Slide 1

September 8, 2005Donatella Lucchesi1 B Physics Today In this lecture: Machine and Detectors review b hadron production mechanisms and results B hadron lifetimes B-B mixing B meson lifetime differences Next lecture from Stefano Giagu CP Violation Rare decays Donatella Lucchesi CDF experiment INFN Padova Slide 2 September 8, 2005Donatella Lucchesi2 Introduction B physics measurements are high precision measurements High statistic data samples High performances detectors Dedicated machine or pp Dedicated trigger needed Tracking, Particle Identification.. Slide 3 September 8, 2005Donatella Lucchesi3 How & where b are produced AcceleratorBeamsSpeciess(GeV) bb (nb) bb / tot B factories e + e - (4s) B 0 & B 10.51.150.25 Tevatron pp ppall1,960 110 5 610 -4 610 -4 LHC pp ppall14,000 510 5 ? B factories produce only 2 species but very low background pp interactions have all species but high background Slide 4 September 8, 2005Donatella Lucchesi4 What is needed: detectors Magnetic field Tracking system high space resolution high momentum resolution Calorimeters o fast and efficient to identify e o good energy resolution to reconstruct decays as 0 or Particle identification: separate K from identify and e Slide 5 September 8, 2005Donatella Lucchesi5 The Y(4s) Machine Slide 6 September 8, 2005Donatella Lucchesi6 The Y(4s) Machine e - 9.0 GeV/c 2 e + 3.1 GeV/c 2 Slide 7 September 8, 2005Donatella Lucchesi7 Detectors: Babar Slide 8 September 8, 2005Donatella Lucchesi8 Detectors: Babar some details Cherenkov ring Silicon detector Slide 9 September 8, 2005Donatella Lucchesi9 The Tevatron Machine Slide 10 September 8, 2005Donatella Lucchesi10 Tevatron Parameter Substantial upgrades for Run II: 10% energy increase s: 1.8 1.96 integrated luminosity increase: x50 6 km long Tevatron ring 82.32.5 Interactions/crossing 396 3500 Bunch crossing (ns) 50173 Ldt (pb -1 /week) 3 10 32 9 10 31 1.6 10 30 Typical L (cm -2 s -1 ) 1.96 1.8 s (TeV) 36 36 6 6 Bunches in Turn Run IIbRun IIa Run I 1992-1995 2001-2009 Slide 11 September 8, 2005Donatella Lucchesi11 Detectors: CDF Slide 12 September 8, 2005Donatella Lucchesi12 Detectors: CDF some details Central Outer Chamber: 96 layers Max drift time 100 ns Gas: Ar-Et-CF 4 (50:35:15) Silicon tracker Slide 13 September 8, 2005Donatella Lucchesi13 Charm and Beauty production at Tevatron oSince m Q >> QCD for c and b quarks, heavy quark production at the Tevatron should be well-calculable in QCD. oPhysics objects: hadrons & leptons (NOT quarks & gluons) oQuarks hadrons: hadronization (fragmentation) oFragmentation: phenomenological models non perturbative Slide 14 September 8, 2005Donatella Lucchesi14 Charm and Beauty production at Tevatron Diagrams at leading order: Full calculations have been done up to NLO (and beyond) Therefore how do we explain Run 1 Tevatron results? Experiment wrong? Theory prediction incomplete? New physics? b-quarks B + mesons Slide 15 September 8, 2005Donatella Lucchesi15 Recent developments In past years many (theoretical) developments: Use B meson rather than b-quarks: less dependent on unfolding and fragmentation uncertainties Beyond NLO: resummation of log(p T /m) terms FONLL ( Cacciari et al ). Important for medium/high p T region. Extraction of fragmentation function parameters from LEP data in this scheme: substantially different b new PDFs MC@NLO match NLO calculation with PS formalism in HERWIG ( Frixione, Nason, Webber ) o Need more B data to compare with theory o Charm? Slide 16 September 8, 2005Donatella Lucchesi16 Raw data Level 1 Trigger Level 2 Trigger Level 3 Trigger Data set Crossing rate 1.7 MHz Inelastic cross Section 56 mb Acc. Rate 40 KHz Latency 5.5 s Pipeline Acc. Rate 300 Hz Latency 20 s Buffer 4 events Acc. Rate 75 Hz Average size 60 Kbytes/event Level 1 Synchronous streams: Calorimeter Fast tracker Muons Level 2 asynchronous systems: Calorimeter clustering Track parameter available electrons Level 3 Offline-like How CDF Collect Data Slide 17 September 8, 2005Donatella Lucchesi17 CDF Secondary Vertex Trigger (SVT) XFT Tracking Chamber hits d 0, 0, P t SVT hits d0d0 20 s !!! Primary Vertex Secondary Vertex d = impact parameter B Decay Length L Slide 18 September 8, 2005Donatella Lucchesi18 Typical CDF trigger Two tracks vertex trigger: 2 tracks reconstructed by SVT with: pt>2 GeV/c 120 m < d 0 < 1 mm pt1+pt2> 5.5 GeV/c Lepton + displaced track trigger: Lepton (e or ) with pt>4 GeV/c Track reconstructed by SVT with pt>2 GeV/c 120 m < d 0 < 1 mm Di-muon trigger : 2 muons pt>1.5 GeV/c Slide 19 September 8, 2005Donatella Lucchesi19 Charm Production: Open Charm meson CDF: 5.8 pb -1 taken with displaced track trigger. >80% prompt production D0D0 D *+ D+D+ D+D+ Ds+Ds+ 36804 28361 5515 851 Slide 20 September 8, 2005Donatella Lucchesi20 Open Charm Cross Section at CDF Prompt charm production component extracted looking at impact parameter of reconstructed D meson Compare to FONLL theory (Cacciari and Nason): Prompt fraction: D 0 : 86.6 0.4% D *+ : 88.1 1.1% D + : 89.1 0.4% D s + : 77.3 3.8% (3-4% syst. err.) Cross sections for |y|5.5 GeV): 13.31.5 b D *+ (p T >6.0 GeV): 5.20.8 b D + (p T >6.0 GeV): 4.30.7 b D s + (p T >8.0 GeV): 0.750.23 b Theory uncertainty: vary renormalization/factorization scales data at upper limits of theory prediction Slide 21 September 8, 2005Donatella Lucchesi21 b-hadron from J/ Production at CDF b-hadron from J/ Production at CDF (J/ from H b ) = 19.9 3.8 nb (H b J/ , |y| September 8, 2005Donatella Lucchesi37 B 0 flavor tag A second stiff (p>1.0 GeV/c ) lepton required to: o Reduce continuum background o Have a precise z reconstruction o Tag the reconstructed B flavor Data sample: 88x10 6 BB events ~50x10 3 B 0 D *- l candidates ~27x10 3 background events Sample composition and background evaluation from M 2 fit Slide 38 September 8, 2005Donatella Lucchesi38 Fit procedure Binned maximum likelihood fit to t [-18 ps, 18 ps] and ( t) [0 ps, 3 ps] Signal fitting function: Detector response on t: Gaussian parameterization Cascade lepton tag: D evaluated from semileptonic BR Fit free parameters: a) Gaussian parameters for detector response b) (B 0 ) & m d ~1 for signal Slide 39 September 8, 2005Donatella Lucchesi39 Fit results B 0 = 1.501 0.008( stat ) 0.030( syst ) ps m d = 0.523 0.004( stat ) 0.007( syst ) ps -1 Main systematic errors: o Analysis bias o Misalignment of the silicon vertex detector m d = 0.509 0.004 ps -1 World Average Slide 40 September 8, 2005Donatella Lucchesi40 How is measured at Tevatron Production time: bb one B 0 d /B 0 s and b-hadron Decay time: B 0 d /B 0 s decays Needs: - Identify B flavor at the decay time and at the production time - Proper decay time (ct) determination with high resolution Slide 41 September 8, 2005Donatella Lucchesi41 B d and B s Mixing differences Due to the different size of CKM matrix elements B d and B s mixing frequencies are very different B d m d = 0.47 ps -1 B s m s = 17 ps -1 Slide 42 September 8, 2005Donatella Lucchesi42 B s collection DsDs BsBs DsDs BsBs ,e P.V. Fully reconstructed hadronic modes: Complete momentum reconstruction Good proper time resolution High B s mass resolution high S/B Selected by Two Track Trigger (SVT) Two displaced tracks Low statistics Partially reconstructed semileptonic modes: Missing momentum carried by the Visible proper time corrected from MC (K factor) Proper time resolution diluted by missing momentum Selected by dedicated trigger ( l+SVT ) High statistics P.V. Slide 43 September 8, 2005Donatella Lucchesi43 N B s = 52633 S/B ~ 2 M 15 MeV B s hadronic decays B 0 s D s - + D s - - ( K + K - ) D s - K *0 K - D s - + - - Partially reconstructed: B 0 s D s *- + D s *- D s - B 0 s D s - + + 0 (not used) Slide 44 September 8, 2005Donatella Lucchesi44 No B s mass peak: missing particles particles Use D s invariant mass Charge correlation l-D s : signal: l + D s - signal: l + D s - background: l - D s - background: l - D s - l + D s - peak not pure signal ~20% background: ~20% background: D s +fake lepton from PV D s +fake lepton from PV B 0,B + D s D X, D l X B 0,B + D s D X, D l X c-c backgrounds c-c backgrounds B s semileptonic decays B 0 s D s - l + X D s - - ( K + K - ) D s - K *0 K - D s - + - - Slide 45 September 8, 2005Donatella Lucchesi45 The amplitude of mixing asymmetry is diluted by a factor m = 15 ps -1 t =100 fs D t =0.32 Vertex resolution (constant) Momentum resolution (proportional to ct) Hadronic mode like Semileptonic mode like m = 15 ps -1 t =167 fs p /p=15% Proper time resolution Slide 46 September 8, 2005Donatella Lucchesi46 B s decay time resolution D s + track: sample of prompt events used to correct ct and parameterize ct as a function of several variables. Hadronic Decays: : ~ 30 m (100 fs) : ~ 30 m (100 fs) p/p < 1% p/p < 1% Semileptonic Decays: : ~ 50 m (167 fs) : ~ 50 m (167 fs) p/p ~ 15% (K factor for missing neutrino) missing neutrino) Slide 47 September 8, 2005Donatella Lucchesi47 Initial state tagging: Opposite side Identify the flavor of the other B: opposite side tagging reconstruct the other b charge search for a lepton or kaon coming from B decay =N tag /N total efficiency D= N tag W -N tag R /N tag dilution D=1 D=0 D 2 =figure of merit Slide 48 September 8, 2005Donatella Lucchesi48 Same side tagging: infer the production B flavor from particle produced close to the B: fragmentation tracks B ** production and B ** B 0 Initial state tagging: Same Side Same side Bs tagger performances can not be measured from data if setting a limit. Must be understood in Monte Carlo. other way? D 2 =2.33(1.0) 0.34(0.35) % B + (B 0 ) Slide 49 September 8, 2005Donatella Lucchesi49 D max ~0.4 30% mistag rate Opposite Side Tagging: lepton D 2 Muon: (0.70 0.04) % D2 Electron: (0.37 0.03)% Find event with Opposite side B (e)X Low momentum lepton Use likelihood method to combine calorimeter muon detector, dE/dx info High purity Low efficinecy Performances and calibration on SVT +lepton data Slide 50 September 8, 2005Donatella Lucchesi50 Cone based jet algorithm: compute Jet Charge of Secondary Vertex tagged jets Jet Probability tagged jet Highest P jet D max ~0.4 30% mistag rate Opposite Side Tagging: jet charge D 2 secondary vertex: (0.36 0.02) % D 2 displaced track: (0.36 0.03) % D 2 highest p jet: (0.15 0.01) % Slide 51 September 8, 2005Donatella Lucchesi51 Amplitude Scan for B 0 d(s) Introduce Amplitude A in the Likelihood the Likelihood Amplitude scan: Fit for the amplitude A and its error (A) at fixed m Fit for the amplitude A and its error (A) at fixed m Repeat the fit for Repeat the fit for different m different m Amplitude consistent with: 1 if mixing present at 1 if mixing present at the frequency m the frequency m 0 if there is no mixing 0 if there is no mixing Example B 0 Hadronic decays Amplitude = 1 at m =0.5 ps -1 Amplitude = 1 at m =0.5 ps -1 Amplitude = 0 at m >> 0.5 ps -1 Amplitude = 0 at m >> 0.5 ps -1 Hadronic B 0 sample ~8K ev. Mix at m d Slide 52 September 8, 2005Donatella Lucchesi52 Amplitude Scan result No sensitivity (yet) but better behaved at high ms semileptonic Slide 53 September 8, 2005Donatella Lucchesi53 CDF + World Combined Result World Average + CDF Run II Sensitivity: 18.6 ps -1 Limit >14.5 ps -1 @ 95% CL CDFII combined result Sensitivity: 8.4 ps -1 Limit: m s > 7.9 ps -1 @ 95% CL Slide 54 September 8, 2005Donatella Lucchesi54 B s,d Lifetime differences: / Standard Model expectation: If bigger new physics Bs / Bs ~ (7-14)x10 -2 Indirect m s measurement (model dependent): Fermilab-Pub-01, 197 Off-shell transition contributes to m On-shell transition contributes to Slide 55 September 8, 2005Donatella Lucchesi55 B s,Light CP even, B s,Heavy CP odd Disentangle different L-components of decay amplitude isolate two B states B s,d / at Tevatron Examine two similar decay B s J/ B d J/ K *0 Look for evidence of two lifetimes in B decays = ( long + short )/2 = ( long - short ) = 1/ Definition: J/ KK K *0 K - + Total J=0 final state Two spin-1 J=0,1,2 Orbital L=0,1,2 3 Different decay amplitudes Slide 56 September 8, 2005Donatella Lucchesi56 B s,d / at Tevatron Transversity angle analysis Slide 57 September 8, 2005Donatella Lucchesi57 Decay Angular Distributions Slide 58 September 8, 2005Donatella Lucchesi58 Fitting functions: Fit simultaneously mass, lifetime and angular distribution convoluted with resolution function mB, L, H and A 0, A || A Slide 59 September 8, 2005Donatella Lucchesi59 Resuslts Slide 60 September 8, 2005Donatella Lucchesi60 B d Results A 0 = 0.7500.0170.012 A || =(0.473 0.034 0.006)e (2.860.220.07)i |A |= (0.4640.035 0.007)e (0.15 0.15 0.04)I Consistent with B factories results Slide 61 September 8, 2005Donatella Lucchesi61 B s Lifetime difference: unconstrained fit +69 s results m s =125 -55 Slide 62 September 8, 2005Donatella Lucchesi62 B s Lifetime difference: B s 0 J/ constrained fit Slide 63 September 8, 2005Donatella Lucchesi63 B s / Combined Fit results from ALEPH, without constraint with constraint CDF and DELPHI data s/s (95% CL range) [ +0.01 ; +0.59 ] [ +0.01 ; +0.59] s/s +0.35 +0.12 -0.16 +0.33 +0.09 -0.11 s +0.25 +0.09 -0.11 ps -1 +0.23 0.08 ps -1 1/s 1.42 +0.06 -0.07 ps 1.405 +0.043 -0.047 ps tau(short) = 1/L 1.21 +0.08 -0.09 ps 1.21 0.08 ps tau(long) = 1/H 1.72 0.19 ps 1.68 +0.08 -0.09 ps Slide 64 September 8, 2005Donatella Lucchesi64 Summary I hope I convinced you that B system is a nice and important laboratory where precisely test the Standard Model. For more tests and for probes for physics beyond Standard Model listen to the next talk Slide 65 September 8, 2005Donatella Lucchesi65 Backup Slide 66 September 8, 2005Donatella Lucchesi66 Future perspectives Add more channels B s D s 3 (130 events +20%) B s D s *p Add semileptonic B s decays from the hadronic trigger X2 semileptonic statistic Improve decay time resolution with PV event by event (detail)detail Incremental changes in existing algorithm (new Jet Charge +20% eD 2 ) Add new tagging algorithm Same Side Kaon Tag New data rolling in, but increasingly peak luminosity: Keep alive as much as possible present triggers SVT upgrade Use new trigger strategies 2 SVT Tracks + opposite side muon (pt>1.5 GeV) at trigger level (already in place since summer 2004 can survive at higher luminosity) Slide 67 September 8, 2005Donatella Lucchesi67 B s mixing sensitivity projection Analytic extrapolation, reproduce present result with current inputs Analytic extrapolation, reproduce present result with current inputs Prediction include a reduced (50%) effective luminosity usable for B-physics from 2007 onwards Prediction include a reduced (50%) effective luminosity usable for B-physics from 2007 onwards Sensitivity to the favorite CKM range Sensitivity to the favorite CKM range In case of no signal 95% C.L. up to 30 ps -1 with 4 fb -1 In case of no signal 95% C.L. up to 30 ps -1 with 4 fb -1 CKM fit will imply New Physics if m s >28 ps -1 by then CKM fit will imply New Physics if m s >28 ps -1 by then Slide 68 September 8, 2005Donatella Lucchesi68 Need high stat. sample to develop and calibrate tagging algorithm: Need high stat. sample to develop and calibrate tagging algorithm: High purity reached after lepton+track mass cut applied High purity reached after lepton+track mass cut applied Statistical Power of a tag: eD 2 Statistical Power of a tag: eD 2 Tagging efficiency (e) Tagging dilution (D = 1-2w) w = mistag rate w = mistag rate Parameterize dilution as a function of relevant variables and wheight events with their event-by-event dilution Parameterize dilution as a function of relevant variables and wheight events with their event-by-event dilution Dividing events into different classes based on tagging power improves combined eD 2 Dividing events into different classes based on tagging power improves combined eD 2 Calibration of the tagger performance requires high statistics! Calibration of the tagger performance requires high statistics! Calibration Sample for Taggers Use inclusive semileptonic decays from the lepton+track trigger (>10 6 events) Lepton charge gives true B flavour Tag the other b Slide 69 September 8, 2005Donatella Lucchesi69 B 0 s Lifetime: B s 0 J/ one component B 0 s Lifetime: B s 0 J/ one component +0.075 s / d =0.980 (stat.) 0.003 (syst.) -0.070 s / d =0.890 0.072 (tot.) L~240 pb -1 337 events D0CDF Ratio respect to B d 0 J/ K *0 Slide 70 September 8, 2005Donatella Lucchesi70 CDF/World Comparison Slide 71 September 8, 2005Donatella Lucchesi71 B d Lifetime difference B d Lifetime difference Fully reconstruct on Bboth in CP and flavor eigenstate decay modes Tag the other B Fit the proper time distribution with: explicitly appears in the hyperbolic term Slide 72 September 8, 2005Donatella Lucchesi72 Data sample Data sample B flavor 31027 events B 0 D *- + ( +,a 1 + ) B 0 D - + ( +,a 1 + ) B 0 J/ K *0 B CP 2603 events B 0 J/ K 0 s B 0 (2s)K 0 s B 0 J/ K 0 L B 0 c1 K 0 s Luminosity =82fb -1 Slide 73 September 8, 2005Donatella Lucchesi73 Results sgn[ ( CP )] / =-0.008 0.037 0.018 The 90% confidence level interval for / sgn[ ( CP )] / [-8.4%, 6.8%] -s* d / d = -0.009 +- 0.037 (BABAR and DELPHI)