Transcript Slide 1

Observation of B0s – B0s Oscillations
The CDF Collaboration
Joseph Kroll
University of Pennsylvania
1st St. Ocean City, NJ, Feb. 7, 2003, H2O 350 F
DPF
Waikiki, HI
2 Nov 2006
Results presented today are contained in two papers:
Parallel session presentations: V. Tiwari (CMU) , J. Miles (MIT)
2 Nov 2006
J. Kroll (Penn)
2
Two-State Quantum Mechanical System
• Produce flavor states:
• Common decay modes ! 2-state QM system
M. Gell-Mann & A. Pais,
Phys. Rev., 97, 1387 (1955)
• States with mass & lifetime (neglecting CP violation)
“Light” (CP-even)
“Heavy” (CP-odd)
2 Nov 2006
J. Kroll (Penn)
3
Antiparticle
exists a time t!
Form asymmetry A(t) = cos(mst)
ms is oscillation frequency
2 Nov 2006
J. Kroll (Penn)
4
Measure Amplitude versus Oscillation Frequency
Time Domain
Frequency Domain
2G
Units: [m] = ~ ps-1. We use ~=1 and quote m in ps-1
To convert to eV multiply by 6.582£ 10-4
2 Nov 2006
J. Kroll (Penn)
5
Start 2006: Published Results on ms
 ms > 14.4 ps-1 95% CL
Results from LEP, SLD, CDF I
see http://www.slac.stanford.edu/xorg/hfag/osc/PDG_2006/index.html
2 Nov 2006
J. Kroll (Penn)
Amplitude method:
H-G. Moser, A. Roussarie,
NIM A384 p. 491 (1997)
6
April 2006: Result from the CDF Collaboration
V. M. Abulencia et al.,
Phys. Rev. Lett.
Vol. 97, 062003 (2006)
Probability
“Signal” is
random
fluctuation
is 0.2%
Under signal
hypothesis:
measure ms
2 Nov 2006
J. Kroll (Penn)
7
Since then CDF has focused on turning
evidence (3) into an observation (>5)
Tevatron has
delivered 2 fb-1
CDF has collected 1.6 fb-1
this analysis
Use the same 1 fb-1 data set with improved analysis
2 Nov 2006
J. Kroll (Penn)
8
Why is this Interesting?
Flavor oscillations occur through
2nd order weak interactions
e.g.
C. Gay, Annu. Rev. Nucl. Part. Sci. 50, 577 (2000)
From measurement of ms derive |V*tbVts|2
All factors known well except “bag factor” £ “decay constant”
Calculated on lattice, uncertainty ~ 10%
2 Nov 2006
J. Kroll (Penn)
9
B Meson Flavor Oscillations (cont)
Well measured: md = 0.507 § 0.005 ps-1 (1%) (PDG 2006)
Measure ms ! ms/md
Theoretical uncertainties reduced
Ratio measures |Vtd/Vts|
This is why ms is
high priority in Run II
from Lattice QCD calculations – see Okamoto, hep-lat/0510113
2 Nov 2006
J. Kroll (Penn)
10
Slide giving example of new physics
2 Nov 2006
J. Kroll (Penn)
11
Experimental Steps for Measuring Bs Mixing
1. Extract B0s signal – decay mode must identify b-flavor at decay (TTT)
Examples:
2. Measure decay time (t) in B rest frame (L = distance travelled) (L00)
3. Determine b-flavor at production “flavor tagging” (TOF)
“unmixed” means production and decay flavor are the same
“mixed” means flavor at production opposite flavor at decay
Flavor tag quantified by dilution D = 1 – 2w, w = mistag probability
2 Nov 2006
J. Kroll (Penn)
12
Schematic of Oscillation Event
opposite-side K
–
jet charge
2 Nov 2006
J. Kroll (Penn)
13
Key Experimental Issues
flavor tagging power,
background
mis-tag rate 40%
2 Nov 2006
displacement
resolution
momentum
resolution
(L) ~ 50 m
(p)/p = 5%
J. Kroll (Penn)
14
What’s Special About CDF & Tevatron
Tevatron delivered required luminosity
Unique trigger (SVT)
large sample of completely reconstructed Bs
Crucial for lifetime resolution & background reduction
Inner layer of silicon (L00)
provided decay distance resolution
Detector for particle identification (TOF)
made kaon identification possible
high efficiency, high purity flavor tag
2 Nov 2006
J. Kroll (Penn)
15
B0s Decay Modes
}
•Fully reconstructed (, 0)
better decay time resolution
•Lower statistics
•Signal 8,700
}
•Not fully reconstructed
poorer decay time resolution
•Higher statistics
•Signal 61,500
Hadronic
{
•
Semileptonic
{
•
Majority of signal collected with displaced track trigger
2 Nov 2006
J. Kroll (Penn)
16
Example: Fully Reconstructed Signal
Cleanest decay sequence
Four charged particles in
final state: K+ K- + Also use 6 body modes:
2 Nov 2006
J. Kroll (Penn)
17
Semileptonic Signals
2 Nov 2006
J. Kroll (Penn)
18
Proper Time & Lifetime Measurement
Decay position
production vertex
25m £ 25 m
Decay time in
B rest frame
2 Nov 2006
(B0s) = 1.??? § 0.0?? ps
(statistical error only)
PDG 2006: 1.466 § 0.059 ps
J. Kroll (Penn)
19
Decay Time Resolution: Hadronic Decays
Maximize sensitivity:
use candidate specific
decay time resolution
<t> = 86 £ 10-15 s
¼ period for
ms = 18 ps-1
Superior decay time
resolution gives CDF
sensitivity at much
larger values of ms
than previous experiments
Oscillation period for ms = 18 ps-1
2 Nov 2006
J. Kroll (Penn)
20
Semileptonics: Correction for Missing Momentum
Reconstructed quantity
2 Nov 2006
Correction Factor (MC)
J. Kroll (Penn)
Decay Time
21
Same Side Flavor Tags
Charge of K tags flavor
of Bs at production
Need particle id
TOF Critical
(dE/dx too)
Our most powerful flavor tag:
D2 = 4-5%
Opposite-side tags: D2 = 1.8%
2 Nov 2006
J. Kroll (Penn)
22
Results: Amplitude Scan
Sensitivity
31.3 ps-1
A/A = 6.1
Hadronic & semileptonic decays combined
2 Nov 2006
J. Kroll (Penn)
23
Measured Value of ms
- log(Likelihood)
2 Nov 2006
Hypothesis of A=1 compared to A=0
J. Kroll (Penn)
24
Significance: Probability of Fluctuation
Probability of
random fluctuation
determined from data
28 of 350 million
random trials
have L < -17.26
Probability =
8 £ 10-8 (5.4)
Have exceeded
standard threshold
to claim observation
-17.26
2 Nov 2006
J. Kroll (Penn)
25
Asymmetry (Oscillations) in Time Domain
2 Nov 2006
J. Kroll (Penn)
26
Determination of |Vtd/Vts|
D. Mohapatra et al.
(Belle Collaboration)
PRL 96 221601 (2006)
Previous best result:
CDF
2 Nov 2006
J. Kroll (Penn)
27
Summary of CDF Results on B0s Mixing
A. Abulencia et al., hep-ex/0609040, accepted by Phys. Rev. Lett.
Observation of Bs Oscillations and precise measurement of ms
Precision: 0.7% Probability of random fluctuation: 8£10-8
( 2.83 THz, 0.012 eV)
Most precise measurement of |Vtd/Vts|
2 Nov 2006
J. Kroll (Penn)
28
Backup & Alternate Slides
2 Nov 2006
J. Kroll (Penn)
29
Weakly Decaying Neutral Mesons
Flavor states (produced mainly by strong interaction at Tevatron)
2 Nov 2006
J. Kroll (Penn)
30
Key Features of CDF for B Physics
• “Deadtime-less” trigger system
– 3 level system with great flexibility
– First two levels have pipelines to reduce deadtime
– Silicon Vertex Tracker: trigger on displaced tracks at 2nd level
• Charged particle reconstruction – Drift Chamber and Silicon
– excellent momentum resolution: R = 1.4m, B = 1.4T
– lots of redundancy for pattern recognition in busy environment
– excellent impact parameter resolution (L00 at 1.5cm, 25m £ 25m
beam)
• Particle identification
– specific ionization in central drift chamber (dE/dx)
– Time of Flight measurement at R = 1.4 m
– electron & muon identification
2 Nov 2006
J. Kroll (Penn)
31
Example of Candidate
Zoom in on
collision pt.
candidate
Same-side Kaon tag
2 Nov 2006
Opposite-side Muon tag
J. Kroll (Penn)
32
Measuring Resolution in Data
Use large prompt D meson sample
CDF II, D. Acosta et al., PRL 91, 241804 (2003)
Real prompt D+
from interaction point
pair with random track
from interaction point
Compare reconstructed decay point to interaction point
2 Nov 2006
J. Kroll (Penn)
33
Time integrated oscillation probability
must measure proper time dependent oscillation to measure ms
2 Nov 2006
J. Kroll (Penn)
34