Performance of the CLEOIII RICH Detector Radia Sia

  • Slides: 25
Download presentation
Performance of the CLEO-III RICH Detector Radia Sia For the CLEO-III RICH Group Syracuse

Performance of the CLEO-III RICH Detector Radia Sia For the CLEO-III RICH Group Syracuse University RICH 2004 Outline: Ø The CLEO-III RICH Detector Ø Physics Requirements Ø CLEO-III RICH at work… Radia Sia RICH 2004 1 Syracuse Univ.

CLEO-III Detector Ø Designed to study the decays of b and c quarks, τ

CLEO-III Detector Ø Designed to study the decays of b and c quarks, τ leptons and Y mesons produced in e+e- collisions near 10 Ge. V cm energy. • Cs. I imposed hard outer radial limit. • Excellent charged particle tracking imposed a lower limit. • Low mass RICH: ~13% of X 0 at normal incidence • SILICON Detector Double sided, r and z strips Radia Sia RICH 2004 2 Syracuse Univ.

Li. F-TEA CLEO-III RICH Radia Sia RICH 2004 3 Syracuse Univ.

Li. F-TEA CLEO-III RICH Radia Sia RICH 2004 3 Syracuse Univ.

Li. F-TEA CLEO-III RICH Consists of 3 components: • radiators: Li. F radiator Methane-TEA

Li. F-TEA CLEO-III RICH Consists of 3 components: • radiators: Li. F radiator Methane-TEA MWPC • expansion volume: Use pure N 2 volume, 15. 6 cm thick to allow the Cherenkov angle to expand expansion gap (15. 7 cm) • photon detectors: Li. F radiator g Ca. F 2 windows K/p • MWPC: CH 4+TEA as VUV photosensor (135 -165 nm) • Charge induced on cathode pads, analog readout to measure photon position. • Ca. F 2 windows RICH 2004 g N 2 expansion gap Radia Sia 4 Syracuse Univ.

Crystal radiators v Array of planar Li. F crystals v Large refraction index (n=1.

Crystal radiators v Array of planar Li. F crystals v 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 Flat Radia Sia RICH 2004 5 Syracuse Univ.

Photon detectors Detection sequence: v g passes through thin UV transparent Ca. F 2

Photon detectors Detection sequence: v g passes through thin UV transparent Ca. F 2 window v Converted to single e by ionizing a TEA molecule. v single Photo-e drifts towards, then avalanches near the 20 um ø Au-W anode wires. v induces a charge signal on the cathode pads v Metallized traces at Ca. F 2 window to act as electrodes Radia Sia RICH 2004 6 Syracuse Univ.

Physics Requirements on CLEO III RICH Design • Require >3σ π/K separation at 2.

Physics Requirements on CLEO III RICH Design • Require >3σ π/K separation at 2. 8 Ge. V/c from RICH (we will also get ~2σ from d. E/dx above 2. 2 Ge. V/c to give ~4σ total PID) • At 2. 8 Ge. V/c: (θp-θK)=12. 8 mrad – need σθtrk = 4 mrad per track – use multiple p-e per track, so σθtrk = σθpe /(Npe)1/2 • Design benchmarks: ��� σθpe = 14 mrad, Npe = 12 pe/trk Radia Sia RICH 2004 7 Syracuse Univ.

Event Analysis • Look for photon hits within ± 3 s of the expected

Event Analysis • Look for photon hits within ± 3 s of the expected angle for each of 5 mass hypotheses (e, m, p, 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 p+, D 0 K-p+ • Combinatorial background subtracted by fitting D 0 mass peak Radia Sia RICH 2004 8 Syracuse Univ.

Single Photon Resolution (Bhabhas) Flat Sawtooth Radiators Flat Radiators Electrons P=5. 2 Ge. V

Single Photon Resolution (Bhabhas) Flat Sawtooth Radiators Flat Radiators Electrons P=5. 2 Ge. V s=12. 2 mrad s=14. 7 mrad Radia Sia RICH 2004 9 Syracuse Univ.

Single Photon Resolution (hadrons) Sawtooth Radiators Flat Radiators Background 12. 8% Background 8. 4%

Single Photon Resolution (hadrons) Sawtooth Radiators Flat Radiators Background 12. 8% Background 8. 4% s=13. 2 mrad s=15. 1 mrad Averaged over kaons with P>0. 7 Ge. V RICH 2004 10 Radia Syracuse Univ.

Photon Yield Sawtooth Radiators Flat Radiators Bhabhas) Mean after Background subtraction RICH 2004 11.

Photon Yield Sawtooth Radiators Flat Radiators Bhabhas) Mean after Background subtraction RICH 2004 11. 9 10. 6 Hadrons 11. 8 Hadrons 9. 6 11 Radia Syracuse Univ.

Per Track Resolution Sawtooth Radiators Flat Radiators Bhabhas s = 3. 6 mrad s

Per Track Resolution Sawtooth Radiators Flat Radiators Bhabhas s = 3. 6 mrad s = 4. 7 mrad s=3. 7 mrad s = 4. 9 mrad Radia Sia Hadrons RICH 2004 12 Syracuse Univ.

Resolution per track: decomposition Flat Sawtooth Measured Flat Expected Tracking error Chromatic error Photon

Resolution per track: decomposition Flat Sawtooth Measured Flat Expected Tracking error Chromatic error Photon position error Emission point error Radia Sia RICH 2004 13 Syracuse Univ.

PID Likelihoods • There are some optical path ambiguities for Cherenkov photon (from the

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 Poptical: which includes the refraction probability and the length of radiation path • Psignal 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 -2 ln(Lp/LK) is expected to have the same behavior as c 2 K – c 2 p Radia Sia RICH 2004 14 Syracuse Univ.

PID Likelihoods: example • Log-likelihoodratio distributions for 1. 0 -1. 5 Ge. V K’s

PID Likelihoods: example • Log-likelihoodratio distributions for 1. 0 -1. 5 Ge. V K’s and p’s identified via D*+a D 0 p+, D 0 a K-p+ K Radia Sia RICH 2004 15 Syracuse Univ.

Integral K/π Separation & Efficiency vs fakes study At 90% efficiency, fake rate is

Integral K/π Separation & Efficiency vs fakes study At 90% efficiency, fake rate is 7% Kaons Pions Radia Sia RICH 2004 16 Syracuse Univ.

PID performance (1) Pion Fake Rate Kaon efficiency = 0. 80 = 0. 85

PID performance (1) Pion Fake Rate Kaon efficiency = 0. 80 = 0. 85 = 0. 90 B physics CLEO-c Radia Sia RICH 2004 17 Syracuse Univ.

PID performance (2) Ds invariant mass from the Y(4 S)data Without PID # of

PID performance (2) Ds invariant mass from the Y(4 S)data Without PID # of Events / 2 Me. V This year, we used The Ds meson yield, reconstructed through the decay mode: Ds π, KK from the Y(5 S) and the Y(4 S) data collected with the CLEO III detector to measure the: Br(Y(5 S) Bs(*) ) which has never been measured before… D π Ds π With PID Ds invariant mass (Ge. V) Radia Sia RICH 2004 18 Syracuse Univ.

RICH at CLEO-c Era e+ e- (3770) Tagged Side D+ D- D 0 Through

RICH at CLEO-c Era e+ e- (3770) Tagged Side D+ D- D 0 Through hadronic channel D 0 -> Need particle ID to Signal Side, ex: keep background low Leptonic: D-> K - p+ RICH 2004 p+ m n DATA Semileptonic: Need electron ID, RICH is useful. . Radia Sia 19 Syracuse Univ.

Summary 1. CLEO Li. F-TEA RICH is providing us with excellent particle identification for

Summary 1. CLEO Li. F-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. 2. It was operating successfully for over 4 years (we made and are making extensive studies of the Upsilon, B and Bs decays). 3. 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 RICH 2004 20 Syracuse Univ.

Backup slides Radia Sia RICH 2004 21 Syracuse Univ.

Backup slides Radia Sia RICH 2004 21 Syracuse Univ.

Efficiency vs fakes study K m. D 0(Ge. V)/c 2 2 K 2 •

Efficiency vs fakes study K m. D 0(Ge. V)/c 2 2 K 2 • cut on 2 K 2 • fit m(D 0) RICH 2004 22 K 2 K 2 x Radia Syracuse Univ.

Efficiency vs fake rate For example at 85% efficiency, fake rate is ~13% Radia

Efficiency vs fake rate For example at 85% efficiency, fake rate is ~13% Radia Sia RICH 2004 23 Syracuse Univ.

Momentum dependence line solid dashed dotted red 0. 7 -0. 9 -1. 1 -1.

Momentum dependence line solid dashed dotted red 0. 7 -0. 9 -1. 1 -1. 3 blue 1. 3 1. 5 1. 7 1. 9 line solid dashed dotted green 1. 9 2. 1 2. 3 2. 5 black 2. 5 2. 7 2. 9 3. 1 Radia Sia RICH 2004 24 Syracuse Univ.

Momentum dependence Momenta 0. 6 -1. 1 Ge. V(red) 2. 5 -2. 7 Ge.

Momentum dependence Momenta 0. 6 -1. 1 Ge. V(red) 2. 5 -2. 7 Ge. V(blue) Radia Sia RICH 2004 25 Syracuse Univ.