Search for the decay BcgBsp

1
Search for the decay BcgBsp

• Masato Aoki
• September 13, 2006

1. BsgJ/yf channel
2. Pion reconstruction efficiency
3. Cut variables and optimization
4. First check BsgDsp,Dsgfp data

2
Pion(from Bc) reconstruction vertexing
efficiency

• BcgBsp, BsgJ/yf Monte Carlo 10,000,000 events
  generated
• Reconstruct only Bs meson
• 54487 events
• Reconstruct pion as well
• VtxFit Prob(c2r-f)gt0.1, SiCOT hits
• 31819 events
• ? 58 (just reconstructed 39613 events 73)

3
Pre-selection

• dataset xpmm0d, 0h and 0i with v13 goodrun list
  (1.1fb-1)
• pT gt 0.35 GeV for all tracks
• Number of Silicon r-f hit layers gt 3 for all
  tracks
• Number of COT axial hit layers (with minimum 5
  hits) gt3 for m and p tracks
• Muons
• pT gt 1.5 GeV, c2stub lt 9 for CMU muons
• pT gt 2 GeV, c2stub lt 9 for CMX muons
• Df(m-m) lt 120degree
• Mass window
• f mass window 10 MeV around PDG f mass
• J/y mass window 50 MeV around PDG J/y mass
  (mass constraint)
• Bs mass window 30 MeV around 6.33601 GeV (mass
  constraint)
• Vertex
• Prob(c2r-f) gt 0.1 for the Bs and Bc vertex
  fitting

4
Variables to be optimized

• pT(Bc), pT(Bs), pT(J/y), pT(f), pT(p)
• Impact parameter of Bc dxy
• DR(Bs,p)
• ct(Bc), ct(BsgBc), ct(BsgPV)
• Signal Monte Carlo (pre-selection 11 events
  expected)
• Background data (sideband events)
• Maximize S / (1.5 Sqrt(B))
• N-1 iteration method
• Cleanup 30 MeV Bsp mass window (corresponding to
  3s)

5
GeV/c
6
cm
7
cm
8
GeV/c
9
cm
10
cm
11
GeV/c
12
GeV/c
13
GeV/c
14
(No Transcript)
15
Optimized cut variables

• pT(Bc) gt 4.15 GeV
• pT(Bs) gt 0 GeV, pT(J/y) gt 0 GeV
• pT(f) gt 1.49 GeV
• pT(p) gt 1.25 GeV
• dxy(Bc) lt 77 mm
• DR(Bs,p) lt 0.801
• 50 mm lt ct(Bc) lt 1740 mm
• -550 mm lt ct(BsgBc) lt 4060 mm
• 200 mm lt ct(BsgPV) lt 3620 mm
• Expected events in 30 MeV Bsp mass window
• ? expected S/N 2.37 / 1.17

16
Blind Box 6.2765 3?0.005(uncertainty)0.005(r
esolution)
17
BsgDsp, Dsgfp

• stripped dataset chdl07, skbh01 (no one uses
  those datasets)
• Obtained 1000 Bs, this should be 1500(mixing
  analysis)
• Whats wrong? Bug? Trigger? Selection? Goodrun?
  Fit modeling?

18
Summary

• Cut optimization was performed
• Expected events S/N 2.37 / 1.17
• Are there much better variables?
• Likelihood? Neural Network?
• Looked at BsgDsp,Dsgfp stripped dataset
• A large disagreement was observed