Prompt χ c1 , χ c2 and X(3872) Production in e + e - annihilation arXiv:0707.1633 (submitted to PRD RC) Woochun Park University of South Carolina August 17, 2007 @Lepton-Photon 2007 8/16/2007 1
Introduction A lot of exciting news about charmonium production in e + e - annihilations • so far from B-factories: – Prompt J/ ψ , ψ (2S) and double charmonium productions are observed. – X(3872), Z(3930), Y(3940), Z(3940), and Y(4260)……… – These are great tools to test NRQCD. • Why have χ c1 and χ c2 not been observed yet in the continuum? – In B decay, inclusive BF is 1.09% (J/ ψ ), 0.31% ( ψ (2S)), 0.39%( χ c1 ), 0.14%( χ c2 ). In e + e - annihilation, J/ ψ, ψ (2S), χ c0 found but not χ c1 , χ c2 yet. – Search for χ c1 and χ c2 with the dominant BF process γ J/ ψ ( 36% and 20%) . – • Test color octet mechanism – Key to explain J/ ψ enhancement. More prominent in χ c . – Estimated cross section is 85 fb -1 and 123 fb -1 for χ c1 and χ c2 . G. A. Schuler and M. V ¨ anttinen, Phys. Rev. D 58 017502 (1998) • Search for X(3872) � γ J/ ψ in continuum X(3872) � γ J/ ψ observed in B-decay . Belle PRL 2002 (33fb -1 ) – Take advantage of the machinery for χ c . – 8/16/2007 2
MC Samples • Double-charmonium MC ( γ * → χ c J/ ψ or ψ (2S) ) is used as our signal MC. • To estimate efficiency inclusively, we use single χ c and X(3872) MC generated flat distribution over cos θ * and p* up to 5 GeV/c. • The χ c from B decays are used as a control sample to validate data-to-MC comparison. • BB MC is used to estimate BB background. • For the X(3872) search, we take advantage of the χ c study because of similar kinematics. 8/16/2007 3
Event Selection • N ch > 4 and R 2 < 0.8 to suppress QED type of backgrounds : ISR ψ (2S) and two-photon fusion events. • Qualified photon candidate must satisfy: – A 42 < 0.1 : A 42 measures the azimuthal asymmetry of the cluster about its peak, distinguishing electromagnetic from hadronic showers. – 0.01 < LAT < 0.5: LAT is a measure of the radial energy profile of the cluster and is used to suppress clusters from electronic noise or hadronic interactions. – 0.41 < θ < 2.41 : Photons in electromagnetic calorimeter fiducial volume – Reject γ from π 0 if M π 0 ∈ [0.114, 0.146] GeV with E γ = 30MeV, Lat < 0.8 – Splitoff rejection by requiring at least 9 ° from any charged track. • p*( χ c , J/ ψ ) > 2.0 GeV/c to suppress B-decay contribution. • Electron id with radiation recovery. Muon identification. • Geometric constraint on the J/ ψ vertex. • -0.05 < M ee – M J/ ψ < 0.03 GeV and -0.03 < Μ µµ – M J/ ψ < 0.03 GeV. • |cos θ H (J/ ψ )| < 0.9 • 0.25 < Δ M (M γℓℓ – M ℓℓ ) < 0.60 GeV for the χ c search • 0.60 < Δ M (M γℓℓ – M ℓℓ ) < 0.95 GeV for the X(3872) search 8/16/2007 4
Event Selection (III) cos θ H ( χ c ) < 0.40 – The figure of merit is N sig 2 /(N cont +N BB ) for the individual cut. – The optimized cut is not sensitive to the scale of N sig . 8/16/2007 5
Unbinned ML Fit CBL Parameter values from MC. • Signal PDF – Crystal Ball Line shape (CBL). – Mass difference between χ c is constrained to PDG 2006 value, 45.5 MeV. – To account for possible energy scale or resolution difference between data and MC, mean is shifted by an offset and resolution is scaled by scale factor. It’s tuned by our control sample (see the next page). • Background PDF 3 rd order Chebyshev polynomial with all parameters floated. – – For the purpose of systematic error study, Exponential function is used. 8/16/2007 6
Control Sample UML fit for the control sample ( χ c from B decays: p*< 1.7 GeV/c) χ c1 χ c2 • To tune offset and scale, we performed an UML fit for the control sample, χ c from B decays (p* < 1.7 GeV/c) . • To cross check, N( χ c2 ) to N( χ c1 ) ratio is calculated and it is consistent with PDG2006. • These values ( δ and β ) will be used and fixed in the UML fit to search for χ c and X(3872) in continuum. 8/16/2007 7
Efficiency (we assume f Ntrk = 1.0) 8/16/2007 8
Reconstruction Efficiency ( ε r ) • The ε r depends on p* and cos θ * of χ c because of p*(J/ ψ ) > 2.0 GeV/c and lower coverage in endcap. • We need to correct the single particle MC ε r using the weight matrix of p* and cos θ * and an efficiency matrix in bins of p* and cos θ *. i =1, 6 j =1, 5 8/16/2007 9
π 0 veto efficiency ( ε v ) • The ε v is dependent on the number of photons in the event. • We need: – The efficiency as a function of photon multiplicity ε v (N γ ) – N γ distribution of signal events. • The ε v is the weighted average of ε v (N γ ) [weighted by the fraction of N γ ] . • Corrected by data-to-MC scale from the control sample (~ 1%). • An alternative N γ distribution (without sideband distribution) is used to estimate systematics. 8/16/2007 10
Splitoff Rejection efficiency ( ε s ) • The ε s is dependent on the N ch in the event. • We need: – The efficiency as a function of charged track multiplicity ε s (N ch ) – N ch distribution of signal events. • The ε s is the weighted average of ε s (N ch ) [weighted by the fraction of N ch ]. • Corrected by data-to-MC scale from the control sample (~ 0.4%). • An alternative N ch distribution (w/o sideband distribution) is used to estimate systematics. 8/16/2007 11
Systematic Error Study Uncertainty from p*/cos θ * distribution Deviation from alternative distribution of N γ / N ch or ε v (N γ ) / ε s (N ch ). , uncertainty of δ and β ISR ψ (2S) background is estimated from MC as 9.4 evnt for the χ c1 and 5.1 evnt for the χ c2 . PDG 2006 8/16/2007 12
Background – Two-Photon Background is estimated to be negligible. – ISR ψ (2S) contribution is estimated using MC sample and subtracted from signal yield. • ISR ψ (2S) background is estimated from MC as 9.4 evnt for the χ c1 and 5.1 evnt for the χ c2 . – Prompt ψ (2S) contribution is estimated using PDG reference value and subtracted from cross section. – Combinatorial backgrounds are modeled by Chebyshev polynomial and systematic error is assigned using alternative pdf. 8/16/2007 13
Yields UML fit for the continuum data (p* > 2.0 GeV/c) • Offset and scale are fixed to the values from the control sample. • For the χ c search: • For the X(3872) search: 8/16/2007 14
Result For the X(3872), we assume BF of X(3872) � γ J/ ψ is 100%. Estimated ISR ψ (2S) contribution is subtracted. 8/16/2007 15
8/16/2007 16
After subtraction of prompt ψ (2S) contribution, prompt χ c production cross-sections in continuum are: y r a n i m i l e Conclusion r P r a B a B While χ c production has been observed in e + e - annihilation ~ 10.6 GeV, the measured cross-sections are compatible with the expected contributions from prompt ψ (2S) production feed-down to χ c. No evidence of prompt χ c production. Theoretical predictions are order of the following values σ (e + e - � χ c1 X)=85 fb and σ (e + e - � χ c2 X)=123 fb (PRD58, 017502(1998)) Prompt X(3872) production in continuum is : No evidence of prompt X(3872) in e + e - annihilation. arXiv:0707.1633 (submitted to PRD RC) 8/16/2007 17
Recommend
More recommend
