tt spin correlations with semileptonic channel

1
tt spin correlations with semileptonic channel
F. Hubaut, E. Monnier, P.
Pralavorio CPPM 01/2004 Top WG meeting

• Goal
• Update the tt spin correlation study in
  semileptonic channel using production and decay
  density matrices
• Reminder
• Last meeting, we reproduced our Prague
  colleagues dileptonic results at the parton level
• Since last meeting
• Tools upgrade for semileptonic analysis
• Event topology reconstruction

first results
2
Tools

• Standard PYTHIA does not include spin
  correlations
• Upgraded to Toprex4.1 and Pythia6.221 with
  Atlfast2.60, Tauola2.6, Photos2.5

• Use TOPREX (developed by CMS) with LO full
  density matrices
• Pythia for hadronization, fragmentation and
  decays
• Tauola package for t decays

• Generalist can do single top, Higgs resonance,
  rare top decays (FCNC, t Hb),
• Evolutive (future integration in Athena
  framework)

3
Event generation and reconstruction

• 2 processes in Toprex are well suited
• CTEQ5L structure function set
• 5.4fb-1 of events in the semileptonic channel
  with process 20, with process 50, and with
  standard Pythia 6.221, w/wo ISR/FSR (no
  pile-up)
• Events reconstructed with Atlfast2.60

• process 20 tt production in the Breit-Wigner
  approach with off-shell W bosons
• process 50 tt production in the on-shell
  approximation, with the possibility to switch on
  an intermediate Higgs resonance (switched off in
  the following)

can help to understand the theoretical
systematics
4
Asymmetry coefficient

• T (T-) angle between chosen spin axis and
  spin analyzer direction in the t(t) rest frame.
  Spin axis is t(t) direction in the parton c.m.s.
  (helicity basis)
• 9 ltCosT CosT-gt unbiased estimator for the
  asymmetry coefficient A, which measures the spin
  correlations
• leptonic side lepton (?1)
• hadronic side W or b (?0.4)
• least energetic jet (lej) (?0.5)

A-0.33? ?-
Standard Pythia Mean 0.00?0.02 A
Entries
In the following, statistical errors 0.02
9 CosT CosT-
5
Asymmetry at parton level
(lej as hadronic spin analyzer)
A0.00
Pythia
Toprex 20
Toprex 50

• Toprex reproduces the LHC expected asymmetry
  (-0.3310.5)
• Good agreement between Toprex processes

6
tt invariant mass
Higher asymmetry for low tt invariant mass
(lej as hadronic spin analyzer)
Asymmetry
Mtt (GeV)
All events
Mtt lt 500GeV
A-0.16
A-0.24
Cut on Mtt can increase asymmetry
7
Signal reconstruction
(similar to top mass analyses)

• Selection
• 1 lepton gt 20GeV (?lt2.5)
• PTmiss gt 20GeV
• at least 4 jets gt 40GeV (?lt2.5)
• at least 2 b jets among them
• Reconstruction
• Jet calibration and B-tagging are performed with
  Atlfast-B
• Select the 2 non-b jets with Mjj closest to MW
• Select the b jet with Mjjb closest to Mt
• Use PTmiss for the PT? and compute Pz? by
  constraining Ml? to Mw
• Select the b jet closest to the lepton
• Select the ? with Ml?b closest to Mt
• Quality cuts
• Select events with lMwREC - Mwl lt 20GeV and
  lMtREC - Mtl lt 35GeV
• Cut on Mttlt500GeV

all cuts 2 efficiency (25 kevents
after cuts)
8
Mass reconstructions
Hadronic W
Mean 80.4 GeV Sigma 6.0 GeV

MW
Hadronic top
Leptonic top
Mean 173.1 Sigma 10.7
Mean 172.4 Sigma 10.9
Mt
Mt

• in agreement with top mass paper
• same results with Toprex or Pythia
• spin correlations do not influence
  top mass reconstruction

9
Asymmetry results
(lej as hadronic spin analyzer)

• at all levels good agreement toprex20/toprex50
  and Pythia compatible with 0 (error 0.02)
• Asymmetry still mesurable after reconstruction
• Necessary to study the systematics induced by
  cuts and reconstruction

10
Idem without ISR/FSR
A0.00
A-0.24
A-0.24
A0.00
A-0.20
A-0.20
A-0.02
A-0.20
A-0.20

• without ISR/FSR, reconstruction level similar to
  parton level after cuts
• ISR/FSR dilute asymmetry

11
W or b instead of lej as spin analyzer

• W (or b) is a less powerful spin analyzer
• different sensitivity to systematics?

12
Conclusions

• Production of semileptonic events performed with
  Toprex4.1 Pythia6.221 Atlfast2.60 for tt
  spin correlation studies
• Analysis at parton level gives expected results,
  as for dileptonic channel
• Event topology reconstruction performed good
  agreement with top mass paper
• Asymmetry still measurable after reconstr., even
  if ISR/FSR degrade it

Our semileptonic results Our semileptonic results Our semileptonic results Our semileptonic results
Pythia Toprex, lej Toprex, W
No ISR/FSR -0.02 ?0.02 -0.20 ?0.02 -0.17 ?0.02
ISR/FSR 0.03 ?0.02 -0.15 ?0.02 -0.08 ?0.02
Prague dileptonic results Prague dileptonic results Prague dileptonic results Prague dileptonic results
ISR/FSR -0.07 ?0.02 -0.18 ?0.02 -0.18 ?0.02
(statistical errors only, no background)
13
Outlook

• Near term
• study systematics
• include background (S/B70 after cuts), pile-up
  effects?

• Look at angles in laboratory frame (less
  sensitive to systematics)
• sign correlation effects
• easier at beginning of data taking
• Perform a likelihood on normalised shapes (as
  Toprex/Pythia) with a well tuned MC (acceptance
  effects, )
• get absolute asymmetry measurement

14
Asymmetry results
(lej as hadronic spin analyzer)

• at all levels good agreement toprex20/toprex50
  and Pythia compatible with 0 (error 0.02)
• Asymmetry still mesurable after reconstruction
• Necessary to study the systematics induced by
  cuts and reconstruction

15
W or b instead of lej as spin analyzer

• W (or b) is a less powerful spin analyzer
• different sensitivity to systematics?