[PPT] - Search for heavy resonances decaying to long- lived neutral PowerPoint Presentation

SLIDE 1

E Clement 5 Apr 2013

Search for heavy resonances decaying to long- lived neutral particles

Emyr Clement

n behalf of the CMS Collaboration

SLIDE 2

E Clement 5 Apr 2013

Motivation

Many new physics scenarios

predict heavy long-lived particles

➡ Weak R-Parity Violating SUSY ➡ Split SUSY ➡ Hidden Valley Scenario ➡ Exotic decays of recently discovered Higgs boson?

Search for neutral long-lived

particles using CMS

➡ Decay to pairs of displaced electrons or muons ➡ Proton-proton data at √s = 7 TeV ➡ 4.1 - 5.1 fb-1 integrated luminosity ➡ Arxiv link to public result

ν

2

SLIDE 3

E Clement 5 Apr 2013

CMS

3

SLIDE 4

E Clement 5 Apr 2013

CMS

4

Pixel Detector & Silicon Tracker

➡ Reconstruct tracks of charged particles ➡ Measure pt of charged particles

SLIDE 5

E Clement 5 Apr 2013

Reconstruction of Displaced Tracks

CMS exploits an iterative tracking

algorithm

➡ First iterations find tracks originating near the primary vertex ➡ Final iterations find displaced tracks

Can reconstruct displaced leptons

with impact parameters (d0) up to ~ 30 cm

➡ Efficiency to reconstruct track decreases at larger displacements

| [cm] |d

10 20 30 40 50

Tracking efficiency

0.2 0.4 0.6 0.8 1 CMS Simulation Data

5

SLIDE 6

E Clement 5 Apr 2013

Displaced Vertex Signature

Simulation mH = 400 GeV mX = 150 GeV

6

Tracks and vertex are displaced from

centre of CMS

➡ Makes analysis almost background free ➡ Only need to reconstruct one displaced vertex Some heavy resonance ( not necessarily a Higgs ) electron or muon pairs Long-lived particle, X

➡ Travel ~20cm in transverse plane before decaying

SLIDE 7

E Clement 5 Apr 2013

Dilepton Mass Spectrum

Look for a narrow resonance in dilepton mass spectrum

]

2

mass [GeV/c

100 200 300 400 500

Entries

2

10

1

10 1 10

2

10

3

10

XX(350), 1 pb → H(1000) QCD tt µ µ → * γ Z/ τ τ → * γ Z/ WW WZ ZZ Data

1

=7 TeV L=5.1 fb s CMS

µ

+

µ

Expected 0.02+0.09-0.02 background candidates

7

]

2

mass [GeV/c

100 200 300 400 500

Entries

4

10

3

10

2

10

1

10 1 10

2

10

3

10

XX(350), 1 pb → H(1000) QCD tt ee → * γ Z/ τ τ → * γ Z/ WW WZ ZZ Data

1

=7 TeV L=4.1 fb s CMS

e

+

e

Expected 1.38+1.78-1.19 background candidates

SLIDE 8

E Clement 5 Apr 2013

Results

See no significant excess
Set 95% CL upper limit on σ(H→XX)×B(X→l+l-)

➡ Range of H and X masses

Main systematic uncertainty due to displaced tracking efficiency

[cm] τ c

1 10

2

10

3

10

) [pb]

µ

+

µ → B(X XX) → (H σ

2

10

1

10 1 10

2

10

2

= 125 GeV/c

H

m

1

= 7 TeV L = 5.1 fb s CMS

Observed 95% CL limits

2

= 20 GeV/c

X

m

2

= 50 GeV/c

X

m Expected 95% CL limits

[cm] τ c

1

10 1 10

2

10

3

10

) [pb]

e

+

e → B(X XX) → (H σ

3

10

2

10

1

10 1 10

2

10

2

= 400 GeV/c

H

m

1

= 7 TeV L = 4.1 fb s CMS

Observed 95% CL limits

2

= 20 GeV/c

X

m

2

= 50 GeV/c

X

m

2

= 150 GeV/c

X

m Expected 95% CL limits

mH = 400 GeV/c2, mX = 20→150 GeV/c2 mH = 125 GeV/c2, mX = 20→50 GeV/c2

8

SLIDE 9

E Clement 5 Apr 2013

Summary

A search for neutral long-lived particles

➡ Decay to pairs of electrons or muons

No significant excess seen

➡ Set upper limits ➡ mH = 125→1000 GeV/c2 ➡ mX = 20→350 GeV/c2

Limits in range 0.7 fb→100 fb

➡ For X with Lxy in laboratory frame of ~2→100cm

Further information : arxiv.org/abs/1211.2472

[cm] τ c

1 10

2

10

3

10

) [pb]

µ

+

µ → B(X XX) → (H σ

2

10

1

10 1 10

2

10

2

= 125 GeV/c

H

m

1

= 7 TeV L = 5.1 fb s CMS

Observed 95% CL limits

2

= 20 GeV/c

X

m

2

= 50 GeV/c

X

m Expected 95% CL limits

9

SLIDE 10

Backup

SLIDE 11

E Clement 5 Apr 2013

Online Selection

Muon channel

➡ Two muons reconstructed in muon systems only ➡ No primary vertex constraint ➡ No tracker requirement ➡ pt>30 GeV/c

Electron channel

➡ Two clustered energy deposits in the ECAL ➡ Et>38 GeV ➡ No tracker requirement

11

SLIDE 12

E Clement 5 Apr 2013

Detailed Candidate Selection

Selection cuts on track/leptons

➡ High purity tracks, 6 valid hits, |η|<2 ➡ pt of tracks > 33 GeV (muon) > 41 (electron) ➡ d0/σ of tracks > 2 (muon), > 3 (electron) ➡ Tracker isolation

Σpt < 4 GeV within ΔR < 0.3
In electron channel, also match

track to offline supercluster

In muon channel, reject back-to-

back muons and require ΔR between muons >0.2

ΔΦ

PV DV

pll

Selection cuts on dileptons/X

candidates

➡ Good vertex fit

χ2/NDF < 5

➡ No more than one tracker hit in front of vertex ➡ Decay length significance (Lxy/σ) cut > 5 (muons) > 8 (electrons) ➡ Reconstructed candidate momentum collinear with vertex flight direction

ΔΦ<0.2 (muons) <0.8 (electrons)

Lxy

12

SLIDE 13

E Clement 5 Apr 2013

Identification of Displaced Vertex

Select lepton pairs that form a vertex

➡ Our X candidates

Require tracks and vertex to be displaced from centre of CMS

➡ Selection makes analysis almost background free

13

xy

σ /

xy

L

2 4 6 8 10 12 14 16 18 20

Entries

2

10

1

10 1 10

2

10

3

10

XX(350), 1 pb → H(1000) QCD tt ee → * γ Z/ τ τ → * γ Z/ WW WZ ZZ Data

1

=7 TeV L=4.1 fb s CMS

e

+

e

Signal MC in overflow Background at low displacements

SLIDE 14

E Clement 5 Apr 2013

Candidate Selection (II)

14

xy

σ /

xy

L

2 4 6 8 10 12 14 16 18 20

Entries

3

10

2

10

1

10 1 10

2

10

3

10

4

10

XX(350), 1 pb → H(1000) QCD tt µ µ → * γ Z/ τ τ → * γ Z/ WW WZ ZZ Data

1

=7 TeV L=5.1 fb s CMS

µ

+

µ

xy

σ /

xy

L

2 4 6 8 10 12 14 16 18 20

Entries

2

10

1

10 1 10

2

10

3

10

XX(350), 1 pb → H(1000) QCD tt ee → * γ Z/ τ τ → * γ Z/ WW WZ ZZ Data

1

=7 TeV L=4.1 fb s CMS

e

+

e

n-1 selection plot for transverse decay length significance (Lxy/sigma)

➡ Good separation of signal from background

SLIDE 15

E Clement 5 Apr 2013

Candidate Selection (III)

Can relax some selection to increase statistics

➡ Check agreement between data and background MC

15

SLIDE 16

E Clement 5 Apr 2013

Systematic Uncertainties

Most crucial systematic uncertainty, how well

do we understand tracking efficiency for displaced tracks

➡ Main method to study: use cosmic muons ➡ Cosmic muon leaves a track in the muon systems ➡ Is cosmic also reconstructed in central tracker? ➡ Assign 20% systematic uncertainty to account for disagreement between data and MC

Trigger Efficiency Uncertainty

➡ Use standard tag & probe ➡ Assign 11% systematic uncertainty in muon channel ➡ Assign 2.6% systematic uncertainty in electron channel

| [cm] |d

10 20 30 40 50

Tracking efficiency

0.2 0.4 0.6 0.8 1 CMS Simulation Data 16

SLIDE 17

E Clement 5 Apr 2013

Limits

[cm] ! c

1

10 1 10

2

10

3

10

) [pb]

e

+

e " B(X XX) " (H #

3

10

2

10

1

10 1 10

2

10

2

= 1000 GeV/c

H

m

1

= 7 TeV L = 4.1 fb s CMS

Observed 95% CL limits

2

= 20 GeV/c

X

m

2

= 50 GeV/c

X

m

2

= 150 GeV/c

X

m

2

= 350 GeV/c

X

m Expected 95% CL limits

[cm] τ c

1

10 1 10

2

10

3

10

) [pb]

µ

+

µ → B(X XX) → (H σ

3

10

2

10

1

10 1 10

2

10

2

= 200 GeV/c

H

m

1

= 7 TeV L = 5.1 fb s CMS

Observed 95% CL limits

2

= 20 GeV/c

X

m

2

= 50 GeV/c

X

m Expected 95% CL limits

[cm] τ c

1

10 1 10

2

10

3

10

) [pb]

µ

+

µ → B(X XX) → (H σ

3

10

2

10

1

10 1 10

2

10

2

= 400 GeV/c

H

m

1

= 7 TeV L = 5.1 fb s CMS

Observed 95% CL limits

2

= 20 GeV/c

X

m

2

= 50 GeV/c

X

m

2

= 150 GeV/c

X

m Expected 95% CL limits

[cm] τ c

1

10 1 10

2

10

3

10

) [pb]

µ

+

µ → B(X XX) → (H σ

3

10

2

10

1

10 1 10

2

10

2

= 1000 GeV/c

H

m

1

= 7 TeV L = 5.1 fb s CMS

Observed 95% CL limits

2

= 20 GeV/c

X

m

2

= 50 GeV/c

X

m

2

= 150 GeV/c

X

m

2

= 350 GeV/c

X

m Expected 95% CL limits

17