Search for Leptoquarks with the D

1
Search for Leptoquarkswith the DØ Detector at
Fermilab
Shaohua Fu Columbia University
OUTLINE

• Theory and Phenomenology
• Fermilab Tevatron
• DØ Detector
• Event Reconstruction
• Background and Signal
• Results and Conclusions

2
The Standard Model and Beyond

• The Standard Model has been very successful
• Elementary particles (fermions)
• six quarks and six leptons in three generations
• Fundamental forces SU(3)SU(2)U(1) group
• describes Strong, Electromagnetic and Weak
  forces
• force carriers are g, ?, W/Z bosons
• (Gravity is not in the framework of the SM)
• Spontaneous Symmetry Breaking mechanism gives
• masses to W/Z bosons and fermions
• (The Higgs boson has not been observed yet)
• However, the Standard Model has its weaknesses
• No explanation of the symmetry between the lepton
  and quark sectors
• Hierarchy problem, fine-tuning problem, dark
  matter, etc.
• Looking for clues of physics beyond the Standard
  Model!

Higgs ?
3
Leptoquarks

• Leptoquarks (LQ) are hypothetical particles which
  are predicted in many Standard Model extensions
  (e.g., the Grand Unified Theories)
• LQ are directly coupled to both Leptons and
  Quarks
• Carry color, fractional electric charge, lepton
  and baryon numbers
• Scalar (spin 0) or Vector (spin 1) vector LQ
  coupling is model dependent
• LQ appear in 3 generations corresponding to 3
  lepton/quark generations, but the
  intergenerational mixing is severely restricted
  by the FCNC constraints.
• At the Tevatron LQ pair production dominates

gluon fusion
quark anti-quark annihilation
4
Leptoquarks

• Leptoquark decays into a lepton (l? or ?) and a
  quark (q)
• LQ ? l?q or ?q We define ? ? Branching Ratio (LQ
  ? l?q)
• LQ pair decay into ll?qq, l??qq, ??qq final
  states with decay fractions of ?2, 2?(1??),
  (1??)2 respectively
• Signals are very energetic leptons and quarks
  (jets) in the detector
• Here I will describe the first generation LQ
  search in ee?qq channel in detail
• Previous search results
• No evidence of LQ has been observed, and the
  lower mass limit has been set
• (If LQ would exist, its mass gt mass limit at 95
  confidence level)
• 1st gen. LQ mass limit from DØ Run I Phys.
  Rev., D64, 092004 (2001)

? BR(LQ?eq) ? BR(LQ?eq) Scalar LQ Mass Limit (GeV/c2) Vector Minimal Coupling (GeV/c2) Vector Yang-Mills Coupling (GeV/c2)
D? 1 225 292 345
D? 0.5 204 282 337
D? 0 98 238 298
5
The Fermilab Tevatron

• Worlds highest energy collider
• 6.2 km in circumference
• Collides bunches of protons with bunches of
  anti-protons at two interaction points, where the
  DØ and CDF experiments sit
• Center of mass energy is 1.96 TeV (was 1.80 TeV
  in Run I)
• Integrated luminosity
• gt 200 pb?1 now in Run II (exceeds the 120 pb?1 in
  Run I)
• The plan is to accumulate 4.3 fb?1 8.4 fb?1 by
  2009

6
Fermilab Accelerator Chain
7
The DØ Experiment

• The DØ collaboration is a group of 600
  physicists from 80 institutions in 19 countries

8
Particle Detectors

• Each collision produces 100 200 particles
• Use a multi-component detector to identify
  particles and measure their energies and momenta

9
The DØ Detector

• The DØ detector weights 5500 tons, measures 17 m
  (L) ? 11 m (W) ? 13 m (H)

10
The DØ Detector
11
Central Tracking System

• Goals detection of charged particles, momentum
  measurement, vertex (interaction point,
  long-lived particle decay)
• Silicon Microstrip Tracker and Central Fiber
  Tracker inside super-conducting Solenoid which
  provides 2 Tesla magnetic field

12
Preshower Detectors

• Central Preshower detector and Forward Preshower
  detectors enhance the electron and photon
  identification

13
Calorimeter

• Measure particle energy, distinguish particle
  type by energy deposition pattern
• A Central Calorimeter and two End-cap
  Calorimeters
• Inter-cryostat Detectors between the cryostats
• Sampling calorimeter
• Uranium absorber
• Liquid Argon active medium
• Stable, uniform response, radiation hard
• Finely segmented (55,000 channels), full coverage

14
Calorimeter

• Projective design
• Cells are arranged into towers of size 0.1?0.1 in
  ?????
• Each tower has electromagnetic, fine-hadronic,
  and coarse hadronic layers

?
15
Calorimeter Electronics

• All 55,000 channels
• Signal from detector ? preamplifier ? sampled
  every 132 ns and pipelined ? waiting for trigger
  decision ? baseline subtraction ? read out

16
Trigger and Data Acquisition System

• Select and record events of interest
• Three levels of triggers (Level 1, Level 2, Level
  3)
• L1 collects prompt readout from sub-detectors and
  makes a very fast decision
• L2 refines L1 trigger information, then combines
  L1 trigger objects from different sub-detectors
  and makes decision
• L3 (software trigger) coarsely reconstructs data
  to find physics objects, performs filtering, and
  records selected events

17
Electron Reconstruction

• Electrons or photons are electromagnetic (EM)
  objects
• Almost all energies deposit in the EM section of
  the calorimeter
• EM shower cluster is isolated and narrow in the
  calorimeter
• Electron has a track detected in the tracking
  system, while photon does not
• EM objects are reconstructed in a
• cone of radius R ???2??2 0.4
• Simulation of electrons of
• transverse energy ET 80 GeV
• (Smaller cone in the plot is for
• isolation measurement)
• EM energy correction
• EM energy scale is determined by
• the known Z0 mass 91 GeV
• using Z0 ? ee? events

R 0.4
18
Jet Reconstruction

• Partons (quarks and gluons) produced from a
  collision develop into jets in the detector via
  hadronization process
• Energies deposit in both the EM and hadronic
  sections of the calorimeter
• Jet shower cluster is large and wide in the
  calorimeter
• A jet usually has several associated tracks in
  the tracking system

• Jets are reconstructed in a cone of radius R
  0.5 (or 0.7)
• Jet energy correction
• Jet energy needs to be corrected due to energy
  lose by out-of-cone showering, the response of
  the calorimeter, and an offset energy (underlying
  events, pile-up, detector noise)
• Jet energy scale can be determined using the ET
  balance in photon jet events, since the photon
  energy is measured and corrected by EM energy
  scale

Jets
Electrons
19
Data Selection

• Data sample
• Data used were collected with the DØ detector
  from Apr 2002 to Sep 2003
• Selected events were required to pass single EM
  triggers or di-EM trigger
• Integrated luminosity 175 ? 11 pb?1
• Event selection
• Two EM objects with ET gt 25 GeV, at least one of
  them with a track match
• At least two jets with ET gt 20 GeV, which are far
  from the EM objects
• To suppress Z0 (jets) ? ee? (jets) background,
  cut out Z0-mass window

Selection Criteria Events
Starting sample gt 10M
Two EM objects (ET gt 25 GeV) 13396
At least two jets (ET gt 20 GeV) 309
Veto Z0 events by Mee cut 85
20
Event Display
?? view
XY view
1st electron 2nd electron 1st jet 2nd jet
ET 95 GeV ET 89 GeV ET 220 GeV ET 33 GeV
21
Signal and Background

• LQ pair ? eejj signals of various LQ mass are
  produced by Monte Carlo simulation
• Next-to-leading order (NLO) theoretical cross
  sections of scalar leptoquark pair production are
  calculated
• Background from the Standard Model physics
  processes
• Z0/Drell-Yan process
• pp ? Z0/? jets ? ee? jets
• Top quark pair production and decay
• pp ? t t, t ? Wb, W ? e?e pp ?
  t t ? ee?bjetbjet
• Background from mis-measurement
• QCD (multi-jet) events with four or more jets,
  where two of the jets are misidentified as
  electrons
• Although the probability of misidentification is
  small, given that the cross section of QCD
  process is several orders higher than processes
  producing real electrons, the contribution of QCD
  background is still significant

22
QCD Background

• Misidentification rate (fake rate)
• Probability of a jet being misidentified as an EM
  object (fj?EM) or as an EM object with track
  match (fj?trk)
• Misidentification background
• Loop over all possible permutations of a QCD
  four-jet event that would give two EM objects and
  two jets
• Multiply fake probability to the weighted number
  of four-jet events in QCD sample, where the
  weight is the number of permutations
• We require at least one EM object with track
  match, so
• fake probability fj1?EM ? fj2?trk fj1?trk ?
  fj2?EM ? fj1?trk ? fj2?trk

23
Z0/Drell-Yan Background

• Di-EM data compared with Z0/? ? ee? Monte Carlo
  (MC)

Invariant mass of two EM objects (ET gt 25 GeV)
Data 13396
Background 12890 ? 834
Z/Drell-Yan 10777 ? 772
QCD fake 2113 ? 317

• To evaluate Z0/Drell-Yan background to the eejj
  signal, further require two jets and veto Z0-mass
  range

24
Top Background

• Top pair decay in di-electron channel Monte Carlo
  samples
• Only a handful of events expected under the
  integrated luminosity
• But since top events have high ET jets, this
  background becomes significant after the final
  cuts

25
Data and Background Comparison

• Require two EM objects (ET gt 25 GeV) and two jets
  (ET gt 20 GeV)

Invariant mass of two EM objects
Data 309
Background 325 ? 71
Z/Drell-Yan 222 ? 49
QCD fake 100 ? 27
Top 3.0 ? 0.3
26
Data and Background Comparison

• Further veto Z events by requiring Mee lt 80 or
  Mee gt 102 GeV

ST distribution
Data 85
Background 118 ? 27
Z/Drell-Yan 36 ? 7
QCD fake 80 ? 21
Top 2.5 ? 0.2
LQ (240GeV) 4.5 ? 0.5

• ST ? ETEM1 ETEM2 ETjet1 ETjet2
• The background and leptoquark signal are well
  separated in ST distribution

27
Optimization

• The data are consistent with the Standard Model
  background, and no evidence for leptoquark
  production is observed.
• To optimize the leptoquark limit setting, apply
  an additional ST cut
• Choose ST gt 450 GeV which optimizes the signal
  over background
• Observed 0 event while 0.43 ? 0.12 event is
  expected

(GeV) Data Background Z/Drell-Yan QCD fake Top LQ (240GeV)
STgt250 15 14.7 ? 3.6 7.4 ? 1.8 6.4 ? 1.9 0.90 ? 0.18 4.5 ? 0.5
STgt300 6 7.0 ? 1.8 4.0 ? 1.0 2.5 ? 0.7 0.42 ? 0.10 4.4 ? 0.5
STgt350 3 3.2 ? 0.8 2.0 ? 0.5 1.0 ? 0.3 0.18 ? 0.05 4.3 ? 0.5
STgt400 2 1.6 ? 0.4 1.2 ? 0.3 0.4 ? 0.1 0.07 ? 0.02 3.9 ? 0.5
STgt450 0 0.43 ? 0.12 0.23?0.06 0.17?0.05 0.032?0.009 3.3 ? 0.4
28
Efficiencies

• Signal acceptance and efficiency
• Acceptance fraction of events passing kinematic
  and geometric cuts
• Trigger efficiency 99.9 estimated using Z?ee
  events
• EM identification efficiency 84.1 (91.5) per
  electron in the central (forward) region
  estimated using Z?ee events
• EM track match efficiency 92.5 (54.6) per
  electron in the central (forward) region
  estimated using Z?ee events
• Jet identification efficiency 97.4 per jet
  estimated in QCD di-jet events

LQ mass (GeV) NLO cross section (pb) Overall efficiency ()
180 0.492 11.6 ? 2.1
200 0.250 16.9 ? 2.6
220 0.131 22.5 ? 3.0
240 0.0705 28.0 ? 3.3
260 0.0387 31.1 ? 3.2
280 0.0214 32.8 ? 3.2
29
Results

• We observed no evidence for leptoquarks
• Use Bayesian approach to set upper limit on
  leptoquark cross section

• Mass limit 1st generation scalar leptoquark mass
  gt 238 GeV at the 95 confidence level when decay
  branching ratio ? 1

30
Search for 1st gen. LQ in e?jj channel

• First generation LQ pair ?e?eqq
• Electron 2 jets missing-ET (ET)
• Decay fractions 2?(1??)
• Data selection (175 ? 11 pb?1)
• EM ETgt35 GeV, 2 jets ETgt25 GeV
• ET gt 30 GeV, MT(e?) gt 130 GeV
• Background
• QCD (from data)
• W2jets and top pair Monte Carlo
• Signal
• Efficiency from 13.1 ? 1.5 (MLQ 160) to 25.3 ?
  2.8 (MLQ280)
• Choose optimized cut ST gt 330 GeV

Data Total Background
ETgt30 GeV 687 666.9
MT(e?)gt130 GeV 15 22.85 ? 1.82 ? 1.58
STgt330 GeV 2 4.73 ? 0.82 ? 0.3

• MLQ gt 194 GeV at 95 C.L. for scalar LQ when ?0.5

31
Combine two channels for 1st gen. LQ
Excluded Region

• Mass limits for first generation scalar LQ at 95
  C.L.
• MLQ gt 238 GeV for ?1, MLQ gt 213 GeV for ?0.5
• Run I limits MLQ gt 225 GeV for ?1, MLQ gt 204
  GeV for ?0.5

32
Search for 2nd gen. LQ

• In ??jj channel (decay fraction ?2)
• Luminosity 104 pb?1 (Aug. 2003)

• In ??jj channel (decay fraction 2?(1??) )
• Luminosity 150 pb?1

• MLQ gt 186 GeV for ?1

• MLQ gt 144 GeV for ?0.5

• Run I limits MLQ gt 200 GeV for ?1, MLQ gt
  180 GeV for ?0.5

33
Summary

• We have searched for first generation leptoquarks
  in the di-electron di-jet final state with the
  DØ detector at the Fermilab Tevatron
• This search is based on the data collected in Run
  II with an integrated luminosity 175 pb?1
• The data are consistent with the Standard Model
  predictions, and no evidence for leptoquark
  production is observed
• We set a lower mass limit of 238 GeV/c2 at the
  95 confidence level for the first generation
  scalar leptoquarks when the decay branching ratio
  ? 1
• In Run II ( 2009), we expect to probe the
  leptoquark of mass up to 350 GeV/c2 for
  exclusion, or 300 GeV/c2 for discovery!