HCP 2005 Particle Identification of the LHCb detector Ann.Van.Lysebetten@cern.ch on behalf of the LHCb collaboration CERN 5th July 2005
The LHCb experiment : introduction precision measurements of CP violation and rare phenomena in B meson decays Standard Model consistency?, new physics? Experimental requirements: Efficient trigger (see talk by Mitesh Patel) Vertex identification High track reconstruction efficiency pp interaction (primary vertex) b-hadron B 0 L l π + π Κ π + π + Control of systematics Particle identification 1. π/k separation RICH 2. lepton id muon + calo
RICH detectors for hadron identification RICH1 Z ~ 1.0 2.2 m aerogel:2 ~10 GeV/c C 4 F 10 :10 ~60 GeV/c RICH2 Z ~ 9.5 11.9 m CF 4 :16 ~100GeV/c pp interaction point Magnet efficient pion/kaon separation up to ~100 Gev/c
Lepton identification : muons and electrons B J/ψ K s access to angle β Rare decays B μμ e h HCAL μfe + scintillator tiles 1468 cells, MWPC s 5.6 λ Ι σ/e ~ 75% / E + GEMs Info used in 1 st level trigger ECAL Shashlik type 66layers 2mm Pb/4mm scintillator 25 X o σ/e ~ 10%/ E
RICH detector requirements Goal: π/k separation in ~ 1-100GeV/c range 400 Polar angle versus momentum Number of tracks 300 (a) 200 100 0 0 50 100 150 200 80 60 40 (b) B ππ decay tagging kaons θ [mrad] 300 200 100 RICH1 RICH-1 25-300 mrad RICH2 B d0 ππ RICH-2 15-120 mrad 20 0 0 5 10 15 20 Momentum (GeV/c) 0 0 50 100 150 200 Momentum [GeV/c ]
RICH detector specifications RICH system divided into 2 detectors with 3 radiators θ C (mrad) 250 200 150 100 50 0 e μ π K π p Aerogel Aerogel C 4 F 10 gas CF 4 gas 1 10 100 Momentum (GeV/c) K C4F10 CF4 θ C max 242 mrad R I C H 1 RICH2 53 mrad 32 mrad
RICH1: low momentum tracks Magnetic shielding Photon Detectors Aerogel Thickness 5 cm n = 1.030 ± 0.001 Aerogel C 4 F 10 250 mrad Spherical Mirror Beam pipe Aerogel Support Structure VELO exit window Track Be Mirrors y Plane Mirror 0 100 200 z (cm)
RICH2: high momentum tracks (CF4) x z Photon Detectors plane
56 spherical mirrors on two planes -> installed and aligned with precision σθ 50 μrad
Photon Detectors Requirements: 2.5x2.5 mm 2 granularity => σ(θ CH ) = 0.6 mrad (RICH1) Coverage of 2.6 m 2 Single photon sensitive λ 200 600 nm Read-out speed 40 MHz B B Hybrid Photon Detector (HPD) S20 multi-alkali photocathode Cross-focusing electron optics (~20kV, demagnification factor of 5) Photo-electrons focused on anode assembly (16x16 mm2), fully encapsulated in the vacuum enclosure 83mm Mu-metal shielding 120mm 140mm
Mumetal shield 20G Simulated B field distribution in RICH1 B B 0G B 30 G Calibration pattern 0 G B 50 G B field in shielding boxes measured! Most important effects from axial fields RICH1 Effect = PARAMETRISABLE in situ calibration pattern runs, Study effect in physics data
RICH simulation and pattern recognition Simulation in Geant4 RICH1 RICH1 B->Kpi events RICH2 get rings out multitude of hits!
RICH simulation and pattern recognition RICH1 RICH2 Aerogel C 4 F 10 CF 4 L (cm) 5 85 167 n 1.03 1.0014 1.0005 N pe 7 31 23 σ totθ (mrad) 2.19 1.29 0.60 RICH Pattern recognition approaches: Track based (local/global) Other algorithms (ring fitting) Difference in log-likelihood under different hypotheses significance of π/k separation for true pions Chromatic 2.07 0.80 0.47 Emission Point 0.34 0.80 0.33 Pixel Size 0.57 0.57 0.16 σ totθ (mrad) RICH +Tracks 2.60 1.60 0.61
RICH hadron ID performance Heavy PID Light PID Kaon identification eff ~88% K K,p Kaon misidentification ~12% K e,μ,π pion identification ~ 97% π e,μ,π pion misidentification ~ 3% π K,p
RICH hadron ID performance: B s -> D s K Time dependant asymmetries in B s -> D s K with CP asymmetry in B 0 S -> J/ψ φ γ with σ(γ) = 15 0 per year Very little theoretical uncertainties Insensitive to new physics ~ 5400 evts per year expected BUT! Entries 200 175 150 125 2000 1500 m Bs = 5.37 GeV/c 2 σ Bs = 13.8 MeV/c 2 D s K D s π m Bs = 5.42 GeV/c 2 σ Bs = 24.0 MeV/c 2 100 75 Bs Ds π Bs Ds K 1000 50 500 25 0 40 20 0 20 40 ΔlnL Kπ 5.3 5.35 5.4 5.45 5.5 mass [GeV/c 0 B s contamination ~ 90% 2 ]
RICH hadron ID performance: B (s) -> h + h - B-> ππ α measurement precision 2-5 0 per year Combination of B-> ππ and B -> KK -> γ measurement precision 5 0 per year -> sensitive to new physics Yields: 26k B 0 π+π and 37k B s -> KK B 0 π + π - Tree diagram B 0 (B 0 ) s b B s K + K - d(s) W Penguin diagram (example) b d(s) W t c u g d(s) u u d(s) d(s) u u d(s) π + (K + ) π (K ) π + (K + ) π (K ) With RICH Purity=84%, ε=79% No RICH Purity=13%
Muon identification Search of hits in muon stations compatible with reconstructed track extrapolations B J/ψ K s ; p > 3 GeV/c μ eff 96.5% π eff 2.1% σ Μ 10 MeV Further reduction of bkg with other algorithms and multivariable analysis =>μ eff ~ 90%, π eff 0.8% Provide global PID with DLL
Electron identification Difference in log likelihood for (DLL): signal (e) and bkg hypothesis based on 4 variables Corrected cluster energy position matching e+/e h/μ Energy deposition in preshower χ 2 e e+/eh/μ Bremsstrahlung correction e+/eh/μ e+/eh/μ E PS (MeV) Energy deposition along extrapolated track in HCAL χ 2 brem E HCAL (GeV)
Electron identification performance B 0 -> J/ψ K 0 S J/ψ -> e+ e - within calorimeter acceptance: Eff 95% Eff 81% π misid 0.7% Combined with RICH B S -> J/ψ φ J/ψ -> e + e - Tail from Bremsstrahlung Bkg rejected with p T cut, mainly coming from secondary electrons and ghosts Performance e eff 78%, π misid 1%
Conclusions Particle identification is essential for the LHCb physics program! Hadron identifcation 3σ π-k separation from 1-100 GeV/c provided by 2 RICH detectors RICH2 construction almost complete, RICH1 construction underway photon detector understood and being produced Installation expected to be completed by October 2006 Electron identification ECAL installed, HCAL underway Muon identification Muon project well advanced (480 chambers produced = 1/3 of the total) Combined PID improves performance
Dipole Magnet Installation progress in point 8 ECAL ½ HCAL Muon filters RICH1 magnetic shielding Eager for first collisions in 2007!