0 (K+-) K*0 500 > 0.6 Bd  D0CP (K+K-) K*0 600 > 0.7 () ~ 8 in a year (55 <  < 105, -20 < D < 20) HCP2005, 4-9 July 2005 Yuehong Xie

a from Bd→ rp → p +p -p 0 ( Further work: systematics; a from B→rr ) reconstructed & selected generated M(p0p-) M(p0p-) M(p0p+) po reconstruction efficienty — merged p0 — resolved p0 M(p0p+) 14k event/year with B/S=0.80 Time dependent Dalitz analysis Theoretically clean extraction of a from a 9-parameter fit, taking into account penguin contribution and resonant background Statistical precision in one year s(a) < 10 ( Further work: systematics; a from B→rr ) HCP2005, 4-9 July 2005 Yuehong Xie

Bs  +- SM: Br = (3.5±1.0) x 10-9 Can be significantly enhanced in some SUSY models Ideal candidate for new physics 17 events/year expected assuming SM Branching ratio Present limit B/S < 5.7 at 90% c.l. from bmx+cc * Require more MC events to obtain precise estimate of B/S Challenges Background suppression trigger and selection efficiency determination Normalization method Worst case mass plot with 5 years data * Events with m from semi-leptonic b decays have been found to be dominating background HCP2005, 4-9 July 2005 Yuehong Xie

Conclusions The LHCb detector and software can efficiently reconstruct many different B decay channels with good performance in proper time resolution, particle identification and mass resolution This enables LHCb to fully explore Bs mixing and decays extract CKM parameters with high precision in different ways perform study of rare B decays which are sensitive to new physics LHCb will have a great opportunity to make precision test of SM in flavour sector in order to find new physics or push possible new physics to a higher mass scale, from the first year of data taking. Thank you! HCP2005, 4-9 July 2005 Yuehong Xie