Top FCNC: preliminary studies in CMS

Leonardo Benucci, Top FCNC: preliminary studies in CMSFLAVOUR IN THE ERA OF THE LHC – Cern Workshop, 7th-10th November 2005

Top FCNC: preliminary Top FCNC: preliminary studies in CMSstudies in CMS

Leonardo BenucciLeonardo Benucci,, Andrea Giammanco, Fabrizio Palla Andrea Giammanco, Fabrizio PallaINFN, PisaINFN, Pisa

Costas Karafasoulis , Costas Karafasoulis , Aristotelis KyriakisAristotelis Kyriakis, Grigoris Vermisoglou, Grigoris Vermisoglou

Institute of Nuclear Physics NCSR “Demokritos”, AthensInstitute of Nuclear Physics NCSR “Demokritos”, Athens

1


OutlineOutline

Why FCNC of top quark?Why FCNC of top quark? The signal we would like to discover The signal we would like to discover

Relevant BackgroundsRelevant Backgrounds Analysis at Generator Level: useful variablesAnalysis at Generator Level: useful variables

Analysis at Generator Level: proposed Analysis at Generator Level: proposed strategy and acceptancesstrategy and acceptances

Detectors simulations and foreseen Detectors simulations and foreseen efficienciesefficiencies

Guidelines for the near futureGuidelines for the near future

1


Top decays at LHCTop decays at LHC

Top remains bareTop remains bare(t) (t) no time to hadronize before no time to hadronize before

decaydecay Top is heavy but point-likeTop is heavy but point-like

it provides clean information it provides clean information on hard processon hard process

it can be the main actor in it can be the main actor in BSM physicsBSM physics

t

b

W

Flavour Changing Neutral Current :Flavour Changing Neutral Current :t t A qA q, ,

wherewhere q = u, c q = u, c , , A= ZA= Z , g, , g, GIM suppressed:GIM suppressed: A ≈ SM x m A ≈ SM x mbb

22/M/MWW22

t

q

A

The only decay seen today:The only decay seen today:t t W bW b

(Br > 99.9%)(Br > 99.9%)

NP

1


Introducing New PhysicsIntroducing New Physics

FCNC FCNC decaydecay Br in SMBr in SM

t t qq 5 10 5 10 -13-13

t t Zq Zq 1.3 10 1.3 10 --

1313

t t g q g q 5 10 5 10 -11-11

Introducing SUSY: significant increase in BrIntroducing SUSY: significant increase in Br

Br in R-parity Br in R-parity violationviolation

<10 <10 -5-5

<10 <10 -4-4

<10 <10 -3-3

Br in Br in SUSYSUSY

<10 <10 -7-7

<10 <10 -8-8

<10 <10 -6-6

Exotic Exotic quarkquark

<10 <10 -5-5

<10 <10 -2-2

<5 10 <5 10 -4-4

R-parity violating models in MSSM (with B non-R-parity violating models in MSSM (with B non-conservation)conservation) Exotic (vector-like) quarkExotic (vector-like) quark

Exp. limits Exp. limits (95% CL)(95% CL)

< 0.003 < 0.003 (HERA)(HERA)

< 0.08 < 0.08 (LEP2)(LEP2)

< 0.29 < 0.29 (CDF)(CDF)

ttt are produced at a rate ≈10t are produced at a rate ≈1066/year at 10/year at 103333/cm/cm22/s /s

FCNC Br might reach a detectable levelFCNC Br might reach a detectable level

ANY OBSERVATION AT LHC WILL BE A SIGNAL OF NEW PHYSICSANY OBSERVATION AT LHC WILL BE A SIGNAL OF NEW PHYSICS

1


The Signal topologyThe Signal topology

tbarb

W

tZ

Tops are almost back-to-Tops are almost back-to-back in the transverse back in the transverse

planeplane

t –t –t production t production t decays SM, t decays FCNC t decays SM, t decays FCNC

l = e,

q = u,c

l = e,

l = e,

Try to find an excess of events over background in 150-200 GeV Try to find an excess of events over background in 150-200 GeV regionregion

cos cos (t(tt)t)

1


Relevant BackgroundsRelevant Backgrounds

ProcessProcess (pb)(pb) commentscomments x Br (pb)x Br (pb)

t‾t t‾t WbWb WbWb 830830 LO+NLO+NNLOLO+NLO+NNLO 4141

WW WW 2l2l ~100~100 NLO, CTEQNLO, CTEQ 4.94.9

Z+jetsZ+jets 2l+jets 2l+jets 12,52812,528 PP00=20 GeV=20 GeV 830830

ZW ZW 3l3l ~19.5~19.5 NLO, CTEQNLO, CTEQ 0.30.3

ZZ ZZ 4l4l 1.2471.247 TopReX, PTopReX, P00=20 =20 GeVGeV

0.0090.009

On tOn tt t Zq Wb signal: Zq Wb signal:

t‾t t‾t WbWb WbWb 830830 LO+NLO+NNLOLO+NLO+NNLO 4141

WW+c‾c+c‾c l+jetsl+jets 10791079 M(M(c‾cc‾c ) > 2m) > 2mcc239.7239.7

WW+b‾b+b‾b l+jetsl+jets 294294 M(M(b‾bb‾b ) > 2m ) > 2mbb65.365.3

W+W+ l+ l+ 56.2156.21 PP00=20 GeV, CTEQ=20 GeV, CTEQ 12.512.5

WW WW 2l2l ~100~100 NLO, CTEQNLO, CTEQ 4.94.9

Z+Z+ 2l+ 2l+ 4646 PP00=20 GeV=20 GeV 3.03.0

On t On t tt q Wbq Wb signal:signal:

GENERATED WITH PYTHIA AND TOPREX @

LO

GENERATED WITH PYTHIA AND TOPREX @

LO

1


A study for selection A study for selection strategystrategyWe present a preliminary study at Generator Level:We present a preliminary study at Generator Level:

identify quarks leptons and identify quarks leptons and by their MC code by their MC code build real particle as W, Z and tbuild real particle as W, Z and t

analyze kinematical (panalyze kinematical (pTT, , ,etc.) and geometrical (,etc.) and geometrical (R, angles, etc) R, angles, etc) variables for signal and backgroundvariables for signal and background define cuts in order to reject backgrounddefine cuts in order to reject background estimate efficiencies of selectionsestimate efficiencies of selections

Afterwards, efficiencies at generator level have to be convoluted with Afterwards, efficiencies at generator level have to be convoluted with effects from detector simulation (reconstruction, b-tagging, effects from detector simulation (reconstruction, b-tagging, misidentification etc.) misidentification etc.)

Actually:Actually: eestimate number of signal and background events for a given stimate number of signal and background events for a given integrated luminosityintegrated luminosity derive a Br(FCNC) upper limitderive a Br(FCNC) upper limitS and B are shown together, normalized to the same numberS and B are shown together, normalized to the same number

Rescaling for x-sections is performed only AFTER selection Rescaling for x-sections is performed only AFTER selection

1


Z reconstructionZ reconstruction cut events with leptons outside CMS cut events with leptons outside CMS acceptance acceptance ::< 2.4< 2.4 ppT T (l) cut can be quite soft(l) cut can be quite soft

cut on cut on R(l+l-) can be useful: Z and W in background R(l+l-) can be useful: Z and W in background are produced back-to-back, so are produced back-to-back, so R(l+l-) is shifted to R(l+l-) is shifted to higher valueshigher values

signalttbar

ZW_3lZZ_4l

ppT T (lept) (GeV/c)(lept) (GeV/c)

a cut on invariant mass 91.19a cut on invariant mass 91.19±10 GeV ±10 GeV results in a reduction of continuous results in a reduction of continuous background down to 3%background down to 3%

Proposed strategy:Proposed strategy: take two isolated, opposite signed and same take two isolated, opposite signed and same flavour leptons cut pflavour leptons cut pT T (l) > 10 (for e), 20 (for (l) > 10 (for e), 20 (for ) ) GeV/c, GeV/c,

< 2.4< 2.4 cut cut R(l+l-) < 2.0-2.5R(l+l-) < 2.0-2.5 cut on Z mass windowcut on Z mass window

WZ

l

l

l

RR (l+l-)(l+l-)

Signal eff.Signal eff. Relative Relative puritypurity

Before Before R(R(l) cutl) cut 89%89% 11

After After R(R(l) cutl) cut 53%53% 3.13.1

1


W reconstructionW reconstruction

Proposed strategy:Proposed strategy: take an isolated lepton with ptake an isolated lepton with pT T (l) > 20 (l) > 20 GeV/c, GeV/c, < 2.4< 2.4 skip leptons used for Zskip leptons used for Z choose MET > 25-30 GeV choose MET > 25-30 GeV cut cut R(l R(l ) < 2.3) < 2.3

N.B.N.B. p pzz(() has two solutions, best solution ) has two solutions, best solution

has to be investigated case by casehas to be investigated case by case

Signal eff.Signal eff. Relative Relative puritypurity

Before Before R(R(l) cutl) cut 79%79% 11

After After R(R(l) cutl) cut 55%55% 1.81.8

RR ((l)l)

MET (GeV)MET (GeV)

cut events with leptons outside CMS acceptance cut events with leptons outside CMS acceptance ::< < 2.4 2.4 missing Emissing ET T (l) not less than 25 GeV to reduce ZZ and (l) not less than 25 GeV to reduce ZZ and ZjetsZjets upper cut to missing Eupper cut to missing ET T (l) cannot be performed(l) cannot be performed

cut on cut on R(R(l) less effective wrt Z casel) less effective wrt Z case

signalttbar

ZW_3lZZ_4l

1


Light and b jet Light and b jet identification identification Light jet Light jet : :

even at parton level, there is a significant even at parton level, there is a significant fraction of light jets in all samplefraction of light jets in all sample in real life, things are much more complicated in real life, things are much more complicated (direct production, ISR, gluon splitting etc.)(direct production, ISR, gluon splitting etc.) cut on pcut on pT T (light-jet) has to be high(light-jet) has to be high

counting not-b jets can be very effectivecounting not-b jets can be very effective

Proposed strategy:Proposed strategy:

take an isolated jet with ptake an isolated jet with pT T (light-jet) > 40 GeV/c, (light-jet) > 40 GeV/c, < < 2.42.4

constrained it to be ONE and ONLY ONEconstrained it to be ONE and ONLY ONE

b jet b jet : : contribution to b spectrum mostly from contribution to b spectrum mostly from ttbar backgroundttbar background cut on pcut on pT T (b-jet) to reject others (b-jet) to reject others backgroundsbackgrounds apply b-tagging algorithm (as secondary apply b-tagging algorithm (as secondary vertex, impact parameter etc.)vertex, impact parameter etc.) counting b jets can be very effectivecounting b jets can be very effective

Proposed strategy:Proposed strategy:

take an isolated jet with ptake an isolated jet with pT T (light-jet) > 20-40 GeV/c (light-jet) > 20-40 GeV/c

tag i.e. with tag i.e. with > 2.0 > 2.0 constrained it to be ONE and ONLY ONE constrained it to be ONE and ONLY ONE

ppT T (q-jet) (q-jet) (GeV/c)(GeV/c)

signalttbar

ZW_3lZZ_4l

ppT T (b-jet) (b-jet) (GeV/c)(GeV/c)

1


Building tBuilding tSMSM and t and tFCNC FCNC

q-Z combinations from Zjets can be reduced q-Z combinations from Zjets can be reduced by cutting on by cutting on R(qZ)R(qZ) a cut on cone between lepton from W and Z a cut on cone between lepton from W and Z R(lZ) > 1.5R(lZ) > 1.5÷÷2 could improve selection of W-Z 2 could improve selection of W-Z opposite in transverse planeopposite in transverse plane

signalttbarZjetsZW_3lZZ_4l

Proposed strategy:Proposed strategy: take the reconstructed and selected b and q jettake the reconstructed and selected b and q jet

take the reconstructed W, Z or take the reconstructed W, Z or choose a pchoose a pz z (() solution (the one minimizing |M(Wb)-) solution (the one minimizing |M(Wb)-175| is a good choice)175| is a good choice)

In analysis, keep In analysis, keep R(l-jet) > 0.5 to avoid l R(l-jet) > 0.5 to avoid l b,c b,c jetsjets

R(q-Z)R(q-Z)

R(l-Z)R(l-Z)

A cut on A cut on coscos(t(tt) < -0.9 t) < -0.9 ÷÷-0.95 strongly suppress ZW, WW and Zjets -0.95 strongly suppress ZW, WW and Zjets backgroundsbackgrounds

1


SelectionsSelections SIGNALSIGNAL ttbarttbar ZZZZ ZWZW WWWW

Z selectionZ selection 0.440.44 0.030.03 0.250.25 0.250.25 0.400.40

x W selectionx W selection 0.150.15 0.018750.01875 0.0250.025 0.0520.052 0.0240.024

x b selectionx b selection 0.140.14 0.017250.01725 0.00050.0005 00 00

x q selectionx q selection 0.11510.1151 17 1017 10-5-5 0.00010.0001 00 00

Final efficiency selectionFinal efficiency selection

Even without rescaling for x-sections, we see Even without rescaling for x-sections, we see that ttbar is the only relevant backgroundthat ttbar is the only relevant background

1


Introducing CMS Introducing CMS simulationsimulationFull/Fast simulations tool are under developmentFull/Fast simulations tool are under development

reconstruct tracks using combinatorial Track-finderreconstruct tracks using combinatorial Track-finder identify isolated muons and electronsidentify isolated muons and electrons match tracks with clusters in ECAL match tracks with clusters in ECAL

a lepton reconstruction efficiency is introduceda lepton reconstruction efficiency is introduced

build jets using an Iterative Cone Algorithm build jets using an Iterative Cone Algorithm implement a b-tag algorithmimplement a b-tag algorithm (Combined BTag)(Combined BTag)

a b-tagging efficiency and a mis-tagging are introduceda b-tagging efficiency and a mis-tagging are introduced

evaluate missing Eevaluate missing ETT from jets in calorimeters from jets in calorimeters

missing Emissing ETT has a finite resolutionshas a finite resolutions

Fast simulationFast simulation package includes Pile-up tuning for high luminosity studies package includes Pile-up tuning for high luminosity studies

1


Reconstruction EfficiencyReconstruction EfficiencyLepton reconstruction Lepton reconstruction :: efficiency for efficiency for is high ( is high (90-95%90-95% , see e.g. , see e.g. CMS Note 2001-054CMS Note 2001-054)) for e, many parameters have to be optimized (E/p, Efor e, many parameters have to be optimized (E/p, Ehh/E/Eee, , conversions, match conversions, match between TK and ECAL etc.) : between TK and ECAL etc.) : ~80%~80% in preliminary studies in preliminary studies

Jet reconstruction Jet reconstruction :: b-tagging capabilities significantly reduce efficiency: b-tagging capabilities significantly reduce efficiency: 40-50% 40-50% in 100 GeV region if in 100 GeV region if we want to contain mis-tagging under 2%we want to contain mis-tagging under 2% (CMS Note 2002-046CMS Note 2002-046)

Trigger efficiency Trigger efficiency :: very high (very high (99%99% in in CMS Note 2001-001CMS Note 2001-001))

High luminosity effects High luminosity effects :: Pile-up rises to 23 evs at 10Pile-up rises to 23 evs at 103434cmcm-2-2ss-1-1

First studies with this scenario seem promisingFirst studies with this scenario seem promising(no extra tracks or efficiency decrease)(no extra tracks or efficiency decrease)

Expected signal efficiencyExpected signal efficiency 3-5%3-5%

Expected surviving tExpected surviving tt events @ 100 t events @ 100

fbfb-1-1220÷310220÷310

1


Estimated BrEstimated BrFCNCFCNC sensitivity sensitivity An evidence for a tAn evidence for a tZq can be claimed at 5Zq can be claimed at 5 (99% CL) if: (99% CL) if:

S/√(S+B) > 5 S/√(S+B) > 5

An FCNC signal with branching ratio An FCNC signal with branching ratio BrBrUPPERUPPER(FCNC) is given by:(FCNC) is given by:

S = S = (t‾t) Br(W (t‾t) Br(Wl) Br(Z l) Br(Z ll)ll) 2 2 ∫ ∫LLdt dt Br BrUPPERUPPER(FCNC)(FCNC)

with with ∫∫LLdt dt integrated luminosity and integrated luminosity and selection efficiency selection efficiency

Expected signal efficiency: Expected signal efficiency: 3-5%3-5% 10 fb10 fb-1-1 100 fb100 fb-1-1

Expected tExpected tt eventst events 22÷3122÷31 220÷310220÷310

Expected SExpected S 39÷4339÷43 88÷10188÷101

Expected Expected BrBrUPPERUPPER(FCNC)(FCNC) (5.3 ÷11.4) 10(5.3 ÷11.4) 10--

44

(2.0 ÷4.1) (2.0 ÷4.1)

1010-4-4

significant improvement to existent limit – close to exotic models significant improvement to existent limit – close to exotic models predictionspredictions

1


Conclusion and outlookConclusion and outlookTwo groups are working on top FCNC inside CMS communityTwo groups are working on top FCNC inside CMS communityParallel work is performed, analyzing both tParallel work is performed, analyzing both t qZ and t qZ and t q q signal. signal. Fast and full simulations results are regularly compared.Fast and full simulations results are regularly compared.

At the present, quite promising resultsAt the present, quite promising results:: large statistic of signal and background is under productionlarge statistic of signal and background is under production signal is clearly understood and quite clean selections are definedsignal is clearly understood and quite clean selections are defined nice agreement in full/fast comparison in most variables (pnice agreement in full/fast comparison in most variables (pTT,,, cos etc.), cos etc.)

To do in the very next futureTo do in the very next future:: several optimization needed: b-tagging algorithms, light jet mistagging several optimization needed: b-tagging algorithms, light jet mistagging etc. etc. increase statistic as much as possible to approach to a HL environmentincrease statistic as much as possible to approach to a HL environment include study of effect of systematics uncertainties (pile-up, UE, jet include study of effect of systematics uncertainties (pile-up, UE, jet fragmentation, PDF etc, see fragmentation, PDF etc, see CMS Note 2005-013CMS Note 2005-013)) include full simulation in all analysesinclude full simulation in all analyses CMS official results with full simulation and systematics will be ready by CMS official results with full simulation and systematics will be ready by April 2006April 2006 ...Intensive work is ongoing!...Intensive work is ongoing!

Top FCNC: preliminary studies in CMS

Documents