Radia Sia Syracuse Univ. 1 RICH 2004 Outline: The CLEO-III RICH Detector Physics Requirements CLEO-III RICH at work Performance of the CLEO-III RICH Detector Syracuse University RICH 2004 Radia Sia For the CLEO-III RICH Group Radia Sia Syracuse Univ. 2 RICH 2004 CLEO-III Detector SILICON Detector Double sided, r and z strips CsI imposed hard outer radial limit. Excellent charged particle tracking imposed a lower limit. Low mass RICH: ~13% of X 0 at normal incidence Designed to study the decays of b and c quarks, leptons and Y mesons produced in e + e - collisions near 10 GeV cm energy. Radia Sia Syracuse Univ. 3 RICH 2004 LiF-TEA CLEO-III RICH Radia Sia Syracuse Univ. 4 RICH 2004 LiF-TEA CLEO-III RICH LiF radiator K/ N2 expansion gap CaF2 windows Methane-TEA MWPC Consists of 3 components: radiators: LiF radiator expansion volume: Use pure N2 volume, 15.6cm thick to allow the Cherenkov angle to expand expansion gap (15.7cm) photon detectors: MWPC: CH4+TEA as VUV photosensor ( nm) Charge induced on cathode pads, analog readout to measure photon position. CaF2 windows Radia Sia Syracuse Univ. 5 RICH 2004 Crystal radiators Array of planar LiF crystals Large refraction index (n=1.5 at 150 nm) ---> number of photons are internally reflected ---> sawtooth radiators implemented in the 4 central raws Flat Sawtooth Radia Sia Syracuse Univ. 6 RICH 2004 Photon detectors Detection sequence: passes through thin UV transparent CaF2 window Converted to single e by ionizing a TEA molecule. single Photo-e drifts towards, then avalanches near the 20 um Au-W anode wires. induces a charge signal on the cathode pads Metallized traces at CaF2 window to act as electrodes Radia Sia Syracuse Univ. 7 RICH 2004 Physics Requirements on CLEO III RICH Design Require >3 /K separation at 2.8 GeV/c from RICH (we will also get ~2 from dE/dx above 2.2 GeV/c to give ~4 total PID) At 2.8 GeV/c: ( p - K )=12.8 mrad need trk = 4 mrad per track use multiple p-e per track, so trk = pe /(N pe ) Design benchmarks: pe = 14 mrad, N pe = 12 pe/trk 1/2 Radia Sia Syracuse Univ. 8 RICH 2004 Event Analysis Look for photon hits within 3 of the expected angle for each of 5 mass hypotheses (e, , , K, p) for every track For each track find the most likely mass hypothesis Do not allow photons which belong to the most likely mass hypothesis to be used by any other track Recalculate PID quantities Performance plots for hadrons are obtained on kaons identified kinematically via D* + D 0 +, D 0 K - + Combinatorial background subtracted by fitting D 0 mass peak Radia Sia Syracuse Univ. 9 RICH 2004 SawtoothFlat Single Photon Resolution (Bhabhas) mrad Flat mrad Electrons P=5.2 GeV Flat Radiators Sawtooth Radiators Radia Sia Syracuse Univ. 10 RICH 2004 Single Photon Resolution (hadrons) mrad mrad Background 12.8% Background 8.4% Averaged over kaons with P>0.7 GeV Flat RadiatorsSawtooth Radiators Radia Sia Syracuse Univ. 11 RICH 2004 Photon Yield Flat RadiatorsSawtooth Radiators Bhabhas ) Hadrons Mean after Background subtraction Mean after Background subtraction Radia Sia Syracuse Univ. 12 RICH 2004 Per Track Resolution mrad mrad Flat RadiatorsSawtooth Radiators Bhabhas mrad mrad Hadrons Radia Sia Syracuse Univ. 13 RICH 2004 Resolution per track: decomposition Measured Expected Sawtooth Flat Chromatic error Emission point error Photon position error Tracking error Radia Sia Syracuse Univ. 14 RICH 2004 PID Likelihoods There are some optical path ambiguities for Cherenkov photon (from the sawtooth radiator and from a track passing two radiators) we chose the closest to the expected one when calculating per track Cherenkov angle There is some loss of information in this approach. use likelihood method to perform PIdentification. The likelihood method weights each possible path by its optical probability P optical : which includes the refraction probability and the length of radiation path P signal is a Gaussian like function with asymmetric high tail (CBL) as seen before. The likelihood folds in Cherenkov photon yield with Cerenkov angle measurements for each hypothesis The distribution of -2ln(L p /L K ) is expected to have the same behavior as 2 K 2 p Radia Sia Syracuse Univ. 15 RICH 2004 PID Likelihoods: example K Log-likelihood- ratio distributions for GeV Ks and s identified via D* + D 0 +, D 0 K Radia Sia Syracuse Univ. 16 RICH 2004 Integral K/ Separation & Efficiency vs fakes study Pions Kaons At 90% efficiency, fake rate is 7% Radia Sia Syracuse Univ. 17 RICH 2004 PID performance (1) CLEO-c B physics Kaon efficiency = 0.80 = 0.85 = 0.90 Pion Fake Rate Radia Sia Syracuse Univ. 18 RICH 2004 PID performance (2) D s invariant mass from the Y(4S)data With PID Without PID This year, we used The D s meson yield, reconstructed through the decay mode: D s , KK from the Y(5S) and the Y(4S) data collected with the CLEO III detector to measure the: Br(Y(5S) B s (*) B s (*) ) which has never been measured before D s D D s invariant mass (GeV) # of Events / 2MeV Radia Sia Syracuse Univ. 19 RICH 2004 RICH at CLEO-c Era e+ e- (3770) D + D - D 0 Tagged Side Signal Side, ex: Leptonic: D-> m n Semileptonic: Need electron ID, RICH is useful.. K - p + p + DATA Through hadronic channel -> Need particle ID to keep background low Radia Sia Syracuse Univ. 20 RICH 2004 Summary CLEO LiF-TEA RICH is providing us with excellent particle identification for all momenta relevant to the past beauty threshold data and present charm threshold CLEO c data. It was operating successfully for over 4 years (we made and are making extensive studies of the Upsilon, B and B s decays). Since last year, we used the detector for the CLEO-c program which is providing us with the Charm factory data to study decays of charm mesons and charmonium decays. The results went already for publication Radia Sia Syracuse Univ. 21 RICH 2004 Backup slides Radia Sia Syracuse Univ. 22 RICH 2004 Efficiency vs fakes study cut on 2 K 2 fit m(D 0 ) 2 K 2 x 2 K 2 m D0 (GeV)/c 2 K K Radia Sia Syracuse Univ. 23 RICH 2004 Efficiency vs fake rate For example at 85% efficiency, fake rate is ~13% Radia Sia Syracuse Univ. 24 RICH 2004 Momentum dependence line solid dashed dotted red blue line solid dashed dotted green black Radia Sia Syracuse Univ. 25 RICH 2004 Momentum dependence Momenta GeV(red) GeV(blue)