Top Quark Properties and Search for Single Top Quark at the Tevatron

1
Top Quark Propertiesand Search for Single Top
Quarkat the Tevatron

• Meenakshi Narain
• Boston University
• Presented at EPS 2005

2
Top Quark at the Tevatron

• Top quark discovered a decade ago
• (in 1995)!
• Run I (1992-1996)
• ?s 1.8 TeV
• Integrated luminosity
• 120 pb-1
• Run II (2001-present)
• ?s 1.96 TeV
• 3 fold increase performance since June03
• Integrated luminosity by June 05
• Delivered gt1fb-1
• On tape 800pb-1
• Analyzed up to 350 pb-1

Worlds only top factory!
3
Top Quark Physics

• Top is very massive
• It probes physics at much higher energy scale
  than the other fermions.
• Top decays before hadronizing
• momentum and spin information is passed to its
  decay products.
• No hadron spectroscopy.
• Top mass constrains the Higgs mass
• Mtop, enters as a parameter in the
• calculation of radiative corrections to
  other
• Standard Model observables
• it is also related, along with the mass of
• the W boson, to the that of the Higgs boson.

Mtop (world average) 172.7 ? 2.9 GeV
?top 10-24 sec
4
The Top Properties Tour
W helicity
Top Charge
Top Width
CP Violation
Top Mass
Top Spin
Anomalous Couplings
Production Kinematics
Production X-Section
Top Spin Polarization
Resonance Production
Y
Rare/non SM decays
Branching Fractions
Vtb
5
Top Quark Decay Properties

• Does top quark decay 100 of the times to Wb?
• B(t? Wb)
• Search for exotic decay modes of the top quark
• t? Hb
• Properties of the W-t-b vertex
• W Helicity
• Top quark Charge

6
Is B(t? Wb) 100?

• Within the SM, assuming unitarity of the CKM
  matrix, B(t?Wb)1.
• An observation of a B(t?Wb) significantly
  different than unity would be a clear indication
  of new physics
• non-SM top decay, non-SM background to top
  decay, fourth fermion generation,..

7
Measurement of B(t?Wb)/B(t?Wq)

• B(t?Wb) can be accessed directly in single top
  production.
• Top decays give access to B(t?Wb)/B(t?Wq)
• R can be measured by comparing the number of
  ttbar candidates with 0, 1 and 2 jets tagged.
• In the 0-tag bin, a discriminant variable
  exploiting the differences in event kinematics
  between ttbar and background is used.

In the SM
Leptonjets (230 pb-1)
Leptonjets and dilepton (160 pb-1)
DØ Run II Preliminary
hep-ex/0505091
Results consistent with the SM prediction
8
Exotic Decays of the top quark

• Since R is about 1
• Top quark decays to a b-quark ? t? Xb
• Is X W ?
• OR
• could X H ?.
• as predicted by generic 2Higgs Doublet Models?

9
Search for t?Hb

• If MHltmt-mb
• then t? Hb competes with t? Wb
• results in B(t?Wb)lt1.
• H? decays are different than W decays
• affect ??(tt) measurements in different channels
  (dileptons, leptonjets, leptontau).
• Perform simultaneous fit
• to the observation in all channels and
• determine model-dependent exclusion region in
  (tan?, MH).

10
Are the other Properties of the Top Quark as
Expected?W-t-b Vertex W helicityTop Charge
11
W helicity in Top quark Decays

• Large top quark mass
• Are there new interactions at energy scales near
  EWSB?
• helicity of the W boson
• examines the nature of the tbW vertex
• provides a stringent test of Standard Model

V-A coupling
gWtb ? Vtb (V-A)
F0 0.7 F- 0.3 F0
V-A SUPPRESSED
12
W helicity

• In the Standard Model (with mb0)
• The PT of the lepton has information about the
  helicity of the W boson
• longitudinal leptons are emitted perpendicular
  to the W (harder lepton PT)
• left-handed leptons are emitted opposite to W
  boson (softer lepton PT)

SM F- 0.3 F0 0.7 F0
Left-handed
Longitudinal
Right-handed
13
W Helicity

• Likelihood analysis of PT spectrum
• Consider dilepton channels
• Fix F00.7, measure F (F-1-F0-F)
• Binned likelihood and estimate F using Bayesian
  method

• Likelihood analysis of cos ?
• Consider leptonjets channels
• Fix F00.7, measure F (F-1-F0-F)
• Two analysis topological and b-tag

Leptonjets (?1 b-tag)
Results consistent with the SM prediction
F00.7, F0
14
Measurement of top quark Charge

• Is it the Standard Model top ?
• OR
• An exotic doublet of quarks (Q1, Q4)
• with charges (-1/3,-4/3) and M 175 GeV/c2
• while M(top) 274 GeV/c2
• W.-F. Chang et al.,hep-ph/9810531
• q -4/3 is consistent with EW data,
• new b-couplings improve the EW fit
• (E. Ma et al. , hep-ph/9909537)

15
Top Quark Charge Measurement

• Goal discriminate between
• Qtop 2e/3 and Qtop 4e/3
• Top quark charge is given by the sum of the
  charge of its decay products
• Determine
• Charge of W (lepton)
• Charge of b-jet Qjet ??qi pTia/ ? pTia
• (here, a0.6)
• Associate b-jets to correct W (charged lepton)
• The charge of the quark is correlated with the
  charge of the highest pT hadron during
  hadronization

16
Top Quark Charge

• We need an observable and an expectation for the
  2/3 and 4/3 scenarios
• Consider only leptonjets double-tagged events
• Two top quarks in the event ? measure the charge
  twice
• The exotic scenario is obtained by permuting the
  charge of the tagged jets
• qb and qB are taken from the data derived jet
  charge templates
• Results coming soon...

qB
qb
qB
qB
qb
17
Top Quark Production Properties

• Since top decay properties look quite consistent
  with SM predictions.
• What about its production?
• Could it be a t-prime?
• Search for tt production (t ?Wq)
• Could the ttbar pair originate from the decay of
  a resonance?
• Model independent search for narrowresonance X?tt
  used to exclude a leptophobic X boson
• What about single top production?

Run I search of X with G1.2M MXgt560 GeV _at_ 95
CL (DØ) and MXgt480 GeV _at_ 95 CL (CDF)
18
Search for Single Top Quark
19
Search for Single Top

• Electroweak Production of top quark
• Measure production cross sections
• Direct measurement of Vtb (s ? Vtb2)
• Top spin physics (100 polarized top quark)
• s- and t-channels sensitive to different New
  Physics
• Irreducible background to associated Higgs
  production
• Exotic Models (FCNC, Top Flavor, 4th Gen)

20
Single Top Status
t-channel
s-channel
d
q
t
u
W
W
?q'
?b
b
t

• Cross sections
  s-channel
  t-channel st
• NLO calculation 0.88pb (8)
  1.98pb (11)
• Run I 95 CL limits, DØ lt 17pb
  lt 22pb
  CDF lt 18pb lt 13pb
  lt 14pb
• Run II CDF 95 CL limits lt 14pb
  lt 10pb lt 18pb
• Other Standard Model production mode (Wt)
  negligible

21
Signature Backgrounds
Signal for s and t channel mostly similar

• Lepton Missing ET Jets
• t-channel extra b tends to be forward
• Similar to top pair production, but with less jets

Harder Signal To Find
(t-channel)
Backgrounds

• W/Z jets Production
• Fake Leptons
• Top Pair Production
• WW, WZ, Ztt, etc.

Much worse than for pair production because of
lower jet multiplicity
Anything with a lepton jets ET signature
22
Discriminating Variables

• Object kinematics
• Jet pT for different jets
• Tagged, untagged,...
• Event kinematics
• H (total energy)
• HT (transverse energy)
• M (invariant mass)
• MT (transverse mass)
• Summing over various objects in the event
• Angular variables
• Jet-jet separation
• Jet pseudorapidity (t-channel)
• Top quark spin

23
Separating Signal from Backgrounds

• Four analysis methods
• Three (Cut, NN, DT) use the same structure
• Optimize separately for s-channel and t-channel
• Optimize separately for electron and muon channel
  (same variables)
• Focus on dominant backgrounds Wjets, tt
• Wjets train on tb-Wbb and tqb-Wbb
• tt train on tb tt ? l jets and tqb
  tt ? l jets
• Based on same set of discriminating variables
• 8 separate sets of cuts/networks/trees

Likelihood Discriminant
Cut-Based
Neural Networks
Decision Trees
24
1. Cut-Based Analysis

• Cuts on sensitive variables to isolate single top
• Separate optimizations for s-channel and
  t-channel
• Loose cuts on energy-related variables pT
  (jet1tagged) H(alljets jet1tagged)H(alljets
  jet1best)HT (alljets)M(toptagged)M(alljets)M(a
  lljets jet1tagged)?s

Factor 2 improvement!
25
2. Neural Network Analysis
Input Nodes One for each variable xi
full dataset
electron
muon
1 b-tag
?2 b-tags
1 b-tag
?2 b-tags
Output Node linear combination of hidden nodes
constructnetworks
Hidden Nodes Sigmoid dependent on the input
variables
2d histograms, Wbb vs tt filter
26
Result

• No evidence for single top signal
• Set 95 CL upper cross section limit
• Using Bayesian approach
• Combine all analysis channels (e, m, 1 tag, ?2
  tags)
• Take systematics andcorrelations into account

Systematic uncertainty
Expected limit set Nobs to background yield
Expected/Observed limit
ss lt 9.8 / 10.6 pb
st lt 12.4 / 11.3 pb
27
Neural Network Output
em ?1 tag
em ?1 tag
em ?1 tag
em ?1 tag
28
Result

• No evidence for single top signal
• Set 95 CL upper cross section limit
• Using Bayesian approach and binned likelihood
• Built from 2-d histogram of Wbb NN vs tt NN
• Including bin-by-bin systematics and correlations

Expected/Observed limit
ss lt 4.5 / 6.4 pb
st lt 5.8 / 5.0 pb
29
3. Decision Tree Analysis
HTgt212

• Replace Neural Networks by Decision Trees
• single tree, 100 nodes
• Remaining analysis steps identical
• Same inputs
• Same filter configuration
• Binned likelihood analysis

Fail
Pass
ptlt31.6
Mtlt352
purity
Expected/Observed limit
ss lt 4.5 / 8.3 pb
st lt 6.4 / 8.1 pb

• Sensitivity comparable to Neural Network analysis

30
4. Likelihood Discriminant Analysis

• New Analysis based on 370pb-1 dataset
• Different btagging algorithm and selection
• Likelihood Discriminant
• Input Variables

31
Result

Expected/Observed limit
ss lt 3.3 / 5.0 pb
st lt 4.3 / 4.4 pb
Best Limit !!!!
Comparison of LH with NN analysis
32
Single Top Summary

• Cross sections
  s-channel
  t-channel st
• NLO calculation 0.88pb (8)
  1.98pb (11)
• Run I 95 CL limits, DØ lt 17pb
  lt 22pb
  CDF lt 18pb lt 13pb
  lt 14pb
• Run II CDF 95 CL limits lt 14pb
  lt 10pb lt 18pb
• RunII DØ 95 Cl Limits
• (230 pb-1)
• Cut Based lt 10.6pb
  lt 11.3pb
• Decision Tree lt 8.3pb
  lt 8.1pb
• Neural Network lt 6.4pb
  lt 5.0pb
• (370 pb-1) (new analysis)
• Likelihood Discriminant lt 5.0pb
  lt 4.4pb

Accepted for publication, hep-ex/0505063
33
Sensitivity to non-SM Single Top
using only muon channel data
using only electron channel data
34
Conclusion

• Measurements of various top quark properties are
  underway and will improve with larger data sets
• The single top cross section limits and
  sensitivity of the analyses are getting to a
  level where we can expect to observe single top
  quark production soon!.
• Stay Tuned.