Search for chargino and neutralino

1

• Search for chargino and neutralino

• High pt Dimuon channel with additional lepton

• Anadi Canepa, O.González, E.Lytken D.Bortoletto
• (Purdue University, IN)

• Talk outline
• Analysis strategy
• Expected sensitivity
• Control region study

CDF note in preparation
2
Analysis strategy

• Blind analysis performed as counting experiment
• Search for trilepton signal 2 muons
  additional lepton (e or ?)

Distribution of transverse momentum of muons in
benchmark point
Point (a) m0 100 GeV, m1/2 180 GeV, tan?
5, A0 0, ? gt 0
Leading muon mostly pt gt 20 GeV/c Exploit single
high pt muon triggers
3
Data set validation bewk2d
MUON_CMUP18 ? 0.899 /- 0.006 MUON_CMX18
? 0.968 /- 0.005 (CDF 7367)
Good Run List v7 (CMUP-CMX-Electron-NoSilicon) Lu
minosity 346 pb-1
Z invariant mass flat in time
Maybe smearing of the Z peak needed since mean
differs from 91.2 GeV ?
4
Lepton scale factors

• Leading muon pt gt 20 GeV/c
• EWK muon ID
• Central loose muon
• Non-leading muon pt gt 5 GeV/c
• INTERMEDIATE pt muon ID
• Same as above but
• pt gt 5 GeV
• Had Energy lt (3.5 pt/8) GeV
• Stub matching relaxed from ?X to ?X or ?2
  requirement
• Third lepton
• Either CMIO (pt gt 10 GeV) or loose central
  electron (Et gt 5 GeV)

MUON_CMUP SF 0.881 /- 0.01 MUON_CMX SF
0.997 /- 0.0017 (CDF 7367)
new pt range
MUON_CMUP id SF 0.99 0.005 MUON_CMX id SF 1.00
0.085 (CDF 7197)
CMIO SF 1.0 (CDF) Loose Electron S.F 1.0 (CDF)
MUON_CMUP reconstruction SF 0.9315 /-
0.0047 (CDF 7367) MUON_CMX reconstruction SF
0.987 /- 0.0026 (CDF 7367) Intermediate pt muon
isolation SF 0.915 /- 0.055
5
Background contributions (3lll)

• We split background contributions into two
  categories
• 3 real leptons (three leptons at generator level)
• tt-bar
• bb-bar
• ZZ WZ
• 2 real muons additional lepton (CMUP, CMX or
  CMIO)
• Drell Yan / WW Fake lepton
• Two ID muons (estimated from data)
• Drell Yan / WW electron from conversion
• Two muons ? radiated off from leptons at
  generator level
• () Fake rate estimated in jet samples with few
  leptons leptons from decay of ? radiated off
  from leptons NOT included in fake rate

6
Background contributions (2ll)

• We split background contributions into two
  categories
• 2 real leptons (two leptons at generator level)
• Drell Yan
• tt-bar
• bb-bar
• ZZ WZ WW
• 1 real muons additional lepton (CMUP or CMX)
• W Fake lepton
• One ID muon (estimated from W decaying to ?1pt)

7
New from last Exotic Meeting

• Heavy flavor
• Oscar generated a large sample of b-bbar decaying
  into trilepton
• Trilepton contribution estimated by applying a
  scaling which takes into account acceptance and
  reconstruction efficiency
• WZ/?
• Found a problem at generator level MADGRAPH
  scales the event depending on the invariant mass
  with a factor based on the charge of the leptons
  involved
• Else produced new samples start looking at that
• For this talk, used WZ sample validated earlier
  (PYTHIA)

8
WZ/? MADGRAPH samples
9
Fake Rate for CMIO

• Late February the calculation of fake rate
  changed
• To get better agreement data-MC, the fakeable
  track must be isolated
• Fake rate for CMIO looks too high (needs
  investigation)
• In signal region, contribution from 2 muons 1
  fake
• 0.05 if CMUP or CMX or loose electron
• 0.15 if CMUP or CMX or CMIO or loose electron
• Plan to run the fake rate calculation

10
DY Conversion

• With the implementation of new cuts, the event of
  DYconversion in the Drell Yan sample removed
• Move to Z? sample for such an evaluation
• Sample is 4.11.2 so the systematics will be
  affected by material

11
Monte Carlo Samples

• Monte Carlo Samples
• tt-bar
• official ttopel Luminosity pb-1
• PYTHIA
• bb-bar
• Multilepton Group Sample
• direct production, gluon splitting, flavor
  excitation
• ISAJET
• WZ
• Multilepton Group Sample Luminosity pb-1
• PYTHIA
• WW, DY and ZZ/?
• Multilepton Group Sample
• WW Luminosity 30,324 pb-1 (? pb)
• DY Luminosity 901 pb-1
• ZZ Luminosity pb-1
• PYTHIA

For scaling only
12
Fake rate implementation
Looks for fake objects in a given dimuon event in
DATA

• If 1 fakeable object (fake rate f.r1)
• the event goes through the analysis cuts with
  the fake object as third lepton and a weight
  f.r1
• If 2 fakeable objects pt1 gt pt2 (fake rate f.r1
  f.r2)
• the event goes through the analysis cuts with
  the fake object (1) as third lepton and with a
  weight f.r1
• the event goes through the analysis cuts a
  second time with the fake object (2) as third
  lepton and with a weight f.r2(1-f.r1)

13
Clean up cuts

• The event is rejected if triggered
  CMUP18_Volunteer only
• The event is rejected if at least one jet with ?
  lt 0.1 and if at least one jet in the chimney
  (improvement up to 4)
• The event is rejected if outgoing pair
• Still cosmic contribution cut on acollinearity

14
Event scaling

• Each event is scaled for ID efficiency and
  trigger efficiency
• The weight of the event is w ?trig SF
• The weight of the event is scaled for the SF
  associated to each lepton in the event
• SF sf(lepton1)sf(lepton2) sf(lepton3)
• The weight of the event is scaled for the trigger
  efficiency
• If 1 CMUP muon, ?trig ?CMUP
• If 1 CMX muon, ?trig ?CMX
• If 1 CMUP 1 CMX muon, ?trig 1.0
• If 2 same type muon, ?trig ? (2- ?)

15
Analysis cuts

• TWO CENTRAL muons and a third central lepton
• lepton pt is Beam Constrained Pt
• ?Rll gt 0.4
• ?Zll lt 4 cm
• if electron, checked for conversion
• maxSep 0.1 ?Cot? lt 0.02
• efficiency ? 99.8 0.3 (Kobayashi)
• Kinematic cuts to reduce background
• Mass cut
• Jet cut
• MEt cut

16
Kinematic variables jet (I)

• New Jet Energy Correction implemented
• JetCorr04a
• In Jet collection, jet counted if
• Raw Et gt 10 GeV
• EmFract lt 0.9
• ?Rjet-lepton gt 0.4
• The Jet Et is corrected at L5 if ? lt 2.5
  (otherwise, Raw Et is used)

N-1 plot We apply jet and mass requirement in
events with at least 3 leptons
17
Kinematic variables jet (II)
TRILEPTON EVENTS Jet multiplicity in events with
at least 3 leptons with INVARIANT MASS CUT
applied. The third object is either a real lepton
or a fake lepton
18
Jet Multiplicity (I)
19
Jet Multiplicity (II)
20
Kinematic variables MEt (I)

• MEt is corrected for
• muons (the energy in the calorimeter is
  substracted and the pt is added to the raw Met
  Met1 in Stntuple)
• jets (any jet with EmFractlt0.9 and ? lt 2.5
  corrected)
• Dilepton region, energy of the muon addition of
  Had and Em energy
• Trilepton region, assume the average deposit of
  energy 2.5 GeV

21
Control regions for the trilepton analysis
Background normalized to 346 pb-1
22
Jet clean up cut
If the events with at least one jet at ?lt1 or at
least one jet in the chimney are removed
23
Invariant mass (I)
Dimuon Event
Mass and Jet cut applied
24
Invariant mass (II)
CR A2
CR Z mass window
25
Corrected Met (I)
Dimuon Event
Mass and Jet cut applied
26
Corrected Met (II)
CR A2
CR Z mass window
27
Raw Met (I)
Dimuon Event
Mass and Jet cut applied
28
Raw Met (II)
CR A2
CR Z mass window
29
?? between muons (I)
Dimuon Event
Mass and Jet cut applied
30
?? between muons (II)
CR A2
CR Z mass window
31
Jet ? (I)
Dimuon Event
Mass and Jet cut applied
32
Jet ? (II)
33
Pt leading muon (I)
Dimuon Event
Mass and Jet cut applied
34
Pt leading muon (II)
CR A2
CR Z mass window
35
Pt non leading muon (I)
Dimuon Event
Mass and Jet cut applied
36
Pt non leading muon (II)
CR A2
CR Z mass window
37
?? between leading muon and Met (I)
Dimuon Event
Mass and Jet cut applied
38
?? between leading muon and Met (II)
CR A2
CR Z mass window
39
?? between non leading muon and Met (I)
Dimuon Event
Mass and Jet cut applied
40
?? between non leading muon and Met (II)
CR A2
CR Z mass window
41
Low Met control regions
NON Z mass region
Control region I nJetlt1
Control region J nJetgt2
42
Low Met control regions
Z mass region
Control region G nJetlt1
Control region H nJetgt2
43
High Met control regions
NON Z mass region
Control region A nJetlt1
Control region A2 nJetgt2
44
High Met control regions
Z mass region
Control region E nJetlt1
Control region F nJetgt2
45
Dilepton control regions

• Large discrepancy in the high MEt regions
• Disagreement seems to improve when using BcPt
• Disagreement present in the old data also
  (version6 of Good Run List)
• Problem is spotted by other independent groups
• Looks unrelated to the tracking resolution since
  cutting on separation between muons and MEt
  doesnt improve the agreement
• See better agreement when running on data
  veto-ing events with more than 1 vertex running
  now on new ntuple with the correct ZVertexBlock

46
Plan

• Solve the discrepancy at large MET
• Chris Hays and David Waters suggest re-weight
  based on MET
• Stephan suggests further investigation splitting
  the events into events with MET aligned to muons
  and MET randomly distributed
• Calculate the fake rate for CMIO muons to cross
  check Liverpool results use new cut for
  isolation at low pt (tot iso lt 2 GeV)?
• Evaluate the heavy flavor contribution
• Calculate the systematics
• Lepton Energy scale
• Lepton ID SF
• Jet Energy resolution
• ISR/FSR for signal and backgrounds (samples ?)
• Luminosity
• Fake rate
• Conversion removal due to material

47
MEt components
Region A (dilepton region outside Z, nJetslt1 and
MEtv gt 15 GeV
Data Drell Yan
Data Drell Yan
48
Calendar for APS

• March 17th STATUS REPORT
• March 24th UNBLINDING APPROVAL
• March 31st PREBLESSING
• April 14th BLESSING
• APS week from April 16th

49
Back up slides
50
DataSet validation
51
(No Transcript)
52
(No Transcript)
53
Fake rate calculation
DEFINITION (object based) probability of a given
jet/track to fake an electron/muon
CALCULATION
The fake rate is measured and averaged in
different jet samples (jet20 jet50 jet70
jet100) the systematic error is assumed to be
100. The fakeable object is used as third lepton
in the event the event goes through the analysis
cuts with a weight f.r.
54
ID Cuts for low pT muons
55
Dimuon event
56
(No Transcript)
57
Z mass region
58
(No Transcript)
59
Step2 Mass and Jet cuts applied
60
(No Transcript)