Top Quark Production at Hadron Colliders

1
Top Quark Production at Hadron Colliders
Fermi National Laboratory Wilson Fellow
Seminar February 15, 2006

• Joao Guimaraes da Costa
• Harvard University

2
How many top quarks are we making?
Top Mass
Higgs Search
Top Cross Section
Top Properties
Event Kinematics (New Physics?)
The Top quark cross section measurement is the
cornerstone to all analysis in the top sample
3
The Top Quark
Fermions
Top discovered in 1995
mtop 172.7 ? 2.9 GeV/c2
1989-1995 110 pb-1 (Run I) A few dozen
events
Mass

• Lingering Questions
• Any New Physics in the top quark sample?
• Is the event excess ONLY from top production?
• What are the top quark properties?
• Top mass, charge, spin, lifetime, branching
  fractions

Force Carriers
4
Top in radiative corrections
Electromagnetic constant measured in atomic
transitions, ee- machines
Weinberg angle measured at LEP/SLC
Fermi constant measured in muon decay
Radiative corrections
5
Top Factories
Tevatron highest energy collider today
LHC highest energy collider starting 2007
1.96 TeV
14 TeV
X 7
1-4 x 10-32
X 30-50
10-23 ? 10-34
Today 318 pb-1 ( 3x run I dataset)
New Results Soon 1 fb-1
6
Top Quark Production at Hadron Colliders
Proton
Hard scattering cross-section
7
Parton Distribution Functions
Parton distribution function xF(x,?2)
Effective center of mass energy
LHC
Gluon parton distribution function diverges at
low x
Tevatron
8
Top Quark Pair Production
LHC
Tevatron
85
10
15
90
9
Theoretical Cross Section

• Theoretical uncertainty ?15
• PDFs
• Renormalization/factorization scale

Mtop 175 GeV/c2
NLO
? 833 pb
Rate 700000 tt/day
120x
? 6.7 pb
Rate 60 tt/day
Cacciari et al. JHEP 0404068 (2004) Kidonakis
Vogt PRD 68 114014 (2003)
10
Single Top Production
Wt-channel
t-channel
s-channel
TeV (vs1.96 TeV) 0.88 0.11
pb LHC(vs 14 TeV) 10.6 1.1 pb
1.98 0.25 pb 246.6 11.8 pb
lt0.1 pb 62.016.6-3.6 pb
Tevatron Limit D0 lt 5.0 pb
lt 4.4 pb
Harris PRD 66 (02) 054024 Cao hep-ph/0409040
Campbell PRD 70 (04) 094012
Tait PRD 61 (00) 034001 Belyaev PRD 63 (01)
034012 Campbell hep-ph/0506289
11
Top Quark Decay
? - decay
Mtop 175 GeV
Virtual W
Br(t ? Wb) 100 in the SM
?top ? 1 GeV ?top ? 10-25 seconds c?top lt 52.5
?m _at_ 95 C.L. (CDF)
12
Top Quark Decay
(2 x)
Br(t ? Wb) 100 in the SM

• All-hadronic
• BR largest (45) but large QCD bkg
• 6 jets
• Dilepton
• BR small (5) and little kinematic constraints
• 2 jets 2 lep ETmiss
• LeptonJets
• BR larger (30)
• 4 jets 1 lep ETmiss

Always 2 jets are b-quark jets
13
Top Event Display from CDF Tagging b-quarks
b-quarks have long lifetime (c? 460 ?m)
d ??c?
B mesons travel few mm before decaying Displaced
Vertex Tagging
b ? cl? (BR ? 20) c ? sl? (BR ? 20)
Soft Lepton Tag
4x lower efficiency 7x larger fake rate
14
The CDF Run II Detector

• Silicon detectors
• L00, SVX, ISL
• hlt2
• Central Outer Tracker
• hlt1
• spT 0.15 pT2
• EM Calorimeter
• sE/E 14/E1/2
• Had Calorimeter
• sE/E 80/E1/2
• Muon Chambers
• hlt1.5

15
Top Cross Section Measurements
CDF s(tt) results
D0 s(tt) results
16
Event Selection
Golden Channel (lepton jets) (30)
MT(W) gt 20 GeV

• HT gt 200 GeV
• 1 SecVtx b-tag
• 2 SecVtx b-tags

• One high PT electron or muon
• ET or PT gt 20 GeV
• Lepton isolated from jets
• Large missing energy from neutrino
• METgt20 GeV
• 3 high ETgt15 GeV jets

or
17
Cross Section Measurement

• Nobs Number of observed events
• Nbkg Number of expected background events
• ?pretag Efficiency before tagging (incl Acc
  and BR)
• ?pretag 7.5
• ?btag Event tagging efficiency
• s L dt Integrated luminosity

18
B-tagging (Secondary Vertex)
Start with set of tracks and general location of
luminous region
JET
30 ?m
19
The luminous region at CDF (2002 to 2005)
z 0
Beam pipe
z direction along beampine (proton direction)
20
B-tagging (Secondary Vertex)
Start with set of tracks and general location of
luminous region
JET
Find primary vertex
Improved b-tagging efficiency 20
Primary vertex
10 ?m
30 ?m
21
B-tagging (Secondary Vertex)
Start with set of tracks and general location of
luminous region
JET
Find primary vertex
Improved b-tagging efficiency 20
Select tracks with large impact parameter inside
jet
Make a seed for secondary vertex and form vertex
Iterate removing tracks with worst chi2
Primary vertex
d0
22
B-tagging (Secondary Vertex)
Start with set of tracks and general location of
luminous region
JET
Find primary vertex
Improved b-tagging efficiency 20
Select tracks with large impact parameter inside
jet
Secondary vertex
Make a seed for secondary vertex and form vertex
L2D
Iterate removing tracks with worst chi2
Primary vertex
d0
Got a vertex! Check if L2D is large enough
23
Updated B-tagging (Tight SecVtx)
Loosen initial track selection More tracks from B
decays
JET
Secondary vertex
L2D
Primary vertex
d0
Tighten secondary vertex quality cuts
24
Fake rate Tags in light flavor jets

• Additional mistags
• Interactions in material
• Long lived particles Ks, ?

• Mistags
• Negative tag rate from dijet data parametrized in
  jet ET, ?, ?, track multiplicity and event ? ET
• Corrections for HF in jet data, material
  interactions, Ks and ?

25
B-Tagging Efficiency Determination MC Scale
Factor
Use inclusive low-pt (8 GeV) electron/muon sample
Fb 25
Tag away jet
Fb 70
Tag electron jet
Improvement from 82 earlier in Run II
Scale Factor Data/MC 91 /- 6

• Concern
• Extrapolate to average jet ET of top events
  ( 50 GeV)

26
Multiple Tagging and Loose SecVtx Tags
New in Run II

• Multiple tagging
• Larger datasets ? full reconstruction of events
• More precise measurement of top mass
• Reconstruction of W hadronic decays
• (4-jet top events)
• New window into new physics
• Higgs (H ? bb)
• Supersymmetry
• (sbottom, stop, charged higgs)

4 b-jets
Loose SecVtx algorithm optimized for double
tagged analysis
Approach Loosen input track quality selection
selection Prove of principle for future
improvements to b-tagging
27
Loose SecVtx Tagging Efficiency
Quark Efficiency () Efficiency ()
Quark Tight SecVtx Loose SecVtx
b 40.3 ? 0.3 47.6 ? 0.3
c 8.7 ? 0.2 12.6 ? 0.2
Mistags() Mistags() Mistags()
0.478 ? 0.003 1.195 ? 0.004

• 20 b-jet efficiency increase
• 2.5 x fake rate increase

Effective luminosity increase 45
28
Top Cross Section Tight / Loose Tagger
Top Tagging Efficiency/Evt () Top Tagging Efficiency/Evt () Top Tagging Efficiency/Evt () Top Tagging Efficiency/Evt ()
PRD Tight Loose Gain
? 1-tag 53 ? 3 60 ? 3 69 ? 4 30
? 2-tag 11 ? 2 16 ? 4 23 ? 4 210
300 gain since start of run II
Efficiency corrected with data/MC scale factor
29
Backgrounds

• Major backgrounds
• Wbb,Wcc,Wc
• Estimated from MC
• ALPGEN HERWIG
• Normalized to pretag data
• W light jet fake b
• Estimated from data
• Non-W
• Estimated from data
• Single top, WW, WZ, Z? ? ?
• Small
• Estimated from MC
• Pythia

30
Top Cross Section (Single Tag)
Sample Events tt fraction ?(tt )
Tight Tagger 138 81
Loose Tagger 174 73
Loose Tagger
Higgs production
31
Sample Composition Single b-tag
32
Top Cross Section (Double Tags)
Sample Events tt fraction ?(tt )
Tight Tagger 33 91
Loose Tagger 54 92
Loose Tagger
Measured cross section
Expectation for mtop 178 GeV
Signal region
33
Sample Composition Double b-tags
34
Major Systematic Uncertainties
Systematic Source () gt 1 tag gt 2 tags
b-tag efficiency 6.6 13.2
Backgrounds 5.3 5.0
Jet Energy Scale 4.3 4.3
Lepton Isolation 5.0 5.0
Luminosity 6.0 6.0
Total 11.4 17.2
Largest systematic

Small improvements possible
LHC Use good prediction from NNLO and higher
rate of W and Z to monitor luminosity
35
Cross Section Mass Dependence
This analysis 16 relative error
Recall Run I ? 5.1 ? 1.5 pb (30 relative
error) Early Run II (before b-tagging
improvements) ? 5.6 ? 1.5 pb (27 relative
error)
mtop 172.7 ? 2.9 GeV/c2
36
TOP MASS MEASUREMENT
37
Top Mass Measurements Summary
38
Top Mass Reconstruction
Lepton Jets
Kinematical Fit
PT balance
mt1 mt2
ml? mW
mjj mW
4 Jets
B-tags Jet-parton combinations
0 12
1 6
2 2
x 2 neutrino Pz solutions
39
Top Mass Reconstruction
Herwig MC Mtop 175 GeV
Resolution of reconstructed mass is dominated by
incorrect combinations
40
Top Mass Reconstruction
Lepton Jets
Kinematical Fit
PT balance
mt1 mt2
ml? mW
mjj mW
4 Jets
B-tags Jet-parton combinations
0 12
1 6
2 2
x 2 neutrino Pz solutions
41
W Mass Reconstruction JES
Lepton Jets
mjj mW
4 Jets
B-tags Jet-parton combinations
0 6
1 3
2 1
42
W Reconstruction Effect from b-Tagging
Mtop 175 GeV, JES nominal
Resolution of reconstructed mass is dominated by
incorrect combinations
Measure JES Mjj templates by varying JES 1 ?
43
Mass Templates

• Minimize ?2 to reconstruct event-by-event top
  mass
• Choose assignment with smallest ?2

44
Templates Sensitivity
45
CDF Ljets Template Method (3)
Top Mass with Jet Energy Scale Calibration
Likelihood fit for top mass and JES
318 pb-1
Mtop 173.5 2.7/-2.6 (stat.) ? 2.5 (JES) ? 1.3
(syst.) GeV/c2
compared to 3.1 GeV wo/ in situ calibration
46
Systematic Uncertainties
47
Statistical uncertainty
We were a bit lucky ..
48
Effect on Higgs Mass Expectations

• Using only CDF Run II Top Mass

? mtop 2.5
? mH 18
World Average mtop 172.7 ? 2.9 GeV/c2

• mtop lt 1.5 GeV/c2

Future Prospects

• mW lt 25 MeV/c2

49
Top Mass Measure Prospects (CDFD0)

• Total uncertainty of
• Dmtop lt 1.5 GeV/c2
• at the end of CDF
• Run II.

50
Physics at CMS
51
New Physics Measurements at LHC

• Standard Model Higgs
• Low mass region Htt ! bbbbWW
• Supersymmetric (SUSY) Higgs
• Charge Higgs (H) tt ! bH bW ! bbW? ?,
• tH ! btt ! bbbWW,
• tbH ! bbtt !
  bbbbWW
• Heavy Neutral Higgs (H/A) bbH/A ! bb??
• SUSY squarks
• Stop tt ! bb ? ? ! bbWW?0 ?0
• tt ! tt ?0 ?0 ! bbWW?0 ?0
• Sbottom gg ! bb bb ! bbbb ?0 ?0

Essential for the understanding of the nature of
new discoveries at LHC

• b-quarks in final state
• Backgrounds from tt ! bbWW production

Common to signatures
52
Searching for new physics with double-tags
Signature based analysis
gt 2 b-jets leptons missing energy

• Identification of high-PT b-quarks (b-tagging)
• Develop algorithms
• Measure efficiency
• Top cross section
• LHC is a top quark factory
• New physics to explore
• Light Higgs (Htt)
• SUSY Higgs (H, H/A)
• Third generation squarks

(8 million tt events in first year)
53
Light Higgs Search ttH ? ttbb
_
_
_
Likelihood based CMS study
MH lt 130 GeV/c2
4 b-jets

• Good b-tagging essential
• eb ? 60
• Backgrounds
• ttbb, ttjb, ttjj
• Reduced by reconstructing both top quarks
• Complementary to H? ??,??
• Best channel to measure H ? bb coupling

_
_
_
_
54
SUSY Charged Higgs boson
Can be discovered in many channels depending on
its mass relative to the top quark.

• m(H) lt m(t)
• t ! bH ! b? ?
• t ! bW ! bl ?
• Similar to SM tt decay.
• Could be seen in top cross section analysis.
• Backgrounds
• pp ! tt, Wjets, bb

• m(H) gt m(t)

2 b-jets
H ! tb
2 top quarks 1,2 b-quarks in final state !!!!
Mass up to 400 GeV/c2 can be discovered
55
SUSY third generation squarks

• Stop and sbottom are expected to be the lightest
  squarks
• Stop
• Sbottom

or
t t signature similar to top quark
Will be seen in top cross section analysis
4 b-jets

ATLAS discovery range m(b) lt 1 TeV
If found, probe SUSY nature by measuring
stop/sbottom properties
56
Summary
Large top quark sample are now available
Top Cross Section (single b-tag) Top Cross
Section (double b-tag)
(54 events) Top Mass
172.7 ?
2.9 GeV/c2
High precision measurements coming next!
1 fb-1 results in Summer 2006
The LHC is a top factory
Calibration tool Background Probe for new
physics
Top quark
57
Backup Slides
58
Increasing B-Tagging Efficiency from tuning

• Include event-by-event primary vertex (20)
• Re-optimize tagger parameters (15)
• Add L00 tracks (5)
• Overall 44 increase of efficiency

59
Measuring B-Tagging Mistag Rate

• Use negative L2D tags in jet data to predict
  mistags
• Parameterize negative tag rate in
• Jet ET
• Number of tracks in jet
• Jet h
• Jet f
• Event ? ET
• Correct for HF in negative tail and long-lived LF
  (aLF)

60
Controlling B-Tagging Mistag Rate
Beam Spot

• Reduced azimuthal asymmetry due to beamline
  offset
• Reduced tags due to conversions and material
  interactions by 50
• Maintained mistag rate 0.5 per jet

Center of SVX
61
B-Tagging Efficiency Determination MC Scale
Factor

• Measure b-tagging efficiency in data control
  sample
• Low ET Electron/Muon (8 GeV) dijet events
• Tag jet opposite to lepton to enhance heavy
  flavor fraction
• Tag lepton jet for b-tagging efficiency
  measurement
• Extrapolate average jet ET of top events ( 50
  GeV)
• Check extrapolation with jet data tag rates

Improvement from 82 earlier in Run II
Scale Factor Data/MC 91 /- 6
62
Top Mass with Jet Energy Scale Calibration

• Minimize c2 to reconstruct event-by-event top
  mass.

Fluctuate particle momenta according to detector
resolution.
Mtop as free param.
Constrain masses of 2 Ws.
t and t have the same mass.
Choose assignment with smallest c2.
63
The Power of Double Tagging
2 b-tags
0 b-tags
1 b-tag
Number of events on each b-tag category
Tagger 0 b-tags 1 b-tag 2 b-tags
Old 40 26 2
Improved 22 42 12