High Precision Measurements of Bs Parameters in BsJ

1
High Precision Measurements of Bs Parameters in
Bs?J/??
Roger Jones University of Lancaster United
Kingdom for the ATLAS B-physics Group
Beauty 2005, Assisi, Italy
2
Overview of ATLAS
Weight 7000 tonnes
Length 46m
Muon chambers
Barrel toroid
Radius 11m
EM calorimeter
Forward calorimeter
End-cap toroid
Hadronic calorimeter
Inner detector
3
The LHC Environment

• pp collisions 14 TeV centre of mass energy
• Luminosity
• 2007 50 days _at_ 0.5 x 1033 cm-2s-1 Tuning
• 2008-2009 200 days _at_ 2 x 1033 cm-2s-1 Low
• 2010 1034 cm-2s-1 High
• Drops by factor 2 during 10 hour run
• 1 proton bunch crossing every 25ns
• 4.6/23 pp collisions/crossing _at_ low/high
  luminosity
• 1 of pp collisions produce a bb pair
• At luminosity 2 x 1033 cm-2s-1
• bb events produced with rate of 106 Hz
• 10Hz output to permanent storage for B-physics
• so highly selective and
  adaptable B-physics
• trigger required

4
B-physics trigger strategies (see Natalia
Panikashvili)
As luminosity drops during the fill, more
triggers are turned on
5
Bs-Bs mixing

• General box diagrams for Bs-Bs mixing

• Gs ½ ( GH GL )
• ?Gs GL GH

• ?Ms MH ML
• Xs ?Ms / Gs

• ?S mixing phase 2 sin ?c sin ? Vub / Vcb
• Arises through the interference of mixing and
  decay
• Highly sensitive to SUSY contributions
• Parameter is small in the Standard Model (0.02)
  so challenging measurement

6
BS?J/??
Extracting mixing parameters requires separation
of CP eigenstate amplitudes
1
Determined by the angular distribution of the
decay, and also proper times and tag
Scalar ? Vector Vector decay final state
described by three helicity amplitudes
2
4
Transversity basis linear combinations of
helicity amplitudes which are CP-eigenstates.
Complete determination yields mixing parameters
3
7
Decay parameterization

• 3 transversity amplitudes
• 2 independent magnitudes and 2 independent
  phases
• A A- d1 d2
• 3 mixing parameters, 1 weak phase
• Gs ?Gs ?Ms ?S
• ? 8 parameters to be extracted from the data

8
Theoretical distribution
h.c.
Accurately modelled by EvtGen
Distribution is model-independent new physics
enters through the modification of existing values
9
Is it a B or a B? (Tagging)
Jet charge tag
For Bd(s)? J/?(µ6µ3)K0S Tagging efficiency etag
0.64 (0.62) Wrong-tag fraction Wtag 0.42
(0.39)
b/b?
b/b?
Signal B/B-meson
etag(electron) 0.012 etag(muon)
0.025 Wtag(electron) 0.27 Wtag(muon) 0.24
b/b?
b/b?
Lepton tag from semi-leptonic decay
Signal B/B-meson
10
The workflow
1
EVENT GENERATION PythiaB, EvtGen
2
SIMULATION/DIGITIZATION/RECONSTRUCTION
3
AS REAL ANALYSIS

PARAMETER EXTRACTION maximum likelihood using
detector performance parameters derived from full
simulation.
4
11
Details of Analysis
BS?J/?(µµ)?(KK)

• All studies based on fully simulated ATLAS events
  and using the current reconstruction software
• 1 000 000 Bs decays produced with PythiaB and
  EvtGen to generate the correct angular
  distribution and mixing
• (model input A-,A, d1,d2,Gs,?Gs, ?s,?Ms)
• Cuts 1 muon gt 6 GeV 1 muon gt 3 GeV kaons gt 0.5
  GeV

All computations performed and all results
stored on the Grid (LCG)
12
As real Analysis

• Fit track pairs to J/? hypothesis
• pT(µ1) gt 3GeV pT(µ2) gt 3GeV
• ?(µ) lt 2.4
• ?2/DoF lt 6 m(J/?) ? (-3s,3s)
• s 38MeV

J/? invariant mass
? invariant mass

• Fit track pairs to ? hypothesis
• pT(K) gt 0.5GeV
• ?(K) lt 2.4
• ?2/DoF lt 6 m(?) ? (1009.2, 1029.6) GeV

13
As real Analysis

• BS fit
• Four-track fit to single vertex ?2/DoF lt 10
• must point at primary vertex
• Bs proper decay time gt 0.5ps
• pT(Bs) gt 10 GeV
• m(Bs) ? (-3s,3s) s 17MeV

Bs invariant mass
Bs proper decay time
14
Analysis Results and projections
Total number of signal events within kinematic
cuts after 30 fb-1 810 000
LVL1/LVL2 trigger di-muon efficiency 77
Number of signal events after trigger 623 700
15
Background analysis
Bd?J/?(µµ)K0(Kp-) (background 1) bb?J/?(µµ)X
(background 2)

• Background 1
• Identical spin structure and topology to signal
• S/B 15.1

• Background 2
• Angular structure assumed to be isotropic
• S/B 6.8

• Simulation/digitization/reconstruction
• Identical to signal
• Analysis
• Same analysis code run over background to
  calculate acceptance

16
Normalised Maximum Likelihood Estimator
Tagging efficiency. B-tag etag1 1 w etag2
w (Anti B)-tag etag1 w etag2 1 w No
tag etag1 etag2 0.5
Convolution with Gaussian to account for proper
decay time resolution
Reconstruction efficiency and acceptance
corrections determined from simulation
Background (level determined from simulation)
Theoretical PDF W for B0 at production W- for
anti-B0 at production
17
Maximum Likelihood test results

• XS fixed can be determined with Bs?Dsp
• Uncertainties on ?s are a function of Xs

Uncertainties
Correlations
s(?Bs)85fs ??s/?0.1
18
Conclusion Estimated reach of ATLAS

• No sensitivity to Standard Model values
• (nor have LHCb or CMS)
• CDF recently made a unexpectedly large
  measurement of ?Gs/Gs
• Study of this Golden Channel in ATLAS should
  provide a rich yield of interesting data