An Attempt to Look for SUSY

1
An Attempt to Look for SUSY
John Zhou Rutgers University
2
Outline

• Brief introduction to SUSY
• A sample analysis
• Results
• Run II outlook
• Conclusions

3
The Standard Model
Standard Model A quantum theory that
includes theory of strong interactions (Quantum
Chromodynamics, QCD) and the unified theory of
weak and electromagnetic interactions
(electroweak theory). It is well tested
experimentally to a very high precision.
4
Limitation of the Standard Model

• There are minimal 18 parameters in the SM that
  have to be put in hand (highly undesirable for an
  ultimate theory.
• Electroweak Symmetry Breaking (EWSB) through an
  ad hoc way The Higgs Mechanism. It is
  responsible for generating mass for SM particles.

• The hierarchy problem

Higgs potential Higgs mass
EW scale If SM valid to GUT scale
f
s
H
H
5
Supersymmetry

• Supersymmetry incorporates additional
  symmetry between fermions and bosons. For each SM
  particle, there is a SUSY partner with spin
  differ by 1/2.

6
Why supersymmetry?

• solves the fine tuning problem.
• naturally gives EWSB.
• GUT candidate Main motivation.

7
SUSY Models

• SUSY is complex. It adds a lot of new particles
  to the SM. The minimal super-symmetric extension
  of the SM (MSSM) has 124 free parameters.
• SUSY must also be a broken symmetry, otherwise
• In order to effectively study SUSY, additional
  constraints on SUSY breaking is needed.

8
SUSY Models

• Different SUSY models were devised based on how
  SUSY is broken. There are two major frameworks.
• mSUGRA - minimal Supergravity
• Gravity mediates SUSY breaking from hidden sector
  to visible sector
• Only 5 SUSY parameters at MSUSY m0, m1/2, tan?,
  A0, sign(?)
• GMSB - Gauge Mediated SUSY Breaking
• R-parity

Supersymmetry breaking origin (Hidden sector)
MSSM (Visible sector)
Gravity
9
Enrico Fermi
10
DØ Single Electron Analysis

• Signal electron ? 4 jets ET

• Motivation
• Search for parameter space which is sensitive to
  chargino and neutralino decays to W and Z
  respectively large jet multiplicity (not as
  sensitive in other channels).
• Complements other mSUGRA search channels at DØ.

is the LSP.
11
Event selection

• DØ 1994-95 data (triggered on electron, jets and
  ET)
• Integrated luminosity 92.7 pb-1
• 1 isolated electron tight electron id selection,
  isolation cut
• GeV,
• or
• 4 jets cone size 0.5, jet id cuts
• GeV,
• Missing Energy
• ET gt 25 GeV
• No second loose electron defined using dilepton
  signal electron definition
• No good muon with
• We observed 72 events.

12
Backgrounds

• Physics
• Standard Model e ? 4 jets ET
• W ? 4 jets
• t t
• WW ? 2 jets
• Instrumentation
• QCD 5 or more jet events with one jet faking an
  electron and inaccurately measured jet energies
  leading to ET

13
Event simulation Fast MC, to explore the large
parameter space
Kinematic cuts
Signal, ttbar, and WW bkgd are simulated with
FMCØ.
Structure of FMCØ
Physics Event Generator (PYTHIA)
Kinematic Ntuple
elecs jets muons ET
s_elecs s_jets s_muons s_ET
Event Weighting (trigger)
Event Weighting (object ID)
particles
Jet (RECO)
Object Smear
Acceptance
14
Multijet background
CC
EC
Fake electron ET distributions are normalized to
the good electron ET distribution in the low ET
region (dominant by multijet background). Tails
in the fake sample in the high ET region (the
signal region is ET gt 25 GeV) models the multijet
background in the signal region. Result 19.1 ?
4.7 events
15
Total number of background events

• t t 17.4 ? 5.5
• PYTHIA generator FMCØ
• ? 5.9 1.7 pb
• WW ? 2 jets 1.4 ? 0.3
• PYTHIA generator FMCØ
• ?WW 10.4 0.23 pb
• Multijet 19.1 ? 4.7
• Estimated from data
• W ? 4 jets 32.2 ? 5.7
• Estimated from data and MC
• Total background 70.1 ? 9.2

Observed 72 events
Data agree with the Standard Model background
expectation very well! But we need to examine
more.
16
Data-Model Comparison
17
NN variables to separate mSUGRA signal from
background

• Variables
• Aplanarity
• .

tt 3C fit
18
Conclusions of Data-Model Comparison

• Conclusions
• The observed data are well explained by the
  Standard Model backgrounds.
• The existence of signal is not conclusive based
  on the observed number of data events and the
  estimated number of background events and error.
• We proceed to set limit on the signal.
• Remember these 72 events survived only the
  initial cuts. More optimized cuts can enhance the
  signal sensitivity. We use NN to do this.

19
NN training result
Signal m0170 GeV m1/260 GeV tan(?) 3
? lt 0 A0 0 ? 31.5 pb (SPYTHIA) a
0.0056 (FMCØ) Nevents 16.3 ? 2.9
20
Significance and NN cut
The expected significance
where
and S(nb) the number of standard deviation
that background must fluctuate to at least n
events

• NN cuts at where the significance is maximal.
• For this particular param. set
• NN cut 0.80
• Nsignal9.5?1.7, Nbkgd4.50.9

21
Result
ALEPH Limit
22
Conclusions

• Supersymmetry is a viable replacement of SM.
• Supersymmetry is complex and has many models.
• We searched for mSUGRA using D? Run I data in the
  electron ? 4 jets ET channels for just one
  particular model with one particular tan?.
• Advanced analysis techniques (e.g., NN) is needed
  to enhance the signal sensitivity.
• Run II offers a great opportunity to further
  search for SUSY. A lot needs to be done.
• LHC claims that they can discovery SUSY in the
  first month of running. So, wed better work hard
  now.

23
tt 3C fit