Heavy Flavour Physics at the Tevatron

1
Heavy Flavour Physics at the Tevatron
Zero to Z0 Conference Fermilab, May 12-14 2004

• Farrukh Azfar, Oxford University (CDF)

2
Overview of this presentation

• Preliminary
• 1) Tevatron performance, Beauty physics at hadron
  colliders
• 2) CDF and D0 detectors, relevance for B-physics
• Physics Results Prospects
• 3) Tests of Heavy Quark Expansion (HQE) Masses
  and Lifetimes of B hadrons. Hadronic Moments.
• 4) Search for Flavour Changing Neutral Current
  (FCNC), Rare decays
• 5) Mixing and CP violation (CPV), Toward
  Bs-mixing CKM angle g.
• 6) Conclusion and Summary

3
Tevatron pp collider upgrade performance,
integrated luminosity
Run-IIa -Goals are ?Ldt2fb-1 (x20 Run-I,
1992-96) Run-II Tevatron Upgrades -Main
Injector for Tevatron -Higher proton intensity
-Anti-proton transfer efficiency increased
-Anti-proton recycler (coming after autumn)
CDF
Performance Improvement -Collision rate 3.5 ms
?396 ns - Bunches 6x6 ?36x36 -Center of Mass
energy 1.8?1.96 TeV/c2 -Peak luminosity
2.4x1031?7.2 x1031cm2s-1 (Below target by x2.5,
but improving)
Data taking efficiency80-90 for CDF D0

Results in this talk CDF analyses 65-250 pb-1
D? analyses 115-250 pb-1 290 pb-1 on tape at
CDF D0
4
Why Beauty at the Hadron-Hadron Colliders ?
s(bb) at ?(4S) 1nb (B-factories)
(Compare bb production s(bb) at Z0 7nb (LEP)
cross section) s(bb) at pp
(1.96TeV/c2)150mb (Tevatron Experiments) More B
_at_ Tevatron but inelastic s is 103 x s(bb)
-Select b-data online, key right detector
triggers -Rewards all B-hadrons e.g. B?, B0,
Bs, Bc ?, ?b (unlike B-factories) higher s
than at Z0 Clever Online B Selection
(Triggers) Traditional Use leptons from e.g.
Bs?Dsm-n CDF D0 (single-lepton) B ? J/yK,
J/y? mm- CDF D0 (di-lepton) Modern long B
lifetimes ?large impact parameter (IP) of
daughter tracks CDF (D0 in progress) SVT
trigger purely hadronic decays of B and
Charm e.g. D0?pp- , Bs?Dsp-, Ds?f p f?KK-
(1st _at_ hadron machine!) CDF Apply High IP
requirement in single-Lepton data as well

5
The CDF D0 Detectors in Run-II
CDF Detector
CDF D0 Detectors are both Multi-purpose
with -Axial Solenoid -Inner Silicon micovertex
detectors -Outer trackers -Calorimetry -Muon
ID -Muon Triggering (CDF D0) -High IP Track
triggering (CDF) D0 Better calorimetry, better
muon tracking coverage CDF Better momentum
measurement, also can select high IP tracks, some
Hadron ID with dE/dX, TOF
D0 Detector
6
Physics Results, Testing HQE B-hadron lifetimes,
massesGoals, Techniques
Goal test the HQE Predicted B Lifetime hierarchy
tBc ltlt tXb0 tLb lt tBd tBs lt t B- lt t
Xb-

• Fully Reconstructed B from J/y?mm- di-muon
  trigger (e.g. Bs ?J/yf) or High IP trigger (e.g.
  Bs?KK-)
• Find vertex, 2-d distance Lxy invariant mass MB
  
• momentum in 2-d PtB Find proper time ct Lxy.
  MB/PtB
• - Fit mass distribution only or mass and lifetime
  distributions

• Partially Reconstructed e.g. Bs ? mDs-n, B?
  mnD0
• 1-lepton (High IP track CDF )trigger
• Missing n means ct Lxy.MB/PtB
  Lxy(D0m).MBK./Pt(mD0)
• -K Pt(mD0) /PtB from MC high statistics but
  worse sct

-Decays selected using SVT trigger have biased ct
-Turn-on near low IP cut -Turn-off at high
IP cut -Bias fix underway Then measure Lifetimes
in Bs?Dsp- etc.
7
Physics Results, HQE Bs ?J/y f, Lifetime and
Mass
J/y ? mm- , f ? KK- (using di-muon (J/y)
trigger) Run-I 60 at CDF. Run-II D0403,
CDF269 Largest sample of fully reconstructed Bs
remains at the Tevatron
CDFM(Bs)5366.01?0.73(stat)?0.0.33 (sys)MeV/c2
D0M(Bs)5360?5 MeV/c2 CDF t(Bs)1.347?0.099?0.01
3 ps t(Bs)/t(Bd) 0.89 ?0.072 D0 t(Bs)
1.190?0.18?0.014 ps (69 events, update in
progress) Mass Lifetime Results from assorted
other fully reconstructed decays CDF t(LB)
1.250.260.10 ps (?b?J/y?), M(LB)5619.71.21.2
MeV/c2 Bs and LB mass measurements remain worlds
best..
8
Physics Results (aside) Bs width difference DGs
angular separation of CP eigenstates
CDF D0 fit 1 lifetime But there are 2
tCP,tCP- DGBs1/tCP-1/tCP- DGs predicted
to be large 10, provides SM consistency check
DGsA.DMBs If DMBs is large DGs is small or
vice-versa?sign of new physics Need to measure
lifetime(s) can do, and determine CP content
Use angular analysis(CDF) and put them together
(when we have higher statistics)
Convenient basis transversity Allows easy
separation of CP content of B?VV decays
Analyse Bs ?J/y f Bd ?J/yK as a
check (J/y?mm-, f?KK-, K(892)?Kp-) PDF has 3
angles f(QT ,FT, QK) with amplitude parameters
A?,A? A0 so that A?2CP odd fraction A?2A02
CP even fraction
9
Physics Results (aside) Bs width difference DGs
angular separation of CP eigenstates
Using 176 Bs?J/yf 993 Bd?J/yK(892) (as a
check)
Compatible with BaBar Belle
Polarization analysis indicates CP 0.770.10
The larger the dominance of a CP eigenstate the
greater the accuracy of DGs Analysis will be
done at D0 as well !
10
Physics Results Testing HQE Charged to Neutral
B-Meson lifetime ratio (D0)
D0 Charged to Neutral B-Meson Lifetime Ratio
t/t0 Use -B ? mnD(2010)-X mostly Bd -B ?
mnD0X mostly Bu -Calculate ratio of
events/lifetime bin N/N0 e-(t/t0-1)t
?(K-factor)?st -Calculate expected ratio using
all BRs in terms of k t/t0, and N (overall
normalization) Minimize c2 determine k and
N D0 Result t/t0 1.093 0.021 0.022,
N1.0010.012 BaBar t(B)/t(Bd) 1.064 0.031
0.026 CDF t(B)/t(Bd) 1.080 ? 0.042
Belle t(B)/t(Bd)
1.0910.0230.014 (B?J/yK
Bd?J/yK0)
One of the Worlds best single measurements
11
Physics Results Testing HQE More B decays used
?b?J/y? (CDF)
Bd?J/y K (D0)
B?J/y K(D0)
B?J/y K(CDF)
12
Physics Results Hadronic Moments from D decays

• 1) HQE G(B?Xclnl)GF2Vcb2 mb5.S(Cn/mb)n with
  Cn lt0OnHQE0gt (non-perturbative, can extract
  from data)
• 2) Free parameters L at O(1/mb), l1l2 _at_
  O(1/mb2) , etc.
• 3) Moments M1,M2 of Xc invariant mass
  distribution
• from B-decays
• have expansions similar to 1) i.e..in terms of L
  l1l2 (sHMXc2)
• 4) By finding pdf hence M1, M2
  -gtconstrain L, l1 , improve Vcb measurements
• 5) Now
• First 2 pieces from D, D0 are well known. f
  (sH) comes from narrow
• wide Dhigher order(resonant non-resonant)

Vcbincl (41.9 0.7exp 0.6theo) ?10-3
contains L, l1
13
Physics Results Hadronic Moments
Reconstruct B-? D0 l- ? Find m and D0
consistent with B parent (vertex). Use lepton
high IP track data.
f(sH) distribution
Reconstruct D0?D?- D0?D ?- decays are
reconstructed, moments m1, m2 calculated wrt
f(sH)
In going from m to M assume -lepton p in B rest
frame gt700 MeV -MD, MD , Branching ratios from
PDG -Only D decays to 1p D, D
M1 M2 from D, D D best single measurement !
14
Rare B decays B s(d)?mm- at CDF

• -Use high-mass di-muon data
• -BRSM(Bs?mm- )(3.8 1.0) 10-9 some extensions
  predict x103 BRSM
• Variables Mass, lifetime, Df from vertex
  Isolation
• 1 background event expected, 1 event seen no
  excess-gtBR limit

BR Upper Limit at 95 CL 7.5x10-7 Bs ? mm-
(Worlds best) 1.9x10-7 Bd ? mm- BR Upper Limit
at 90 CL 5.8x10-7 (Bs ? mm-) 1.5x10-7 (Bd ?
mm-)
Blind analysis cuts optimization before
looking at the signal mass region
Bd result Belle 1.6x10-7 BaBar 2.0x10-7
Submitted to PRL
15
Physics Results Bs ?mm- limits from D0
-Use MC for signal data, background for cut
optimisation -Expect 7.3 ? 1.8 background events
in signal region
The analysis has not been unblinded yet (signal
region still hidden). It is still being
optimized (without bias) and expected to improve
?180 pb-1
Expected limit (Feldman/Cousins) Br(Bs ? ? ?-)
lt 9.1 ? 10-7 _at_ 95 CL (stat only) Br(Bs ? ?
?-) lt 1.0 ? 10-6 _at_ 95 CL (stat
syst) (expected signal has been normalised to B?
? J/? K? for BR limit calculation)
16
Rare Decays Bs? ff Observation BR (SVT
Trigger) CDF

• 1) Bs? ff decays via second order weak decay
• 2) SUSY coupling could enhance the SM BR (10-5)
• 3) Comparison of angular distributions of various
  B ?VV decays can determine a and g

First observation (4.8s) ! Blind analysis
1) Normalization Mode Bs?J/y f 2) Relative
Efficiencies from MC 3) N(Bs ? J/y f) is
corrected for Reflections from Bd? J/y K 4)
J/y ? mm- f?KK- Bs?J/y f BRs taken from PDG
BR (1.4 0.6 0.2 0.5 (BR))x10-5 (SM
3.7x10-5) Upper Limit lt2.7x10-5 _at_ 95 CL
17
Mixing and CP violation (CPV) in Bd,s decays,
basics
-Mixing tag B-flavour at birth, decay to flavour
specific state asymmetry AmixCos(Dmd,st) -CPV
tag B at birth, decay to CP eigenstate
asymmetry(t) AcpvAcpv,direct.Cos(Dmd,st)Acpv,mi
xing.Sin(Dmd,st)
Issues Tagging Flavour Correctly
.... being able to tag at all
Statistical power N tagged events eD2N pure
events
Opposite side tagging
Same side tagging
ConceptLook for B on opposite side of B of
interest -Look for m,e -Use weighted jet-charge
Disadvantages Opposite B not in acceptance
(60) or mixes (B0)
Concept Look for p (K) from hadronization of B
(Bs) of interest, Higher e
Check algorithms in known b-flavour decays eg B
? J/yK Prepare for Bs mixing by first doing Bd
mixing
18
Proof of principle Bd mixing at D0
-Data sample lepton triggers -Bd ?mD(2010)-X
(D-?D0p-,D0 ?Kp-) -Find m, D0,p- consistent
with B -Select events within DM(D-,D0)PDG-
DM(D-,D0)lt0.04GeV/c2 -Opposite-side m tags
flavour -Use PDG BRs to calculate expected
observed asymmetry(t) -DMd Purity are free
parameters fit
Preliminary results ?md 0.506 ? 0.055? 0.049
ps-1 Consistent with world average 0.502 ?
0.007 ps-1 Tagging efficiency 4.8 ? 0.2
Tagging Purity 73.0 ? 2.1 First D0
mixing Measurement !!
19
Proof of principle CDF (Run-II) DMd measurement
CDF Run-I Dmd (all methods) 0.495 0.026
0.025 ps-1
First Run-II mixing result same side tagging
(SST) Find fragmentation p? from B, track near B
with lowest relative PT
B?J/y K, ?D0p to check tag, B0?J/y K0, D?p?
for Dmd
-1.1KB0?J/yK0 (J/y data) -4.9K B0?D?p? (SVT
Trigger!) DMd0.550.10 ps-1 Dilution (D) 12.4
eD21.00.5
Work on jet-charge opposite side muon Tagging
continues
20
Physics Prospects Toward Bs mixing at CDF
fully reconstructed decays B0s? Ds???
First observation of mode Bs?Dsp-with (Ds ?fp,
f ?KK-) ! Flagship Mode for Bs mixing !
Decays we plan to use B0s? Ds???, B0s?
Ds??????? Proper time resolution ?t 67 fs ? t
? ?PT/PT

-Need to tag initial B flavour -projection awaits
final eD2
-Currently have reconstructed only
Ds??fp? -Reconstruct with more Ds decays eg
K0K?, ????? to improve yields
21
Physics Prospects Toward Bs mixing semi-leptonic
decays
Use leptons (CDF leptonhigh IP track) select
Bs?mDs-X
Find leptonDs-?fp- f ?KK- lepton has charge
opp. to Ds
Plots have different mass resolution and S/B

• Also Lifetime measurement provides valuable
  constraint on DGBs
• t(tcp2tcp-2 )/(tcptcp-) as in B0s? Ds???

22
Physics Prospects CP violation in B?hh- (SVT
data) decays determining angle g (CDF), Method
B?hh- from hadronic trigger Data ! (891 events)
Tree gt penguin in B?pp- vice-versa in Bs?KK-
Four unknowns In Asymmetry(t) dratio of
penguin/tree hadronic matrix elements q phase
of d g,b weak phases Constrain b from
B-factories, measure g by fitting asymmetry (t)
Proposed by R.Fleischer, PLB459 1999 306
dE/dx check Use D?D0p?, D0? K?p?
1st Stage Statistically Separate Bd ?pp-, Bd
?Kp-, Bs ?Kp-, Bs?KK- - Use M?? vs
a(1-p1/p2)?q16 distinct shapes for pp- KK-,
(Bd,Bs?) Kp-, pK- -Use dE/dX distinguishes Kp
to 1.16s in the future use DmBd DmBs too
23
Physics Prospects CP violation in B?hh- decays
determining angle g (CDF)
Yields (Results from 65 pb-1) Bd?pp-
148?17 Bd?K?p? 39?14 Bs?K?p?
3?11 Bs?KK- 90?17 (Bs?KK- First
Observation !)
Results use 65 pb-1 sample, 1.16s dE/dX Kp
separation Update with dE/dX (1.4s) 180 pb-1
underway !
Sanity check (spot on !) Measure Ratio of
Branching Ratios CDF G(Bd?pp-)/G(Bd ?Kp-)
0.26 0.110.055, PDG
Ratios of BRs (CDF) ACP(Bd?pp-) (B-factories)
Check SM consistency (D.London)
Finally we expect (Fleischer method) (2fb-1)
?(?) 10?(stat) 3?(syst SU(3) breaking)
24
Conclusions

• CDF D0 are completing 1st phase (250pb-1) of
  data taking
• a) Current s(t(Bu)/t(Bd0) ) surpasses
  theoretical accuracy. Also tests of vertexing
  tracking (for future DMBs and CPV)
• b) Search for FCNC set limits on rare BRs
• c) Prepare for Bs mixing Establish by
  measuring Bd mixing first !
• 2) Next phase (gt250pb-1 lt500pb-1) will
• a) set limits on (or observe) Bs mixing
• b) set limits on DGBs
• c) search for CPV in the neutral B system
• d) Continue to improve limits of Rare Decay
  BRs
• 3) Final Phase (end Run-IIa) (gt500pb-1 and
  lt2fb-1) all of the previous
• a) Achieve better than 1 accuracy on
  s(t(Bs)/t(Bd) ) s(t(Bd)/t(Lb))
• b) Measure Bs mixing parameter xs expect to
  measure d(DGs)5
• c) Measure CKM angle g
• d) and search for unexpectedly large CPV in
  Bs ?J/yf
• Last phase will be mostly complementary to the
  B-factories

25
Backup Slides
26
Aside Physics Results Ratio of branching
ratiosof Bs?Ds?p? to Bd?D?p?
Interest in Bs?Ds?p? is mostly due to Bs mixing
butweve also measured the ratio of branching
ratios G(Bs?Ds?p?)/G(Bd?D?p?)
Normalization mode is Bd?D?p?, D?
?K?pp- Kinematically Bs?Ds?p?, Ds? ?fp?,
f?KK-
Ratio of Bs to Bd signals is
Where e are determined from Monte-Carlo D?, Ds?
BR are from PDG, obtain
Using fs/fd 0.2730.034 from PDG obtain
were beginning to fill in PDG section on the Bs
27
Data Samples B and Charm from the hadronic
trigger
0.5M Charm decays at CDF ?10-20 come from B
Great Potential for B and Charm Physics, opens
at least as many avenues as J/y trigger
Some charm is prompt
D from direct charm Points back to beam spot
..Some charm is from B
D from B has a impact Parameter wrt beam spot
An example of B reconstructed Using data from
this trigger
..to separate prompt Ds from Ds coming from B
Prompt Charm D0?K?
86.5 ? 0.4 (stat) D??D0 87.6 ?
1.1 (stat) D??K?? 89.1 ? 0.4
(stat) Ds??? 72.4 ? 3.4 (stat)
We have B and tons of Charm as well !
28
Physics Results Average B-hadron lifetime from
partially reconstructed B?J/yX decays
This is a sanity check of our B?J/y sample
Obtain Average B hadron tB From all B?J/y (other
stuff) decays B is not fully reconstructed
D0 Inclusive B Lifetime
Partially reconstructed B -Correct for missed
daughters F(PT) (from by Monte-Carlo) -tB
is an estimate -it is the average lifetime of
all hadrons decaying to J/y
Results from D0 and CDF tB1.561?0.024?0.074 ps
D0 (40 pb-1) tB1.526?0.034?0.035 ps CDF (18
pb-1) Both consistent with PDG tB 1.564 ?
0.014
29
Bs width difference DGs and angular variable
separation
Two CP states transversity

• One lifetime(width) has been fit in this mode
• But contains two distinct lifetimes CP CP-
  Bs, significant lifetime (width) difference
• DGs1/tB1-1/tB2
• 2) Extract DGs fit two lifetimes, use single
  angle to separate CP and CP- Bs (Transversity
  angle q)
• SM prediction for DGs 0.10Gs also DGs A.xs
  (xs Bs mixing parameter) if DGs is small and xs
  is large or vice-versa Sign of new physics
• CDF prediction for 2fb-1 d(DGs)0.05

Two CP states lifetime
Total function and normalization
Current limit (LEP) ??s / ?s lt0.31, from
branching ratio of Bs?Ds()Ds?()
Note SM CP violation in this mode O(3), if
large ?new physics CP asymmetry sin2e ? DGs,
measured DGs,SM.Cos2e (complementary)
30
Physics Results lifetime, mass, from fully
reconstructed B? J/y X modes Standard Technique
Probability Density Function (pdf)
Data from J/y?mm- di-muon trigger 1)
Reconstruct vertex according to decay topology 2)
Calculate decay proper time mass errors 3) If
fitting for massfit mass only 4) If fitting for
lifetimeFit mass and lifetime using bi-variate
Probability density function (PDF) in likelihood
1) Signal Lifetime


2) Signal Mass
An Example B -gtJ/y K at CDF
3) Signal for Mass only analyses
4) Signal pdf in mass and lifetime
5) Signal for lifetime analysis
6) Normalization mass lifetime
Both the mass and lifetime distributions are fit
in a single step. Technique is applied to B
gJ/y K, B0 gJ/y K0 (K0 g Kp), Bs g J/y f (f
gKK), LbgJ/yL (Lgpp)
31
B physics prospects(with 2fb-1)
Both competitive and complementary to B
-factories

• Bs mixing Bs ?Dsp(Ds3p) (xs up to 60, with xd
  meas. one side of U.T.)
• Angle ? B0? J/? Ks (refine Run1 meas. up
  to ?(sen2?) ? 0.05)
• CP violation, angle ? B0? pp(pK), Bs ? KK(Kp)
  
• Angle ?s and ??s/ ?s Bs? J/? ? (probe for
  New Physics)
• Precise Lifetimes, Masses, BR for all B-hadrons
  Bs, Bc, ?b
• (CDF observed Bc ? J/? e(?)?. Now hadronic
  channels Bc ? Bs X can be explored)
• HF cross sections (beauty and charm)

• Stringent tests of SM or evidence for new
  physics !!

32
Physics Results Average B-hadron lifetime from
partially reconstructed B?J/yX decays.
This is a sanity check of our B?J/y sample
Obtain Average B hadron tB From all B?J/y (other
stuff) decays B is not fully reconstructed
If Fully reconstructed B -ct c.(time in B rest
frame) -Lxy 2-d decay length -MB mass -PT
transverse momentum
D0 Inclusive B Lifetime
If Partially reconstructed B -Correct for missed
daughters F(PT) (from by Monte-Carlo) -tB
is an estimate -it is the average lifetime of
all hadrons decaying to J/y
Results from D0 and CDF tB1.561?0.024?0.074 ps
D0 (40 pb-1) tB1.526?0.034?0.035 ps CDF (18
pb-1) Both consistent with PDG tB 1.564 ?
0.014
33
Sin(2?) in B0?J/y Ks
?ms/?md
a
g
b
N(B0)(t) - N(B0)(t)
ACP(t)
D?sin(2b)?sin(Dmd t)
N(B0)(t) N(B0)(t)
In Run1 measured
B0 ? J/? Ks J/? ? ??
sin(2b)0.790.390.16
(400 events)
sin(2b)0.910.320.18
(60 B0 ? ? (2S) Ks)
With 2fb-1 can refine this measurement Although
no way to compete with B-Factories !

• N(J/? Ks) from scaling Run I data
• x 20 luminosity
  8,000
• x 1.25 tracks at L1 trigger 10,000
• x 2 muon acceptance 20,000
• Trigger on J/? ? ee? 10,000

Stat. Error
Expect s(sin2b) ? 0.05
Systematic 0.5xStatistical (scales with control
sample statistics)
Combined eD2 from 6.3 to 9.1 (Kaon b-tag)
Same S/B 1
34
Tevatron Performance
3.8 x 1031

• Tevatron operations
• Startup slow, but progress steady !
• Now L 3.5 x 1031 cm-2s-1
• integrating 6. pb-1/week
• still factor 2-3 below planned values
• additional improvements (10-20) expected
  from Jan. 3weeks shutdown

Initial Luminosity
July 01
Now

• CDF operations
• Commissioning Summer 2001
• Physics data since February 2002
• Running with gt90 Silicon integrated
• since July 2002

On-tape Luminosity
110 pb -1

• Luminosity (on-tape)
• 20pb-1 until June (analyses in this talk)
• Additional 90pb-1 July December
• Reach 300- 400 pb-1 by October 2003

July 02
Feb 02
35
Quadrant of CDF II Tracker
TOF 100ps resolution, 2 sigma K/? separation for
tracks below 1.6 GeV/c (significant improvement
of Bs flavor tag effectiveness)
TIME OF FLIGHT

• COT large radius (1.4 m) Drift C.
• 96 layers, 100ns drift time
• Precise PT above 400 MeV/c
• Precise 3D tracking in ?lt1
• ?(1/PT) 0.1GeV 1 ?(hit)150?m
• dE/dx info provides 1 sigma K/? separation above
  2 GeV

• SVX-II ISL 6 (7) layers of double-side
  silicon (3cm lt R lt 30cm)
• Standalone 3D tracking up to ? 2
• Very good I.P. resolution 30?m (20 ?m with
  Layer00)

LAYER 00 1 layer of radiation-hard silicon at
very small radius (1.5 cm)
(achievable 45 fs proper time resolution in Bs ?
Ds p )
36
CDF II Trigger System
3 levels 5 MHz (pp rate) ? 50 Hz (disk/tape
storage rate) almost no dead
time (lt 10)

• XFT EXtremely Fast Tracker
• 2D COT track reconstruction at Level 1
• PT res. DpT/p2T 2 (GeV-1)
• azimuthal angle res. Df 8 mrad
• SVT Silicon Vertex Tracker
• precise 2D SiliconXFT tracking at Level 2
• impact parameter res. ?d 35 ?m
• Offline accuracy !!

CAL
MUON
CES
COT
SVX
XFT
XCES
Matched to L1 ele. and muons
XTRP
enhanced J/? samples
L1 CAL
L1 MUON
L1 TRACK
GLOBAL L1
SVT
L2 CAL
CDF II can trigger on secondary vertices !!
Select large B,D samples !!
GLOBAL LEVEL 2
TSI/CLK
37
SVT Triggering on impact parameters
150 VME boards

• Combines COT tracks (from XFT) with Silicon Hits
  (via pattern
• matching)
• Fits track parameters in the transverse plane
  (d, ?, PT) with offline res.
• All this in 15ms !
• Allows triggering on displaced impact
  parameters/vertices
• CDF becomes a beauty/charm factory

38
B triggers conventional
Need specialized triggers
?(bb) / ?(pp) ? 10-3

CDF Run1, lepton-based triggers

• Di-leptons (??, PT ? 2 GeV/c) B ? J/? X, J/? ?
  ??
• Single high PT lepton (? 8 GeV/c) B ? l ? D X

Suffer of low BR and not fully rec. final state
Nevertheless, many important measurements by CDF
1 B0d mixing, sin(2?), B lifetimes, Bc
observation,

• Now enhanced, thanks to XFT (precise tracking at
  L1)
• Reduced (2?1.5 GeV/c) and more effective PT
  thresholds
• Increased muon and electron coverage
• Also J/? ? ee

39
XFT performance
Efficiency curve XFT threshold at PT1.5
GeV/c ? 96.1 0.1 (L1 trigger)
Offline track
XFT track
11 pb-1

• XFT L1 trigger on tracks
• better than design resolution
• ?pT/p2T 1.65 (GeV-1)
• ?? 5.1 mrad

53.000 J/? ? ??
40
SVT performance

• I.P. resolution as planned
• ?d 48 ?m 35 ?m ? 33 ?m

intrinsic
D0 ? Kp used as online monitor of the hadronic
SVT triggers
transverse beam size

• Efficiency

S/B ? 1
90
soon
80
41
TOF performance

• TOF resolution (110ps) within 10 of design value

Background reduction in ? ?KK Low PT (lt 1.5
GeV/c) track pairs before and after a cut on TOF
kaon probability x20 bkg reduction, 80 signal
efficiency
with TOF PID
S/N 1/2.5
S/N 1/40
42
CDF J/y cross section
0ltptlt0.25 GeV
5.0ltptlt5.5 GeV
10.0ltptlt12.0 GeV
s(ppgJ/y pTgt0 hlt0.6) 240 ? 1 (stat)
?35/28(syst) nb
43
Lots of charm from hadronic triggers
With 10 pb-1 of hadronic trigger data
Relative Br. Fractions of Cabibbo suppressed D0
decays
Already competitive with CLEO2 results (10fb-1 _at_
?(4S)) !!!!!
?(D?KK)/?(D?K?) 11.17 ? 0.48(stat) ?
0.98 (syst) ?(D?? ? )/?(D?K?) 3.37 ?
0.20(stat) ? 0.16(syst)
O(107) fully reconstructed decays in 2fb-1

• ? Foresee a quite interesting charm physics
  program
• D cross sections,
• CP asymmetries and Mixing in D sector, Rare
  decays,

44
B0s mixing expectations with 2fb-1
xs ?ms??(B0s)
Bs ? Ds?, Ds ? ? ? Ds ? ??, KK, ???

• Signal 20K (fp only) - 75K (all) events
• with SVT hadronic trigger
• BR (Ds ?) 0.3 BR (Ds ? ? ?) 0.8
• Resolution
• ?(c?) 45 fs (with Layer00)
• eD2 11.3 (with TOF)
• S/B 0.5-2 (based on CDF I data)

S.M. allowed range 20. lt Xs lt 35.
5s sensitivity up to Xs 63 (S/B 2/1) Xs
53 (S/B 1/2)
Can do a precise measurement or evidence for
new physics !
45
Data Samples The J/y?mm- t CDF and D0 (Run-II)
75K at D0, completely new capability ! (40 pb-1)
0.5M at CDF (70 pb-1)
Two Fully Reconstructed B-hadron?J/y states at
CDF D0
B?J/y KS CDF220, D045 (Run-II) (D0 had none
in Run-I)
LB?J/yL CDF53, D016 (Run-II)
46
Data Samples B and Charm Using the high Impact
Parameter (IP) (Hadronic) trigger
Select events by requiring -2 tracks with
IPgt100 mm - track PT gt 2GeV/c - sum 2-track
PT gt 5.5 GeV/c
0.5M Charm decays at CDF ?10-20 come from B
Great Potential for B and Charm Physics, opens
at least as many avenues as J/y trigger
47
Data Samples B()?luD decays using hybrid
trigger
Select events with 1 lepton (PTgt3 GeV/c) 1 high
IP (gt120mm)track -High IP track means we can go
lower in lepton PT -gtMuch higher than Run-I due
to lower PT thresholds (x4-5 increase) Used
for 1) High statistics lifetime and mixing
analyses 2) calibration samples for tagging
(B?lnD) Drawback worse vertex
resolution due to missed neutrino Some
numbers BglD0X (D0gKp) 10000 events, BglDX
(DgKpp) 5,000 events also Bs decays (later)
48
Physics Results Lifetimes from partially
reconstructed decay
Decays included
Accounting for missed neutrino Bs ? Ds?l, Ds?l
(Ds?????, K0K?, ?????) expect 40K events in 2
fb-1 st is worse due to missed u (K factor) ?t
60 fs ? t ? ?K/K, ?K/K 14 If one Bs lifetime
is fit in any flavour specific mode tfit
(tBsCP2tBsCP-2)/(tBsCPtBsCP-) from which DGs
can be determined as well
Use new hybrid displaced tracksingle lepton
trigger
49
Physics Results LB, leptondisplaced track and
purely hadronic data samples (have shown J/y mode
already)
Protons are easiest to separate using Time of
Flight Particle ID in left plot using TOF and
dE/dX
?b ? ?cl? ? pK? l?
Lifetime in hadronic, hadronlepton modes
require correction for IP cut bias missing ?
Expect results after this summer
?b ? ?cp ? pK? p
Note on LB A search for CP violation in Baryon
decays is planned using LB?pp
50
Mixing and CP violation (CPV) at Hadron colliders
Proof of principle
Run-I, CDF were able to do 2s measurement of
sin2b competitive xd (Dmd/G) measurements can
tag b-flavours in hadron collider environment
Sin2b0.790.39(stat)0.16(sys) (CDF 1996) CDF
have not repeated this measurement yetcannot
compare to B-factories
CDF In Run-II with 40-50 x more Bd?J/yKS decays
can get d(sin2b)0.05 D0 Similar statistics
Cant be competitive with BaBar (insert
current) and BELLE (insert current) Redo the
measurement because -Its an important benchmark
-Gives credence to other CPV measurements eg.
in B?hh- Bs?J/yf
51
Physics Results Charm physics at CDF Search for
CP violation (CPV) in Charm decays
1) c and u quarks dont couple to t ?box diagram
contributions are tiny 2) CPV in charm decays ?
due to interference in decay (direct CPV) 3) SM
prediction O(0.1-1) CP violation effects in
Charm Decays
How Compare rate of Decay of D0, D0 to CP
eigenstates fKK- and pp-

• Method Using data from Hadronic Trigger
• -Collect D?D0p sign of p tags flavour of D
• -Search for D0? KK-, D0? pp-, D0 ? pp- D0?
  KK-
• Correct for tracking efficiency for vs - p from
  D?D0p
• -Count number of decays in each mode after
  corrections

52
CPV in charm decays
Cross-check Measure Ratio of Branching Ratios
_at_CDF G(D0? pp-)/G(D0? Kp-)9.380.180.10
G(D0? KK-)/G(D0? Kp-)3.686 0.076
0.036 FOCUS G(D0? pp-)/G(D0?
Kp-)9.930.140.14 G(D0? KK-)/G(D0?
Kp-)3.530.120.06 CDF accuracy is comparable
and consistent with FOCUS (2003) and World
average 2.880.15 (PDG)
93560 D?D0p with D0?Kp-
8320 D?D0p, D0? KK-
CLEO Result (2001)
First CPV measurement at CDF in Run-II
ACP(D0? (pp-)) 0.02.20.8 ACP(D0? (KK-))
1.9 3.20.8
ACP(D0? (pp-))2.01.70.6 (PDG
0.51.6) ACP(D0? (KK-))3.01.90.6(PDG
2.12.6)
53
Physics Results Search for Flavour Changing
Neutral Current decay D0?mm-
SM predicts a branching ratio (BR) of O(10-13)
for D0?mm-
Some R-parity violating SUSY models predict
branching ratios upto O(10-6)

• Technique
• D0?pp- BR is well known identical acceptance
• to D0 ?mm-
• 2) Use D0? D0p to tag D0 in D0?K-p (thus no
  K vs p ambiguity)
• 3) See how many ps fake ms per PT
• 5) Look for D0 ?mm- in same sample
• 6) Subtract D0?pp- faking D0?mm-
• 0 events found in 2s search
• window

CDF Result BR(D0?mm) ? 2.4x10-6 better than
most recent world average ( PDG 90CL lt 4.1 x
10-6 )
54
Physics Results,Testing HQE A summary of results
HQE Predicted B Lifetime hierarchy tBc ltlt tXb0
tLb lt tBd tBs lt t B- lt t Xb-

• D0 (240 pb-1) t(B)/t(B0) 1.0930.021 0.022
  ps (from semi-leptonics)
• CDF (240 pb-1) t(B ) 1.66?0.03?0.01 ps, t(B0
  )1.54 ?0.05?0.01 ps
• t(B)/t(Bd) 1.080 ? 0.042 (B?J/yK
  Bd?J/yK0)
• t(Bs)/t(Bd) 0.89 ?0.072(Bs ?J/yf)
• t(LB) 1.250.260.10 ps (?b?J/y?)
• CDF Mass Measurements
• M(Bs) 5366.01 ? 0.73?0.33 MeV/c2
  Worlds best measurements
• M(LB) 5366.01 ? 0.73?1.2 MeV/c2 of Bs
  LB masses

BELLE (PRL 88 171801 2002) using
Bd?D()-(p,r), J/yKS,J/yK0 and B?D0p,
J/yK t(B)/t(Bd) 1.0910.0230.014
BABAR fully reconstructed decays Bd?D()-(p,r,
a1), J/yKS,J/yK0 and B ?D0p,
J/yK t(B)/t(Bd) 1.0820.0260.012 BABAR
partially reconstructed decays(B?D,D l
n) t(B)/t(Bd) 1.064 0.031 0.026
55
Physics Results Charm physics at CDF Search for
CP violation (CPV) in Charm decays
1) c and u dont couple to t ?box diagram
contributions are tiny 2) CPV in charm decays ?
due to direct CPV SMO(0.1-1) CPV, good test of
SM !
How Compare N(D0), N(D0)to CP eigenstates KK-
pp-
Data from SVT 1) Find D?D0p sign of p tags
flavour of D, 2) Find D0? KK-, D0? pp-, D0 ?
pp- D0? KK-
1) Cross-checkRatios of BRs (_at_CDF) G(D0? pp
)/G(D0?Kp)9.380.180.10 G(D0? KK-)/G(D0?
Kp-)3.686 0.076 0.036 FOCUS 9.930.140.14
3.530.120.06 CDF consistent with FOCUS PDG
2.880.15
First CPV result at CDF in Run-II
AD0? pp-2.01.70.6 AD0?KK- 3.01.90.6
CLEO Result (2001) PDG
AD0?pp- 0.02.20.8 AD0?KK- 1.9
3.20.8 (0.51.6 2.12.6)
56
Physics Prospects CP violation in B?hh- decays
determining angle g (CDF)
Monte-Carlo Bd ?pp-, Bs ?KK- Bs ?K?p?, Bd
?K?p? (From Monte-Carlo)-all pile up
B?hh- from hadronic trigger Includes B ?pp-,Bs
?KK- Bs ?K?p?, and Bd ?K?p?
Must disentangle each mode from signal We (will)
use -dE/dx based K and p ID -Kinematical
variable M?? vs a(1-p1/p2)?q1 -Width of
signal -Frequency of oscillation in CP asymmetry
57
Physics Prospects CP violation in B?hh- decays
determining angle g (CDF), Method
Five observables,
Tree and penguin graphs for B?pp- Bs ?KK-
Tree gt penguin in B?pp- vice-versa in Bs?KK-
-CP Asymmetry in B?pp- Sin2(gb) (without
penguin) -CP Asymmetry in Bs ?KK- Sin2g
(without penguin) -Assume SU(3) symmetry
replace s?d Hadronic matrix element ratios
penguin/tree same for both modes
Four unknowns dratio of penguin/tree hadronic
matrix elements q phase of d g,b weak phases
Constrain Sin2b from B-factories, CDF/D0
results and measure g
Proposed by R.Fleischer, PLB459 1999 306
58
Physics Prospects CP violation in B?hh- decays
determining angle g (CDF)
Mpp vs a for each B?hh- mode
Yields (Results from 65 pb-1) Bd?pp-
148?17 Bd?K?p? 39?14 Bs?K?p?
3?11 Bs?KK- 90?17 (Bs?KK- First
Observation !!)
Numbers from 65 pb-1 sample 1.16s dE/dX
separation Update from re-calibrated dE/dX (1.4s)
180 pb-1 in progress
Sanity check Measure Ratio of Branching
Ratios CDF G(Bd?pp-)/G(Bd ?Kp-) 0.26
0.110.055, PDG
Ratio of BRs along with ACP(Bd ?pp- ) from
B-factories Helps constrain g
Fleischer method Expect (2fb-1) ?(?)
10?(stat) 3?(syst SU(3) breaking)
59
Physics Results, Testing HQET lifetime, mass,
from fully reconstructed B decays modes,
Technique
Data from J/y?mm- di-muon trigger or High IP
trigger - Reconstruct vertex - Calculate
decay proper time, mass errors - Massfit mass
distribution only - Fitting for LifetimeFit mass
and lifetime distributions in single step


Probability Density Function and normalization
Technique applied to several decays B gJ/y K,
B0 gJ/y K0 (K0 g Kp), Bs g J/y f (f gKK)
LbgJ/yL (Lgpp)etc
-Those Decays selected using SVT trigger have
biased ct (Also leptonhigh IP track
data _at_CDF) -Fix bias and then measure Lifetimes
in Bs ?Dsp-, and other purely hadronic decays
High IP track selection efficiency
60
Physics Results,Testing HQELifetime, mass
summary
HQE Predicted B Lifetime hierarchy tBc ltlt tXb0
tLb lt tBd tBs lt t B- lt t Xb-

• D0 (240 pb-1) t(B)/t(B0) 1.0930.021 0.022
  (from semi-leptonics)
• CDF (240 pb-1) t(B ) 1.66?0.03?0.01 ps, t(B0
  )1.54 ?0.05?0.01 ps
• t(B)/t(Bd) 1.080 ? 0.042 (B?J/yK
  Bd?J/yK0)
• t(Bs)/t(Bd) 0.89 ?0.072(Bs ?J/yf)
• t(LB) 1.250.260.10 ps (?b?J/y?)
• CDF Mass Measurements
• M(Bs) 5366.01 ? 0.73?0.33 MeV/c2
  Worlds best measurements
• M(LB) 5366.01 ? 0.73?1.2 MeV/c2 of Bs
  LB masses

BABAR exclusive decays BABAR
inclusive decays t(B)/t(Bd) 1.0820.0260.012
t(B)/t(Bd) 1.064 0.031 0.026 BELLE
t(B)/t(Bd) 1.0910.0230.014
-Projection s(t(Bs)/t(Bd) ) s(t(Bd)/t(Lb)) lt1
at 2fb-1 -Current s(t(Bu)/t(Bd0) ) surpasses
theoretical accuracy -Measurements test vertexing
tracking Crucial for DMBs and CPV
61
Physics Results Rare B decays B s(d)?mm-
-No observed excess -Expected backgrounds
(events) 1.05?0.3 (Bs) and 1.07 ?0.31
(Bd) -Observed 1 event for both modes ?
branching ratio limit is possible -SM Prediction
BR10-9
BR limits vs. luminosity
BR Upper Limit at 95 CL 7.5x10-7 (Bs ?
mm-) 1.9x10-7 (Bd ? mm-) BR Upper Limit at 90
CL 5.8x10-7 (Bs ? mm-) 1.5x10-7 (Bd ? mm-)
Submitted to PRL
Bs result surpasses previous worlds best result
(by x2 CDF)
Bd result bit better than Belle (1.6x10-7) and
BaBar (2.0x10-7)
62
Rare Decays Bs?ff branching ratio
1) Approach calculate branching ratio by using
N(Bs J/y ? f) in the same data (SVT triggered)
sample
1) All BRs are taken from PDG 2) Efficiencies
calculated from MC 3) N(Bs ? J/y f) is corrected
for a) Reflections from Bd? J/y K(892)
b) Requirement of a muon match (check of signal
in SVT data)

• Calculate Branching ratio using corrected
  Ncorr(Bs J/y f)

SM Prediction for branching ratio 3.7x10-5
hep-ph/0309136 BR (1.4 0.6 (stat) 0.2(syst)
0.5 (BR))x10-5 Upper Limit BR (1.4 0.6
(stat) 0.2(syst) 0.5 (BR))x10-5
63
Physics Results Hadronic Moments
Calculating D, D are known ,
measure only f, contains wide and narrow D0
and non-resonant part. Reconstruct only B-? D0
l- ? (p0s not possibe _at_ CDF). Find m and D0
consistent with coming from B (vertex), Masslt5.3
GeV D0 ?Dp, Dp-, D0p0, reconstruct or use
Isospin for Mass pdf
Using LeptonHigh IP data A) D0 ? D ?- B)
Cant do D0 ? D0 ?0 But 0.5A same shape
C) D0 ? D ?- D) D ? D0 ? E) D ? D
?0 Cant do Feed-down to D ?- is corrected
for. F) D0 ? D0 ?0 Cant do but 0.5.A (same
shape)
64
Physics Results Hadronic Moments analysis
D0 mass distribution.
.gives moments wrt D0 only
Moments from all D, D
Best single measurement Of M1 M2 in the world !
And finally from M1 M2 we get
65
Physics Prospects CP violation in B?hh- decays
determining angle g (CDF), Method
B?hh- from hadronic trigger
Tree gt penguin in B?pp- vice-versa in Bs?KK-
Four unknowns In Asymmetry(t) dratio of
penguin/tree hadronic matrix elements q phase
of d g,b weak phases Constrain Sin2b from
B-factories, CDF/D0 results and measure g by
fitting asymmetry Proposed by R.Fleischer,
PLB459 1999 306
dE/dx check Use D?D0p?, D0? K?p?
Problem Separating Bd ?pp-, Bd ?Kp-, Bs ?Kp-,
Bs ?KK-, Use M?? vs a(1-p1/p2)?q1, dE/dX, to
separate Bd ?pp-, Bd ?Kp-, Bs ?Kp-, Bs ?KK-,
in the future Will use oscillation frequencies as
well
66
Toward Bs Mixing Proof of principle CDF (Run-II)
DMd
CDF Run-I Dmd (all methods) 0.495 0.026
0.025 ps-1
First Run-II result Bd-Bd oscillations using
same side tagging (SST) Look for fragmentation p?
from B, track with lowest relative PT to B
-Use B?J/y K (J/y data) B?D0p (SVT data)
to tune tagging -Use B0?J/y K0(? K?p?)and
B0?D?p? to measure Dmd
Flavour Tagging -Look for fragmentation p?
from B -Calculate Ptrel variable -Want maximally
collinear B and p -Pick p candidate with lowest
Ptrel -B flavour is correlated with p sign
67
Physics Prospects CP violation in B?hh- decays
determining angle g (CDF), Method
B?hh- from hadronic trigger
Tree gt penguin in B?pp- vice-versa in Bs?KK-
Four unknowns In Asymmetry(t) dratio of
penguin/tree hadronic matrix elements q phase
of d g,b weak phases Constrain Sin2b from
B-factories, CDF/D0 results and measure g by
fitting asymmetry Proposed by R.Fleischer,
PLB459 1999 306
dE/dx check Use D?D0p?, D0? K?p?
Problem Separating Bd ?pp-, Bd ?Kp-, Bs ?Kp-,
Bs ?KK-, Use M?? vs a(1-p1/p2)?q1, dE/dX, to
separate Bd ?pp-, Bd ?Kp-, Bs ?Kp-, Bs ?KK-,
in the future Will use oscillation frequencies as
well
68
Physics Results Testing HQE Lifetimes from
partially reconstructed decays
Data are selected using high PT leptons (D0)
leptonhigh IP track (CDF)
Examples of decays
Accounting for missed neutrino Bs ? Ds-l, Ds-l
(Ds???, K0K, ????) Bu- ? D0l-n, D0l- nX (D0
?pK-) Advantage Very high statistics Drawback
st is worse due to missed u (K factor) However
Large numbers provide opportunities for lifetime
mixing
D0 Charged to Neutral B-Meson Lifetime Ratio
t/t0 -B ? mnD(2010)-X decays mostly Bd -B ?
mnD0X decays mostly Bu -Calculate ratio of
events/lifetime bin -Account for all decays BRs
(PDG) -ratio of events expected N/N0
e-(t/t0-1)t ?(K-factor)?st D0 Result t/t0
1.093 0.021 (stat) 0.022 (syst) Competitive
with worlds best results