CLEO-c and CESR-c:
A New Frontier of Electroweak And QCD Physics
Kamal Benslama
Seminar at the University of Montreal November 14, 2002
CLEO-c
CLEO-c CESR-cCESR-c
D
oD
o, D
oK
-
+K- K+
+
The CLEO Collaboration
Albany Caltech CMUCornell Florida Harvard Illinois Kansas Minnesota Ohio State Oklahoma Purdue Rochester SLACSMU
UCSDUCSB Syracuse Vanderbilt Wayne State
• Membership:
• ~20 Institutions
• ~155 physicists
• ~1/2 DOE, 1/2 NSF
• Currently expanding…
• Publication history 1980-
• ~320 papers
• diverse physics:
CESR
The CESR Storage Ring
The Storage Ring
Symmetric e+e- accelerator at or near (4S)
(PB ~ 300 MeV/ c)
On the (4S):
e+e- (4S) BB ( ~ 1 nb)
e+e- qq, (q = u, d, c, s) ( ~ 3 nb)
1/ 3 running at OFF (4S) for continuum bkg subtraction
CLEO III collected ON: 6.9 fb-1
OFF: 2.3 fb-1
- -
CESR and CLEO
What is CESR-c and CLEO-c ?
• Modify for low-energy operation:
addwigglers
for transverse cooling• Expected machine performance:
• E
beam~ 1.2 MeV at J/
Ecm L (1032 cm-2 s-1)
3.1 GeV 2.0
3.77 GeV 3.0
4.1 GeV 3.6
One day scan of the ’:
(1/29/02)
1.5 T now,... 1.0T later 93% of 4
p/p = 0.35% @1GeV dE/dx: 5.7% @minI
93% of 4
E/E = 2% @1GeV = 4% @100MeV
83% of 4
% Kaon ID with 0.2% fake @0.9GeV
85% of 4
For p>1 GeV Trigger: Tracks & Showers
Pipelined
Latency = 2.5s
Data Acquisition:
Event size = 25kB Thruput < 6MB/s
CLEO III Detector
CLEO-c Detector
CleoIII Tracking
Silicon (r = 2 - 12 cm)
4 Layers, double sided
Suffering from rad damage
Drift Chamber (DR) (r = 12 – 82 cm)
47 anode layers, 9796 cells
Inner part axial, outer part all stereo
Central Drift Chamber
• DK mass resolution ~ 5.5 MeV
• Ks mass resolution ~ 3.1 MeV
• p/p = 0.35% @1GeV
• dE/dx: 5.7% @minI
Average hit resolution 85 Sweet spot ~ 55-65
dE/dx in the Drift Chamber
•CleoIII dE/dx vs Momentum showing the , K and p bands.
•The K/ separation from dE/dx, in standard deviations.
Inner Tracking (2 < r < 12cm)
• Now: 4 Layer Silicon Detector
• Double sided wafers,
• 125000 channels
• 1.6% X0 thickness
• early radiation damage:
Layers 1 & 2 r eff
• Proposed: 6 layer drift chamber
• all stereo ( 10-150 tilt)
• 300 channels: smaller evt size
• 1.1% inner wall + 0.1% gas
• improves p/p (less material)
• worsens z0 (stereo)
• preserves mass resolution
• no effect on pattern recognition
• build from spare parts
Silicon Radiation Damage
(rad)
• Efficiency losses exhibit wafer geometry, not detector geometry.
• rphi side only. Z side so far shows no eff loss.
Wire vertex Chamber
RICH Detector
Performance with Bhabhas (only RICH tracking):
single resolution N resolution/track Flat radiator 14.0 mrad 10.2 5.3 mrad
Sawtooth radiator 12.2 mrad 12.0 4.2 mrad
Thinnest (19 cm) <0.12% X0
Proximity Focusing
Solid Radiator (LiF)
45o saw tooth outer face (40%)
Rest flat radiator
15.7 cm N Expansion Gap
MWPC photon detector
Photosensitive gas = TEA (Triethylamine)
abs=0.5 mm, <165 nm
3 K/ separation at 2.65 GeV/c
RICH
Cherenkov Images
Flat radiators
Sawtooth
The RICH Smiles
e
+e
-e
+e
-PID using the RICH
# 1
#
1
)
| (
)
| (
i
back i
signal i
back K
i signal K
P P
L
P P
L
•Likelihoods are defined as
where
- CB line shape (Gaussian with power law tail on the high side)
- constant
2K2= -2log(LK/L)
•Likelihood method folds in photon count and Cherenkov angle measurement
together
signal
P
Pback
Efficiency vs fakes study
•cut on
2K
2•fit m(D
0)
2K 2 x
2K 2
K
K
mD0(GeV)/c2
Number of events
CsI Calorimeter
•93% solid angle coverage - uniform resolution in barrel & E.C.
E/E = 4% @ 100 MeV 2% @ 1 GeV
•~5.5 MeV
0mass resolution
Typical Hadronic Event
(From online event display)
Look What RICH did to our Data !
Signal characteristics:
Two stiff back-to-back particles
Simple, robust analysis.
Statistics limited, not systematics
Backgrounds:
Entirely from “continuum”
Continuum background characterized by event shape variables
CLEOIII Rare B Analysis
Continuum occasionally produces high momentum, back-to-back, 48%
35%K , 17%KK.
Roughly 25% from , 75% from , ,
Primary handle is in global event shape characteristics
π π π
d c d
c uu
s s
Signal Continuum
Spherical event 2-jets
Shape Structure
Continuum Background
Analyses Strategy
Study backgrounds (using Monte Carlo and OFF-resonance data)
Reconstructions variables
Maximum likelihood fit
PS:
I would prefer Neural Network
Monte Carlo Experiments:
Study possible biases
Apply the fit to data
signal
continuum
Energy difference
Beam constrained mass
~ 2.5 MeV (3.0 MeV if 0 )
Resolution is mode
dependent, but generally
~ 20-25 MeV (2 worse if 0 )
2
| p |
m 2 beaE
BM
m beaE
i ersE
ght
dau B E
dE/dx RICH
Reconstruction Variables
Signal is isotropic, continuum is jetty Use shape variables
Sphericity angle
R2 (Fox Wolfram moment)
Momentum flow in nine cones around the sphericity axis
Combine the variables in a Fisher discriminant
Reconstruction Variables
N
bN
sand are the yields to be determined
PDF shapes for signal are determined from MC
PDF shapes for continuum are determined from OFF-resonance data
Maximum Likelihood Method
Fit variables:
Fit component fractions:
Outputs:
plus background, at the maximum of the Likelihood function.
Efficiency: Use of Maximum Likelihood fit doubles the efficiency of the analysis compared to a normal counting analysis
Systematics
:
All shapes are systematically varied to study effects. Correlations: typically at few % level, except
- where kinematics leads to ~ 20%
correlation.
M
BΔE F
dx 2
dE
cosθB
f
KπNππ
dx 1
dE
f
ππf
KKNKπ NKK
ΔE M
BMaximum Likelihood Fit
3 . 1 8
. 18 6
. 18 2
.
29 67..14 44..1533..40 22..68
Yield BR(BK)(x10-6) BK CLEOIII CLEO(1999)
Good agreement: between CLEOIII & CLEO II &
with BaBar/Belle.
How well does CLEO III work?
1
stresults from CLEO III data at Lepton-photon
2001
FINAL Results VERY Soon
The CLEO-c Program
20 02
20 03
20 04
20 05
Prologue: Upsilons ~1-2 fb-1 ea.
Y(1S) ,Y(2S), Y(3S)…
Spectroscopy, Matrix Elements, ee
10-20 times existing world’s data
Act I: (3770) -- 3 fb-1
30M events, 6M tagged D decays
(310 times MARK III)
Act II: √s ~ 4100 -- 3 fb-1
1.5M DsDs, 0.3M tagged Ds decays
(480 times MARK III, 130 times BES II )
Act III: (3100) -- 1 fb-1 1 Billion J/ decays
(170 times MARK III 20 times BES II)
•Charm events produced at
threshold are extremely clean
•Large , low multiplicity
•Pure initial state: no fragmentation
•Signal/Background is optimum at threshold
•Double tag events are pristine
–These events are key to making absolute branching fraction
measurements
•Neutrino reconstruction is clean
•Quantum coherence aids D mixing and CP violation studies
Why run on threshold Resonances ?
A typical Y(4S) event:
Tagging Technique
•
Pure DD or D
sD
sproduction
Many high branching ratios (~1-10%)
High reconstruction eff
Two chances
high net efficiency ~20% !
D K tag.
S/B ~ 5000 Ds KK tag.
S/B ~ 100
Beam constrained mass
6M D tags 300K Ds tags 6M D tags
300K Ds tags
• We expect great advances in flavor and electroweak physics during the next decade:
– Tevatron (CDF, D0, BTeV,CKM).
– B-Factories (BaBar, Belle).
– LHC (CMS, ATLAS, LHC-b).
– Linear Collider (?).
• What could CLEO-c possibly have to offer this program?
Why CLEO-c ? Why Now ? Why CLEO-c ? Why Now ?
To score nice goals we
Absolutely need an excellent player who can make the Perfect passes at the perfect time
CLEO-c
Precision Standard Model Tests
f
D+and f
Dsat ~2% level
.Absolute hadronic charm branching ratios with 1- 2% errors.
Semileptonic decay form-factors (few % accuracy).
Contribution to CKM
Measurements
Absolute Branching Ratios
~ Zero background in hadronic modes
Decay Mode PDG2000 CLEOc (B/B %) (B/B %)
D0
K 2.4 0.5
D
+ K 7.2 1.5
D
s 25 1.9
Set absolute scale for all heavy quark meas.
The importance of absolute Charm BRs
Stat: 3.1% Sys 4.3% theory 4.6%
Dominant Sys: slow, form factors
& B(DK) dB/B=1.3%
V
cbfrom zero recoil in B D*
+
CLEO LP01
10
3) 1 . 2 0
. 2 4
. 1 4
. 46
(
cb
V
Decay Constants : |f
D|
2|V
CKM|
2Ds
D
f
Ds
f
Ds 2.1%
f
Df
D 2.6%
KL
(Now: ±35%)
(Now: ±100%)
Br(Ds
Importance of measuring f
D& f
Ds: Vtd & Vts
1
2
3 2
1
10 8 . 8 50 200
.
0
d B B Vtd
MeV f ps B
M d d
d d
B B
B B
d d
B f
B f
M
M ) ( )
5 ( . ) 0
(
1.8% ~15%
Lattice predicts fB/fD & fBs/fDs with small errors
if precision measurements of fD & fDs existed (they do not)
could substitute in above ratios to obtain precision estimates of fB
& fBs and hence precision determinations of Vtd and Vts Similarly fD/fDs checks fB/fBs
(lattice)
2 2
ts td B
B B B s
d
V V f
B f B M
M
s s
d d
i
1
i
1
~5% (lattice)
COMPARISON
0 5 10 15 20 25 30
Error (%)
f
D) K
D (
Br
) D
(
Br S
) K D
(
Br 0
BaBar 400 fb-1 Current
Comparison between B factories & CLEO-C
Systematics &
Background limited
CLEO-c
3 fb-1 Statistics limited
abcdefghi
f
Ds|f(q2)|2
|VCKM|2
Absolute magnitude & shape of form factors is a great test of theory.
b
c
u
d HQET
l
l
1) Measure D form factor in Dl (CLEO-c): Calibrate LQCD to 1%.
2) Extract Vub at BaBar/Belle using calibrated LQCD calc. of B form factor.
3) Precise (5%) Vub is a vital CKM cross check of sin2.
4) Absolute rate gives direct measurements of Vcd and Vcs.
B
D i.e.
Semileptonic Form Factors. Semileptonic Form Factors.
2 2 2 3
2 2
2
( )
24 G V P f q dq
d
K F cs
Semileptonic Decays |V
CKM|
2|f(q
2)|
2Decay Mode PDG2000 CLEOc (B/B %) (B/B %)
D0 Kl 5 1.6
D0 l 16 1.7
D+ l 48 1.8
Ds l 25 2.8
Plus vector modes...
U = Emiss - Pmiss LowBkg!
e e D
e K D
e K D
e D
e D
e K D
e K D
e D
e D
e K D
e K D
c s
s s
: 12
: 11
: 10
: 9
: 8
: 7
: 6
: 5
: 4
: 3
: 2
: 1
0
* 0 0 0 _
0
* _
0 0
0
* 0
0
Semileptonic dB/B, Vcd, & Vcs Semileptonic dB/B, Vcd, & Vcs
0 10 20 30 40 50 60 70 80 90 100
1 2 3 4 5 6 7 8 9 10 11 12
Decay modes
Error (%)
CLEOc PDG
Vcs /Vcs = 1.6% (now: 11%) Vcd /Vcd = 1.7% (now: 7%)D0 K e
e D0
Use CLEO-c validated lattice + B factory Blv for ultra precise Vub
D0 l
D0 Kl
How can CLEO-c Contribute to CKM Measurements ?
An illustration using a variant of the 95% Scan method.
Allowed regions of the - plane using:
• current experimental results and
• conservative theoritical uncertainties
Allowed regions of the - plane using:
• current experimental results and
• theoritical uncertainties of O(1%)
•2% decay constants and 3% semileptonic form factors
CLEO-c: Probes of new Physics
Mixing sensitivity at the 1% level.
CP violation sensitivity at the 1-2% level.
•
Rare Decays.
Sensitivity: 10-6D0
D0
K-
K-
D0
D0
K+
K-
Ratio of Rates:
Charm Mixing Charm Mixing
x = /
y = /2
To 1st order, where
e+e ” D0D0
NO DCSD
Quantum coherence
At the (3770)
e+ e
D0
D0
+
K+
e+e ” D0D0
JPC = 1
i.e. CP+
Suppose both D0’s decay to CP eigestates f1 and f2: These can NOT have the same CP :
CP Violation CP Violation
Observing this is evidence of CP
Ex:
(K
K
)(π
π
)
CP(f1 f2) = CP(f1) CP(f2) (-1)l = CP
+ (since l = 1)
Compare to B Factories
CLEO-C BaBar Current
2-4fb-1 400 fb-1 Knowledge
f_D |Vcd| 1.5-2% 10-20% n.a.
f_Ds |Vcs| <1% 5-10% 19%
Br(D+ -> ) 1.5% 3-5% 7%
Br(Ds -> ) 2-3% 5-10% 25%
Br(D->l) 1.4% 3% 18%
Br(c -> p) 6% 5-15% 26%
A(CP) ~1% ~1% 3-9%
x'(mix) 0.01 0.01 0.03
Systematics & background limited.
Statistics limited.
• and Spectroscopy
–Masses, spin fine structure –Leptonic widths for S-states.
–EM transition form factors
•run on resonances winter ’01-summer’02
•~ 4 fb-1 total
Uncover new states of matter
Glueballs G=|gg Hybrids H=|gqq
Requires detailed understanding of ordinary hadron spectrum in 1.5-2.5 GeV mass range.
Study fundamental states of the theory
Calibrate and test theoretical tech.
Probing QCD
•Gluons carry color charge: should bind!
• But, like Elvis, glueballs have been sighted too many times without confirmation....
• CLEO-c: find it or debunk it!
huge data set
modern detector
95% solid angle coverage
• Radiative decays are ideal glue factory:
Gluonic Matter
X
c
c¯
• CLEO-c: 109 J/ ~60M J/ X
• Partial Wave analysis
• Absolute BF’s: ,KK,pp,,…
1
• perfect initial state
• perfect tag
• glue pair in color isosinglet
The dubious life of the f
J(2220) (A case study)
Now you see it…
(1996)BES MARKIII
(1986)
L31997
Now you don’t…
Crystal barrel:
pp
?? LEAR
1998
OPAL1998
L3 Signal
2 3
MKK
2 3
… or do you?
… or don’t you?
f
J(2220) in CLEO-c?
f
J(2220) in CLEO-c?
Two Photon Data: fJ2220
– CLEO II: B fJ /KSKS < 2.5(1.3) eV – CLEO III: sub-eV sensitivity
Upsilonium Data: (1S): Tens of events
5000 –
8500 32
pp
5300 23
KSKS
18600 46
K+K-
13000
18
32000
74
CLEO-C BES
CLEO-c has corroborating checks:
2
3
Inclusive Spectrum J/ X
10-4 sensitivity for narrow resonance Eg: ~25% efficient for fJ(2220)
Suppress hadronic bkg: J/X
Additional topics
• ’ spectroscopy (10 8 decays) ’chc…
•
at threshold (0.25 fb
-1)
• measure mto ± 0.1 MeV
• heavy lepton, exotics searches
•
c
cat threshold (1 fb
-1)
• calibrate absolute BR(
cpK)• R= (e
+e
- hadrons)/ (e
+e
-
+
-)
• spot checks
If time permits Likely to
be added to run plan
-
The CLEO-c Program: Summary
•
Huge data set• 20-500 times bigger than previous experiment
•
Modern and understood detector•Experienced Collaboration
•
Powerful physics case
• Precision flavor physics -
• Nonperturbative QCD -
• Probe for New Physics
•
Very small and well-controlled backgrounds•Very small and well-understood systematic errors
• A large number of and wide variety of precision measurements to challenge and validate theory
CLEO-c Physics Impact
•CLEO-C workshop (May 2001) : successful ~120 participants, 60 non-CLEO
•Snowmass working groups E2/P2/P5 : acclaimed CLEO-c
• HEPAP endorsed CLEO-c
•CESR/CLEO Program Advisory Committee Sept 28 Endorsed CLEO-c
•Proposal submission to NSF was on October 15,2001
•Site visit on Jan/Feb 2002: Endorsed CLEO-c
•Science Board March 2002,
• Expect approval shortly thereafter
•See http://www.lns.cornell.edu/CLEO/CLEO-C/
for project description
Invitation
If interested in CLEO-c program you are more
than welcome to join my ex-colleagues. They have room for you !
More information is available in CLEO Web page: www.lns.cornell.edu/CLEO/CLEO-c
Contact person: spoke@mail.lns.cornell.edu
CLEO-c
CLEO-c CESR-cCESR-c