The Elephant in the B Factory
An Asymmetric Fable
E.I. Rosenberg
September 29, 2000
Outline
•
•
•
•
•
Review of CP Violation
SM Constraints
The B-factory
The BaBar Detector
Preliminary results (∼ 9 fb-1)
• Status and outlook
E. I. Rosenberg
September 29, 2000
2
The Discrete Symmetries
Parity(P) – inversion of space
+
+
+
+
Γ( K → µ ν ) = Γ( K → µ ν )
L
µL
R
µR
Charge Conjugation(C) –
interchange of particle and
antiparticle (inversion of strong
charges)
+
+
Γ( K → µ ν ) = Γ( K − → µ −ν )
L
µL
Time Reversal (T) – inversion of time
+
+
+
L
µL
+
Γ( K → µ ν ) = Γ( µ ν → K )
L
E. I. Rosenberg
µL
L
µL
September 29, 2000
3
CP and CPT Symmetries
•P, C and T are all conserved by the strong and
electromagnetic interactions.
•The weak interaction violates P and C, but conserves
the combination CP (almost).
+
+
−
−
Γ( K → µ ν ) = Γ( K → µ ν )
L
µL
R
µR
• The combination CPT is always conserved.
•⇒ CPT conservation requires:
mK = mK
+
−
Γ (K ) = Γ (K )
total
total
0
0
Γ (K ) = Γ (K )
total
total
E. I. Rosenberg
m -m
m
K0
September 29, 2000
average
K0
< 10−18
4
Types of CP Violation
• Direct CP Violation:
f HM ≠ f HM
(A ≠ A )
f
f
search with neutral or charged states
•CP Violation in mixing: the neutral states mix -- mass
(decaying) and strong interaction eigenstates are different.
ML = p M
0
+q M
0
; M = p M 0 -q M 0
H
•If |q|≠ |p|, the mass eigenstates are not CP eigenstates
•CP Violation in the interference
between mixing and decay:
q A fcp
λ=
;
p A fcp
λ ≠1
λf = 1; Im λf ≠ 0
E. I. Rosenberg
September 29, 2000
5
Experimental Observation of CP Violation
•1964 First observation:
Γ(K → ππ) ≠ 0
0
L
Only observed for neutral kaons and described by three
measurements all of which involve the quark transition s → u :
0
q 1−ε
A(K L
→ π +π − )
=
η +− =
= ε + ε′
p 1+ ε
A(K s0 → π +π − )
0
-3
A(K L
→ π 0π 0 )
η
≈
η
≈
2.3x10
η00 =
= ε − 2ε ′
+−
00
0
0 0
A(K s → π π )
δ=
Γ(K 0L → π - l +ν ) − Γ(K 0L → π + l −ν )
Γ(K 0L → π - l +ν ) + Γ(K 0L → π + l −ν )
E. I. Rosenberg
= (0.327 ± 0.012)%
September 29, 2000
6
CP Violation and the Matter Universe
• At early times, nB = nB
• At some stage in the evolution: CP violating and
baryon number changing processes can occur so
( n − n ) /( n + n ) ≠ 0
B
B
B
B
• universe is temporarily out of equilibrium
• a short time late B violating processes are no
longer active
• B + B → nγ converts all the antibaryons and all
but the excess baryons to photons so
n −n
n
→
n +n
n
E. I. Rosenberg
B
B
B
B
B
γ
September 29, 2000
7
Standard Model and CP (I)
 u  c  t 
 d  s  b 
   
•The flavor eigenstates of
the quarks are not the weak
interaction eigenstates--there are transtions
between the families.
This mixing is described by
the CKM Matrix. Where the
three diagonal elements
Vud, Vcs, and Vtb ≈ 1, family
is almost a good quantum
number.
E. I. Rosenberg
 Vud Vus Vub 
V V V 
 cd cs cb 
V

V
V
 td
ts tb 
September 29, 2000
8
Standard Model and CP (II)
•The CKM Matrix is unitary and in the case of three
generations can be parameterized by three Euler angles
and six phases. Five of the phases can be absorbed into
the quark phases. One observable phase remains.
•THE IMAGINARY PART OF THE CKM MATRIX IS
NECESSARY TO DESCRIBE CP VIOLATION
•All CP violating amplitudes in the SM are proportional to
JCP = |Im (VijVklVil*Vkj*)|; i≠k, j≠l

λ2
1
Wolfenstein
2

parameterization 
-λ


 A λ 3 (1 -ρ -iη )



E. I. Rosenberg
λ
λ2
12
-A λ 2
September 29, 2000

A λ (ρ -iη ) 


2
Aλ



1



3
+ 0(λ4)
9
Standard Model and CP (III)
•In the Wolfenstein parameterization, the CP violating
amplitude JCP ∝ A2λ6η ≈ O(10-4)
Meson
Dominant
quark decays
Resonance
width
CP violating
effect
K
D
B
s→u
c→s
b→c
∝ λ2 x p.s
∝ 1 x p.s
∝ A2λ4 x p.s
∝ A2λ4η
∝ A2λ6η
∝ λ2η
A, λ ,η are all < 1; λ = 0.02205 ± 0.0018
THE LARGEST EFFECTS ARE IN THE B SYSTEM
E. I. Rosenberg
September 29, 2000
10
Standard Model and CP (IV)
•The CKM matrix represents rotations between the
quark generations and so is unitary: VV = 1
•There are six equations of the form VijVik* = 0 (j≠k) which can be
represented as a closed triangle in the complex plane -- all of which
have area ½JCP.
•The most useful contains the most poorly known elements of the
CKM matrix: VudVub*+VcdVcb*+VtdVtb*=0
Vud Vub*
( ρ ,η )
V td V tb*
Vcd Vcb*
α
V cd V cb*
γ
E. I. Rosenberg
β
September 29, 2000
(1,0)
11
Physics Goals: Precision Measurement of the sides
and angles of the CKM Unitarity Triangle to
overconstrain the SM and look for new physics
At an asymmetric collider, such
as PEP-II and KEK, we can
measure the time-dependent
CP-violating asymmetries in the
decay of neutral B-mesons.
sin 2β: B0 → J/ψKs0, B0 → J/ψKL0, B0 → J/ψK*(892)
B0 → D+D-, B0 → D*+D-, B0 → D*+D*sin 2α: B0 → π+π-, B0 → ρπ, B0 → ρρ, B0 → a1π
(difficult due to penguin contributions)
|Vcb | and |Vub| using semileptonic decays
constrain γ
E. I. Rosenberg
September 29, 2000
12
BABAR Expected CKM Constraints
Before
BABAR
E. I. Rosenberg
September 29, 2000
13
Differences Bs and Ks
The neutral Ks:
Γ(K0L) = 1.24 x 10-8 s
Γ(K0S) = 0.89 x 10-10 s
mass eigenstates have very different lifetimes
∆m = 3.49 x 10-12 MeV
m =497 MeV few final states are available
The neutral Bs:
Γ(B0L) ≈ Γ(B0H) ≈ 1.28 x 10-14 s we must work with strong
eigenstates and need to tag the flavor of the B
∆m = 3.05 x 10-10 MeV
m = 5280 MeV rich variety of final states to measure mixing
and CP violation (but also more background)
E. I. Rosenberg
September 29, 2000
14
B Flavor Tagging
c
b
lν
W-
In semileptonic decays -- the lepton charge “tags” the B flavor
One needs good lepton ID
b
W+
E. I. Rosenberg
l+
c
ν
September 29, 2000
15
CP Violation in the B sector using the
asymmetry to a CP eigenstate
(
)
)
(
=
0
0
Γ (B → f ) + Γ (B → f )
1 − λ ) cos ( ∆m t ) − 2 Im λ
(
=
0
Γ B0 → f − Γ B → f
cp
af
cp
cp
afcp
fcp
cp
fcp
B0
cp
2
fcp sin
B
1 + λf
( ∆m B t )
B
0
2
cp
if |λ |= 1; af = − Im λf sin ( ∆m B t )
cp
E. I. Rosenberg
cp
September 29, 2000
16
The Golden Channel
J/ψ
c
b
B
c
s
W-
d
d
K0
*
*
*





V
V
V
V
V
Vcs 
0
tb
td
cs
c
b
cd
λ B d → J /ψ K s = − 
 V V*   V V*   V V* 
 tb td   cs cb   cd cs 
(
)
B mixing
Direct CP K Mixing
Im λ (Bd → J/ψ K s0 ) = sin 2 β
E. I. Rosenberg
September 29, 2000
17
The PEP-II Collider Uses the 2-mile Long
Linear Accelerator
E. I. Rosenberg
September 29, 2000
18
E. I. Rosenberg
September 29, 2000
19
Updated 6/14/2000
PEP-II Luminosity Performance
Parameter
Luminosity
Units
cm-2 sec-1
Design
3 x 1033
Achieved
2.56 x 1033
Updated 9/14/2000
Specific
Luminosity cm-2 sec-1 mA-2 /bunch 3.1 x 1030
Horizontal Spot Size µm
Vertical Spot Size µm
220
6.6
2.9 x 1030
190
6.0
PEP-II delivered 20 fb-1 from June 99 through Sept. 2000
E. I. Rosenberg
September 29, 2000
20
B Decays at PEP-II
L = 3x1033 cm-2s-1 ≈ 30fb-1yr -1
+ -
At the ϒ 4s σ (e e → bb) = 1.05nb
⇒ N BB / year ≈ 3 x 10
7
•Good Signal to Background: 1:4
•Only Bd and Bu
•Clean events <ncharged tracks> = 11
•Kinematic constraints
•High efficiency for full reconstruction
E. I. Rosenberg
September 29, 2000
21
Detector Requirements
•
•
•
•
•
To measure decay dependent
BABAR uses:
asymmetries we need:
Large uniform acceptance • Asymmetic layout with a highly
and high efficiency
instrumented forward region
Precision vertex resolution • Five layer Si strip detector with
double sided readout
High resolution particle
• 1.5T B-field and minimum
tracking up to 4 GeV/c
multiple scattering drift chamber
Identification of leptons and • Segmented flux return (µ),
charged kaons
segmented CsI calorimeter (e),
DIRC (π/K separation)
• CsI crystal calorimeter
Efficient, high resolution
detection of neutral pions
E. I. Rosenberg
September 29, 2000
22
The Detector
Si Vertex Tracker
Drift Chamber
CsI Electromagnetic Calorimeter
Particle ID -- DIRC
1.5 T Superconducting Coil
Iron Yoke with Resistive Plate
Chambers for µ and K0L ID
E. I. Rosenberg
September 29, 2000
23
BABAR Is An International Collaboration
E. I. Rosenberg
September 29, 2000
24
E. I. Rosenberg
September 29, 2000
25
E. I. Rosenberg
September 29, 2000
26
E. I. Rosenberg
September 29, 2000
27
SVT and DCH are used to
Reconstruct Charged Tracks
SVT: 5 layers
DCH: 40 layers
1.5 T solenoidal field
Typical values:
1. 0.41 < θ < 2.54 in lab fiducial angle
2. σpt/pt = 0.45% + 0.14% × pt
3. Impact parameter resolution ~60µ @ 1
GeV/c.
E. I. Rosenberg
September 29, 2000
28
SVT - Silicon Vertex Tracker
E. I. Rosenberg
September 29, 2000
29
SVT - Silicon Vertex Tracker
Five layers of double-sided Si
300 µm thick; rad hard
Gives (z,φ) coordinates
150k channels
r = 3.2 …….14 cm
cos θlab = -0.87 … + 0.94
readout pitch: 100 + 50 µm
(inner)
210 + 100 µm
(outer)
E. I. Rosenberg
September 29, 2000
30
Drift Chamber
7104 Signal Wires
40 layers
42…70 mrad Stereo angles
Gas:
He(80%)+iso-C4H10(20%)
E. I. Rosenberg
September 29, 2000
31
Drift Chamber Resolution
All tracks reconstructed-mainly Bhabhas
Average cell resolution
125 µm
TDR resolution 140 µm
E. I. Rosenberg
September 29, 2000
32
Detection of Internally Reflected Čerenkov
Light
Particle ID by Čerenkov
ring detection
12 boxes of 12 quartz bars
each
11000 photomultipliers
E. I. Rosenberg
September 29, 2000
33
DIRC Cerenkov Angle
for tagged π’s and K’s
E. I. Rosenberg
September 29, 2000
34
D0 Signal Improvement with DIRC PID
Momentum of D in Υ4s rest
frame > 1.5 GeV/c
Kaon momentum 0.5-4.0 GeV/c
and point with quartz bar
acceptance
Kaon Cerenkov angle within 2σ
E. I. Rosenberg
September 29, 2000
35
Electromagnetic Calorimeter
6580 CsI Crystals
cos θlab= -0.77 to 0.96
¾90% 4π in cm
¾σE/E = 1.33%⋅E-1/4 ⊕ 2.1%
E. I. Rosenberg
September 29, 2000
36
Instrumented Flux Return
65 cm Fe
Resistive Plate Chambers inserted
with area > 1000 m2
Cylindrical RPCs outside calorimeter
Gas:
Ar(48%)+C2H2F4(48%)+isoC4H10(5%)
Maximum of 21 layers traversed
E. I. Rosenberg
September 29, 2000
37
Calorimeter Performance
Generic B events
with < 6 EMC clusters
Bkgd from combinatorics
Eγ > 30 MeV
Eγ > 100 MeV
E. I. Rosenberg
Eγ > 100 MeV
September 29, 2000
38
Lepton ID Efficiency
E. I. Rosenberg
September 29, 2000
39
Pion Rejection
π/µ Separation from DIRC
π/e Separation from DCH
Sample of identified π’s
from K0s decays
E. I. Rosenberg
September 29, 2000
40
Computing: The Experiment
Within the Experiment
BaBar has chosen an Object Oriented Approach
•Online (2.5 years from scratch)
C++ (written by novices in OO coding)
Java (GUIs for histograms)⇒ JAS
CORBA
•Offline
Reconstruction and analysis all C++
Simulation moving to GEANT 4 (C++)
•Data Storage (40 Tbytes thru April 2000)
Objectivity database
E. I. Rosenberg
September 29, 2000
41
L1 Trigger
E. I. Rosenberg
September 29, 2000
42
L3 Trigger Event Display
Multihadron
Event
E. I. Rosenberg
September 29, 2000
43
D0 Lifetime Measurement
Based on 800 pb-1
PDG98: .415±.004 ps
E. I. Rosenberg
September 29, 2000
44
J/Ψ -> e+e- candidates
J/Ψ -> µ+µ- candidates
The red points are the data. The blue curve is a sum of a Gaussian, an
exponential background. For the electron data, the blue curve also
includes another exponential that describes the Bremsrahlung tail
E. I. Rosenberg
September 29, 2000
45
B’s to Charmonium Final States
E. I. Rosenberg
September 29, 2000
46
CP Violation at PEP-II
1. With e- energy of 9 GeV and e+ energy of 3.1 GeV, the
Υ4s is produced with βγ=0.56
σ(ϒ 4S ) ~ 22% of σ(qq)
2. The Υ4s resonance
decays intoBB pairs in a
coherent L=1 state
3. Tag one B, we know the flavor of the other B at that time.
4. By having a Lorentz boost, the second B decays at a later time
and we have can study CP asymmetries as a function of time
1 - Γ ∆t
1 ± Dsin 2β ( sin ∆m∆t ) 
f ± (∆t,Γ ,∆m,Dsin2β )= Γe
4
D is a dilution factor, D = (1-w), is related to the fraction of
missed tag events, w.
E. I. Rosenberg
September 29, 2000
47
A Tagged B → J/ψ
0
K
s
Event
π–
π+
π–
π+
µ+
K–
µ+
π+
π+
π+
µ–
Ks
K- π+
π–
π+
µ-
K–
µ+
µ–
π+
7-2000
8558A7
E. I. Rosenberg
−
September 29, 2000
48
The BCP Sample
J/ψ KS (KS → π π )
0
0
+
−
124 ±12 events
purity 96%
J / ψ K S0 ( K S0 → π 0π 0 )
∆E (GeV)
purity 91%
0.1
0.05
ψ (2 S ) K S0
0
27 ± 6 events
-0.1
60
purity 93%
BABAR
B0 → J/ψK0 (π+π-)
40
20
∆E (GeV)
2
-0.05
entries/2.5 MeV/c
0
5.2
5.225
5.25
5.275
5.3
2
beam energy substituted mass (GeV/c )
0.1
0.05
0
-0.05
2
-0.1
∆E=E*
entries/2.5 MeV/c
PRELIMINARY
18 ± 4 events
*
reco-E Beam vs.
m ES = E
*2
Beam
E. I. Rosenberg
r *2
− preco
BABAR
10
B → ψ(2S)KS
0
7.5
5
2.5
0
5.2
5.225
5.25
5.275
5.3
2
beam energy substituted mass (GeV/c )
September 29, 2000
ψ (2 S ) K
0
49
The Experimental Asymmetry
There are four time
distributions:
f + : B tag =B; ∆t > 0 f − : B tag =B; ∆t > 0
B tag =B; ∆t < 0
B tag =B; ∆t < 0
f + (∆t) - f − (∆t)
=
A
CP f + (∆t) + f − (∆t)
= D sin 2β sin(∆m∆t)
E. I. Rosenberg
f-
f+
f+
September 29, 2000
f-
50
E. I. Rosenberg
September 29, 2000
51
E. I. Rosenberg
September 29, 2000
52
E. I. Rosenberg
September 29, 2000
53
The Blind BABAR
Analysis
The blinded approach allows systematic studies of
tagging, vertex resolution and their correlations to be
done while keeping the value of sin2β hidden
•The amplitude in the asymmetry, ACP(∆t) , was
hidden by arbitrarily flipping its sign and adding an
offset
•The CP asymmetry in the ∆t distribution was
hidden by multiplying ∆t by the sign of the tag and
by adding an arbitrary offset
E. I. Rosenberg
September 29, 2000
54
Distributions for
the tagged CP
eigenstate decays
of the B’s.
A likelihood fit is
used to extract
the asymmetry A
entries/0.4 ps
The Experimental Distributions
BABAR
10
7.5
B0 tags
5
2.5
0
10
7.5
-5
0
5
0
5
B0bar tags
5
2.5
0
-5
E. I. Rosenberg
September 29, 2000
∆t (ps)
55
Sin 2β from Likelihood Fit
sin 2β = 0.12 ± 0.37 (stat.) ± 0.09 (sys.)
BA BA R
E. I. Rosenberg
September 29, 2000
56
Systematic Error on sin 2β
Uncertainty
Source
B0 Lifetime
Mass Difference
∆z resolution for CP sample
Time resolution bias for CP sample
Measurement of mistag fraction
Different mistag fraction for CP and
non CP samples
Different mistag fractions for
Background in CP sample
Total systematic uncertainty
E. I. Rosenberg
September 29, 2000
0.012
0.015
0.019
0.047
0.059
0.050
0.005
0.015
0.091
57
Sin 2β for the
different tag
catagories
Identify flavor of Btag using
electrons or muons (p*>1.1
GeV/c), or charged kaons.
Or use neural net.
Charge of slow pion from D*Lower momentum leptons;
p* of charged particles
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
sin2β
E. I. Rosenberg
September 29, 2000
58
Asymmetry fits using sin 2β
+1σ
-1σ
E. I. Rosenberg
September 29, 2000
59
Results for channels that should
have no CP asymmetry
Sample
Apparent CP
asymmetry
0.03 ± 0.07
hadronic charged
-0.01 ± 0.08
hadronic neutral
J/ψ K+
0.13 ± 0.14
J/ψ K*0 (K*0 → K+π-) 0.49 ± 0.26
E. I. Rosenberg
September 29, 2000
60
Unitarity Triangle Constraints
The set of ellipses represents the allowed range of (ρ,η) based on our
knowledge of the magnitudes of CKM matrix elements, for a set of
typical values of model-dependent theoretical parameters:
2σ
Experimental
inputs
Experimental
inputs
measurement
jV cb j
j VV j
ub
cb
mBd (ps);1
mBs
jK j (10;3 )
central value
exp. error
.0402
.0017
.085
.008
.472
.017
from A (Moriond 2000)
A
2.271
.017
Theoretical
inputs
Theoretical
inputs
Theoretical est. lower bound higher bound
< VVubcb >
0.070
0.100
p
fBd BBd
0.185
0.255
2
s
1.14
1.46
BK
0.72
0.98
E. I. Rosenberg
1σ
2σ
1σ
_
ρ
1σ
2σ
September 29, 2000
61
Suppose we include direct CP violation
2
2
(1+ λ )
Asymmetry
A
=
CP
D sin 2β sin(∆m∆t)+(1- λ )cos(∆m∆t)
1
0.5
sin 2β = 0.12 ±0.37
0
-0.5
BABAR
-1
Preliminary
-5
1+ λ
2
5
∆t (ps)
Asymmetry
1− λ
2
0
= 0.26 ± 0.19
1
BABAR
Preliminary
0.5
0
-0.5
-1
-1
E. I. Rosenberg
September 29, 2000
-0.5
0
0.5
62
1
sin(∆m∆t)
Sin 2β
Osaka 2000
E. I. Rosenberg
September 29, 2000
63
Summary and Outlook
•PEP-II and BABAR have had an excellent first year
•First events May 16, 1999; preliminary results
July, 2000 (9 fb-1)– more than just sin 2β presented
today
•We expect >25 fb-1 by the end of October and will
reblind the analysis
•Design luminosity almost here; will upgrade to
3x1034 over next five years.
THE FUN IS JUST STARTING
E. I. Rosenberg
September 29, 2000
64