New Phenomena at the Tevatron

1
New Phenomena at the Tevatron

• Susan Blessing
• Florida State University
• CERN
• March, 1998

2
Possible Topics

• Supersymmetry
• Minimal Supersymmetric SM
• gauge mediated
• gravity mediated
• with photons
• without photons
• Supergravity
• R-parity violating
• Specific production and decay modes
• By final state
• Leptoquarks
• First, second, third generation
• Pati-Salam
• Monopoles
• New gauge bosons

• New particles decaying to dijets
• New particles decaying to diphotons
• Higgs
• SM
• Charged
• Bosonic
• Technicolor
• Fourth generation heavy neutrino
• Fourth generation charge -1/3 quark
• Compositeness

3
Leptoquarks

• SM fermions show symmetry between quarks and
  leptons -- perhaps there is some coupling between
  them and a mediating particle exists
• occurs naturally in GUT theories where quarks and
  leptons share multiplets
• Characteristics
• fractional charge
• QCD color
• lepton and baryon quantum numbers
• scalars and/or vectors
• couple only within one generation, otherwise
  large contributions to ??????e, ????e?,?K
  ?????,?K ? e?
• or be very massive

4
Leptoquarks, continued

• H1 and ZEUS reported an excess of high Q2 events
  which could be explained as first generation
  scalar LQs around 200 GeV/c2
• or as something boring - like statistical
  fluctuation
• Lepton exchange process depends on the unknown
  leptoquark-lepton-quark coupling
• for MLQ lt 200 GeV/c2, the coupling is known to be
  small from ee- ? qq total cross section and
  forward-backward asymmetry measurements
• Primarily pair produced at the Tevatron
• independent of LQ-l-q coupling

_
First generation LQs decay to eq (?q) with
branching fraction ???????
Three search channels eejj, enjj, and nnjj.
5
Leptoquarks at HERA

• 20 pb????e?p data
• H1 -- 7 events (1.0 bkg)
• ZEUS -- 4 events (0.9 bkg)
• 1 pb????e?p data
• no excess reported
• Difference between e?p and e?p can be explained
  by a particular type of LQ with ?1 (only decays
  to eq) and q 5/3

6
First Generation Scalar LQ
eejj channel
123 pb-1
0 candidates 0.44 0.06 bkg events
ST vs fit mass
Also looked at mass fitting
7
First Generation Scalar LQ
enjj channel
115 pb-1
Began with a base sample of events
MT - data and prediction
14 candidates 17.8 2.1 events predicted
No improvement with simple rectangular
or diagonal cuts, especially in top quark mass
region.
Use a neural network with two inputs for MLQ gt
120 GeV/c2. Train a separate NN for each MLQ.
Output DNN 0 - bkg DNN 1 - signal
For MLQ ? 120 GeV/c2 ST gt 400 GeV
8
enjj Neural Network
dM/M(180) vs ST
DNN - data and prediction for MLQ 180 GeV/c2
Limit contour for each NN
Cut on DNN gt 0.8 or 0.85 0.29 lt background lt 0.61
events
No events remain in the data
9
First Generation SLQ Results
? vs MLQ Exclusion Region
Excludes the interpretation of the HERA events as
first generation LQs in models without extra
fermions or intergenerational mixing.
nnjj channel
10
First Generation Vector LQ

• Cross section depends on type of coupling
• Fundamental gauge bosons
• Yang-Mills coupling (k 0, l 0)
• Composite particles
• Anomalous (or minimal vector) coupling (k 1, l
  0)
• Vector minimum (k 1.3, l -0.2)
• smallest pair production cross section at the
  Tevatron, most conservative limit
• All couplings lead to cross sections larger than
  those for scalar LQ
• Use new calculation of cross section by Boos,
  Belyaev, and Solomin (1997) at Q2 MLQ2
• MC events generated using PYTHIA v5.70
• separate MC for each coupling
• final states identical to scalar LQ
• kinematic distributions are similar
• apply scalar LQ analysis
• mostly

11
Vector LQ Analysis

• eejj channel
• identical to scalar LQ analysis
• enjj channel
• MLQ ? 220 GeV/c2
• identical to scalar LQ analysis
• ST gt 400 GeV for MLQ ? 120 GeV/c2
• NN for MLQ gt 120 GeV/c2
• MLQ gt 220 GeV/c2
• NN used a mass-based variable, trained for
  particular values of MLQ
• NN of primary use in top mass region
• use ST gt 400 GeV
• nnjj channel
• same Run 1a top squark analysisacolinear jets
  and ET

/
12
Vector LQ
Efficiencies identical within errors for three
VLQ couplings and SLQ
Comparison of three VLQ cross sections
and experimental limit for b1.
Recall no candidates for eejj and enjj channels.
13
Vector LQ Limits
95 CL limits for Yang-Mills coupling by channel
95 CL limits for all three couplings
For Yang-Mills coupling, b1, MVLQ gt 340
GeV/c2 b1/2, MVLQ gt 329 GeV/c2 b0, MVLQ gt
200 GeV/c2
14
Monopoles
Introduced by Dirac in 1931 to explain quantizatio
n of electric charge.
Should have been produced in profusion in the
Big Bang.
Extensive searches for relic monopoles, flux is
severely constrained.
From the theoretical point of view one would
think that monopoles should exist, because of the
prettiness of the mathematics. Many attempts to
find them have been made, but all have been
unsuccessful. One should conclude that pretty
mathematics by itself is not an adequate reason
for nature to have made use of a theory.
Dirac, 1981
15
Monopoles
Relic monopole searches have no sensitivity to
monopole mass.
L3 studied Z ? ggg, M gt 510 GeV/c2
Possible to produce monopoles at the
Tevatron within loops. Very strong coupling to
photons.
2 high ET photons jets escape down beam pipe
16
Monopoles - Event Selection

• Base Sample - 69.5 ? 3.7 pb-1
• two photons, ETg gt 40 GeV, hg lt 1.1no jet with
  ETj gt 15 GeV and hj lt 2.5no significant ET, ET
  lt 25 GeV
• Background
• physics backgrounds with similar loop diagrams
  with other particles are negligible
• QCD multijet misidentification
• dijets, direct photons, diphotons
• Drell-Yan misidentification
• Total estimated background
• Optimization
• vary ST SETg cut until expected background is
  0.4 events
• ST gt 250 GeV corresponds to an expected
  background of 0.41 ? 0.11 events
• no events remain in the data

/
/
17
Monopoles - Distributions
ST - Data and background
DY
QCD
ST - Signal
hg2 vs hg1 - Signal
Nearly independent of monopole mass for M gt 300
GeV/c2.
18
Monopoles - Limits
Overall efficiency for gg events (52.8?1.4)
Production of 2 or more photons with SETg gt 250
GeV and hg lt 1.1
Calculation by Ginzburg and Schiller
(hep-ph/9802310)
Monopole interpretation
Monopole mass limits
at the 95 CL
19
Supersymmetry

• New symmetry between fermions and bosons
• Predicts superpartners for all known particles
• Solves fine tuning problem
• if new particles have masses below 1 TeV/c2
• Allows for unification of gauge couplings
• Only known way to include gravity
• Provides a candidate for dark matter

20
SUSY Continued

• Lightest SUSY particle
• c01 in gravity mediated models
• dark matter candidate
• G in gauge mediated models
• c01 ? Gg
• Assumptions
• SUSY couplings SM couplings
• R-parity ? (-1) 3BL2S is conserved
• 1 for SM, -1 for SUSY
• SUSY particles are produced in pairs
• decay of a SUSY particle produces another SUSY
  particle
• the LSP (lightest SUSY particle) is stable

21
SUSY Models
MSSM

• Minimal supersymmetric extension of the SM
• direct supersymmetrization of the gauge theory
• minimum number of additional particles
• c0i and c?i described by four parameters
• M1, M2, m, tan b
• U(1) and U(2) gaugino mass parameters at the EW
  scale, higgsino mass parameter, ratio of vacuum
  expectation values of the two Higgs doublets
• Scalar sector described by many mass parameters
• Different SUSY breaking leads to different
  classes of models
• MSSM with assumptions about physics atthe GUT
  scale
• Five parameters

SUGRA
Mtop also required
22
The CDF Event
This led to many attempts to explain it within
various SUSY models.
Also possible to have photon final states with
gravity mediated SUSY.
In 1996, 26 papers on Gauge Mediated SUSY In
1993, 1994, 1995, 0 papers!
23
More Events with 2gs?
CDF and DØ have made extensive searches for more
diphoton events.
CDF has searched for exclusive gg X final
states in 85 pb-1
No excess
/
24
Inclusive gg ET
/
ETg gt 12 GeV, hg lt 1.1
Note cuts are different for two
experiments, cannot compare directly.
Z ? ee events
CDF
use Z ? ee data as model of ET
spectrum normalize to inclusive ggET
spectrum ETg1 gt 25 GeV ETg2 gt 12 GeV
/
/
gg events
Single gg event stands out
25
What we rule out
Neither experiment sees any excess diphoton
events. We are able to rule out some, all or
none of various models.
Everything from this model!
Some of this model
(examples, not an exhaustive list!)
26
And ...
A little of this model
Almost none of this model!
27
c02 c01 g Models


No gaugino mass unification!
28
Search for ggET
/
DØ has published a ggET analysis (PRL 80, 442
(1998))
/
2 events observed 2.3 ? 0.9 events expected
All of these limits are well above the
theoretical cross sections
29
Search for gET ? 2 jets
/
Almost no SM backgrounds at the parton level
Significant instrumental backgrounds
Base Sample
ETg gt 20 GeV Two or more jets ETj gt 20
GeV hj lt 2.0 ET gt 25 GeV
/
99.4 ? 5.4 pb-1
378 events 74 with ? 3 jets 10 with ? 4 jets
Expect 375.6 ? 35.7 events with 3 jets 74.9 ?
16.3 with ? 3 jets 7.6 ? 4.5 with ? 4 jets
30
gET ? 2 jets, Signal
/

• pp c?iq, c?ig, c0jq, c0jg, qq, qg, gg c02
  X ? g ET jets modeled using SPYTHIA
• M1, M2, tan b, m fixed
• M1 M2 60 GeVtan b 2m - 40 GeV
• Other parameters varied within model
• Sleptons assumed to be heavy so they do not
  affect decays of other sparticles
• All top squark production ignored
• Three cases studied
• m(q) m(g)m(q) gtgt m(g)m(q) ltlt m(g)
• For m(q) m(g), in base sample expect351 events
  for m(q) 200 GeV/c2 19 events for m(q) 300
  GeV/c2

/










31
gET ? 2 jets
/
Comparison of data, background, signal
ETg Distribution
Njets Distribution
32
gET ? 2 jets
/
5 events remain in the data 8.1 ? 5.8 events are
expected


If m(q) m(g) 300 GeV/c2, expect 11.3 events
Limits
Kane et al.s models not ruled out for heavy
squarks/gluinos
33
Search for g b ET
/
CDF 85 pb-1
/
ET Distribution
Njets Distribution
Very similar to DØs model, but with tanb 1
(rather than 2) and fixed top squark mass and
decay.
34
Limit
Limit in squark/gluino mass plane



35
So what is the CDF Event?
SM Source?
_
/
/
Unlikely!
36
So what is the CDF event?
Forward electron may not be an electron. It
passes all electron cuts, but the SVX track stub
misses the cluster.
Df between SVX track and EM centroid.
Central electrons
Forward electrons
Electron? missing SVX track Photon? other
nearby tracks Hadronic tau decay? Hadronic jet?
passes all electron cuts - rare
It can be whatever you like.
J. Conway, CDF
37
Supergravity

• Search within the SUGRA parameter space so that
  results can be combined
• Look for final states regardless of origin
• generate MC events based on various sets of
  parameters
• search for final states with dileptons, single
  leptons, or jets (and ET)

/
38
Parameter Space
Work primarily in m0, m1/2 space
Vary tanb to a lesser extent
Set sign(m) -1 (most of 1 parameter space we
can reach has been ruled out)
A0 affects only the top squark mass at the
Tevatron
Contours show constant squark and gluino masses
39
Gluino Cascade Decays



M(g) gt M(q)
0
? e?? ?1


? ??? ?1
_

0
? q q ?1

??2 jets
/
large ET



M(g) lt M(q)
0
? e?? ?1


? ??? ?1
_


0
g ? q q ?2

_


? q q ?1

??2 jets
_

0
? q q ?1

40
Relative Squark/Gluino Production
41
SUGRA -- jets ET
/
Base requirements 3 jets with ET gt 25 GeV 1 jet
with ET gt 115 GeV HT ?ETjet gt 100 GeV
(excluding leading jet) Lots of cuts to ensure
well-measured events
42
Signal varies within parameter space
Optimize ET and HT SETjet cuts for each point
in parameter space
HT gt 100 -- 250 GeV ET gt 50 -- 150 GeV
/
43
Compare prediction and data
Points - prediction Histogram - data
44
SUGRA -- ee jets ET
/
tanb 2
Number of leptonic decays depends on
parameters, tanb is important here
tanb 6
Background 3.2 ? 1.3 events Data
2 events
45
The Dip
46
SUGRA Limits
Squark and Gluino Mass Limits
Curved contours are squark masses Horizontal
contours are gluino masses
47
Conclusion

• Lots of new phenomena activity
• Limits on many models
• No discoveries ..
• Looking forward to the remainder of the Tevatron
  Run 1 analyses and to Run 2
• more energy
• more luminosity
• more data
• discoveries?

48
Dijet Resonances
95 CL Exclusion Regions
DØ
Mq lt 725 GeV/c2 340 lt MW lt 680
GeV/c2 365 lt MZ lt 615 GeV/c2
CDF
200 lt Mq lt 750 GeV/c2
49
eejj Optimization
Optimized efficiency for MLQ 200 GeV/c2 for a
fixed background of 0.4 events (68 chance no
events will remain).
Tried over 20 variables, in over 50
combinations energies, masses, mass differences,
event shape
Comparison of ST and dM/M.