Julien MOREL Fabienne LEDROIT Benjamin TROCME ATLAS Exotic group LPSC - Grenoble

Download Presentation

Julien MOREL Fabienne LEDROIT Benjamin TROCME ATLAS Exotic group LPSC - Grenoble

Loading in 2 Seconds...

- 147 Views
- Uploaded on
- Presentation posted in: General

Julien MOREL Fabienne LEDROIT Benjamin TROCME ATLAS Exotic group LPSC - Grenoble

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Discovery and identification of a new neutral gauge boson in the e+e- channel with the ATLAS detector

Julien MOREL

Fabienne LEDROIT

Benjamin TROCME

ATLAS Exotic group

LPSC - Grenoble

23 August 2006 - Laboratoire René-J.-A.-Lévesque - Montréal

- Introduction and motivations
- The different theoretical Z’ models
- The LHC and the ATLAS experiment
- The ATLAS Z’ discovery potential
- How can we infer the underlying theory ?
- Conclusions and outlook

- It is very well verified
- It makes very good prediction
- Hypothetical particle : Higgs boson
- Lot of parameters
- Divergences
- Number of fermion familly
- The forces are not describe by the same gauge theory

We need to search beyond the standard model

Many theories beyond the standard model predict new neutral gauge bosons (Z’) :

- Grand Unified Theory (GUT)
- Z’y, Z’c, Z’hfrom E(6) and Z’LR from SO(10), CDDT parameterization

- Little Higgs theory
- New gauge bosons come from new gauge groups.

- Almost all theories with extra-dimensions
- New gauge bosons are standard Z/g Kaluza-Klein excitations.

- …

Backgrounds

For our studies

- Hadronic channel

- We focus on the channel

Signal over background ratio very small

- Leptonic channel

- To study the discovery potential and the underling Z’ theory

Small physic background (mainly Z/g process or rare processes)

Mass limit with 200 pb-1

Tevatron ultimate limit

With 2 fb-1, Tevatron Run II can probe up to Mz’≈ 1 TeV

- Introduction and motivations
- The different theoretical Z’ models
- The LHC and the ATLAS experiment
- The ATLAS Z’ discovery potential
- How can we infer the underlying theory ?
- Conclusions and outlook

Grand unified theories

- Based on the existence of a large gauge group including the SU(3)×SU(2) ×U(1) SM gauge group
- Provide a framework for the unification of the SM forces

Extra-dimension theories

Original ADD : [ N.Arkani-Hamed, S.Dimopoulos ,G.Dvali : Phys. Rev D59 086004 (1999) ]

- 4D brane + n compactified X-dim in which only the graviton can propagate
- Provide an explanation of the weakness of gravity

Original RS : [ L.Randall, R.Sundrum, Phys. Rev. Lett. 83 3370 (1999) ]

- 5D bulk with a warped geometry bounded with two 4D brane (Plank and TeV)
- Provide a reduction of the Plank scale on the TeV Brane

Theoretical assumptions and experimental constraints :

- Z-Z’ mixing small (LEP)
- Flavour changing neutral currents constraints
- No Z’ decay into new particles
- Anomaly cancellations

It’s based on the existence of a additional U(1) gauge group :

Up to now, we study GUT Z’ from specificmodels (E6 models : Z’y, Z’h, Z’c SO(10) model : Z’LR)

Carena, Daleo, Dobrescu, Tait (CDDT) propose a model independent parameterisation

[Phys. Rev. D70, 093009 (2004)]

4 classes of solutions are found :

Each model fully described by 3 free parameters :

- Z’ mass
- Coupling strength normalisation gZ’
- An x parameter (fermions coupling related)

These new 4 classes contain :

- The E6 models:
- Some little higgs models
- ...

- Fermions confined on the 4-brane
- Graviton propagates in 4-brane + 1 large extra dimension
- Gauge fields propagate in 1 small extra dimension
- Masses of the KK modes Mn2= M02 + (nMc)2
- Z’ADD = Z / g first KK mode
- (mass degenered)

R >>1 TeV-1

compactified on S1/Z2

R ~1 TeV-1

[T.G.Rizzo : Phys. Rev D61 055005 (2001) ]

Mc is the only parameter

RS with bulk matter :

[ G.Moreau, J. I. Silva-Marcos, Hep-ph/0602155 ]

Planck Brane

- Gauge fields are in the bulk.
- Higgs field remains on the TeV brane.
- Fermions are in the bulk with different localizations along the extra-dimension.
- Z’ gauge coupling non universal

TeV Brane

t

u

H

fermions

W

g

G

Z

γ

- 3 important features :
- New interpretation of the fermion mass hierarchy.
- Compatible with a Grand Unified Theory [hep-th/0108115] .
- KK excitation provides WIMP candidate.

RS model : 1 spatial X-dim compactified over with radius Rc

Fermion 5D masses :

Effective 4D masses matrix:

ci = new dimensionless parameters

kij = new parameters related to the yukawa coupling

Experimental constraints :

- SM charged Fermions masses and mixing angles (5% uncertainty)
- SM neutrino masses and mixing angles (4s)
- Flavor Changing Neutral Current
- S and T parameters

We study two sets of parameters (labeled A and B) :

Point A = Realistic model

Point B = Strong coupling

Grand Unified Theories

- Standard Pythia : process n°141

Extra-dimension theories

- Pythia with an user-defined process developed by T.Rizzo and interface with pythia by G.Azuelos and G.Polesello for the ADD model.
- Pythia with an user-defined process developed by G.Moreau based on G.Azuelos and G.Polesello code for the RS model

These generators provide Z’RS calculation with full interference Z/Z(1)/Z(2)/g/g(1)/g(2)

- Introduction and motivations
- The different theoretical Z’ models
- The LHC and the ATLAS experiment
- The ATLAS Z’ discovery potential
- How can we infer the underlying theory ?
- Conclusions and outlook

The installation of the LHC's magnets is progressing rapidly

The beam pipe closure date will be August 2007

LHC will start in 2007 with 450 GeV per beam

- 7 TeV per beam
- Instantaneous luminosity = 1033 cm-2 s-1 (low lumi)
- = 1034 cm-2 s-1 (high lumi)

2008 :

Inner detector is about to be installed (mid 2007)

Calorimeters are already installed

288 muon Stations have been installed (47%)

Fast simulation :

Simulation using a parameterization of the detector resolutions

Full simulation :

Real simulation of the whole detector using Geant4

- Introduction and motivations
- The different theoretical Z’ models
- The LHC and the ATLAS experiment
- The ATLAS Z’ discovery potential
- How can we infer the underlying theory ?
- Conclusions and outlook

=

Effective cross section

To compute the Z’ ATLAS Discovery potential we need :

- The detector efficiency (e)
- The cross section (sZ’)
- The DY cross section (sDY)
- A significance convention (S12)

According to hep-ph/0204326 we use the significance S12 (realistic) :

We ask |S12| > 5 for a discovery

CMS efficiency (acceptance, trigger, reconstruction) lies in the range 70-75 %

- We use the channel with the ATLAS detector efficiency
- (see next slides)

- We also use the channel for the Z’RS
- with a CMS like detector efficiency inspired from

CMS-NOTE-2005-002

The ATLAS detector efficiency …

Selection criteria :

- 2 identified e±
- 2 e± with |h|<2.5
- Opposite charges
- back to back in the transverse plane

The efficiency of the event selection depends on :

- The di-lepton mass
- The angle between the electron and the beam in the lab frame

The efficiency depend on the model due to the Z’ boost :

dileptons coming from are more boosted than

di-leptons coming from because of different pdfs.

Y

Z

'

SSM

This angular dependence is related to the Z’ boost :

1 model-dependent

combination

(different couplings)

Z’ rapidity: model-independent shapes

The ATLAS detector efficiency …

Selection efficiency vs di-electron mass

For and events separately (low masses):

- All models compatiblefor a given parton flavour
- Efficiency only depends on initial parton flavour (for a given mass)
- Efficiency for events lower than efficiency for

A model-independent method to take into account the efficiency …

Selection efficiency vs di-electron mass

For , , events separately (all masses and all models)

In the effective cross section calculation

We assign the right efficiency depending on the initial parton flavour and the invariant mass, event by event.

3 free parameters in the CDDT parametrization : x , mZ’ and gZ’

MZ’/gZ’ as a function of x for different values of gZ’

CDF exclusion plots

ATLAS discovery plots

[hep-ex/0602045]

Exclusion plots

Discovery plots

Good hope to discover model not yet excluded by cdfin 2008 with atlas

Di-lepton invariant mass in the RS model

According to the G.Azuelos and G.Polesello idea,

to discover a Z’ we are looking for :

- An excess of cross section due to a resonance
- A lower cross section due to a destructive interference

Lack of events

Excess of events

The parameter M1 represent the integration bounds

We chose it model-independent such as :

M1 depend on the luminosity and represents the end of the DY process. We keep 15 events above M1 to allow a S12 calculation with a non-zero background value

M1

We calculate the significance S12 in two regions of the mass spectra :

- In the resonance region
- Above M1

- In the interference region
- Between 500 Gev and M1

Z’RS discovery potential - RS model : Point A

Z’RS discovery potential - RS model : Point A

- We combined :
- the two analyses (interference and resonance)
- the two channels (e+e- and m+m-)

≈3 TeV

≈4 TeV

≈6 TeV

Point A

Z’RS discovery potential

- We can discover up to 3 TeV with 10 fb-1(already excluded)

- We can discover up to 6 TeV or 4 TeV with 300 or 100 fb-1

- We can discover point B up to 10 TeV with 100 fb-1

We have studied the ATLAS discovery potential

Assuming we have 100fb-1 and a Z’ signal

How can we infer the underlying theory ?

Useful observables :

- Total decay width
- Forward-Backward asymmetry

- Introduction and motivations
- The different theoretical Z’ models
- The LHC and the ATLAS experiment
- The ATLAS Z’ discovery potential
- How can we infer the underlying theory ?
- Conclusions and outlook

Strong dependence on model parameter

TeV

TeV

Estimated at tree level

With the formula :

TeV

TeV

Fit of the Z’η invariant mass spectrum

M=1500 GeV

(500 fb-1)

Detector resolution on Mll:

≈9 GeV at 1.5 TeV

≈ 30 GeV at 4 TeV

Fit function for the invariant mass spectrum :

Resonance peak

DY-Z’ interference

DY

GUT

X-dim

- Fully simulated events for GUT models and ADD
- Generated events for RS model

Total decay width

- Well mesured with high accuracy
- The different values provide a model discrimination

*

*

where

*

*

The forward-backward asymmetry is defined by :

backward

forward

*

P

P

q *is the angle between the quark and the electron in the Z’ rest frame

TeV

TeV

On peak asymmetry

computed with only events in the window

[M-4G;M+4G]

M=1.5 TeV

TeV

TeV

Strong dependence on model parameter

Forward-Backward asymmetry – Generated events - GUT

Huge statistic : 6M events

Big deformation of the forward backward asymmetry in the resonance region

Forward-Backward asymmetry – Generated events – X-dim (ADD)

Huge statistic : 6M events

Deformation of the forward backward asymmetry on the resonance

Huge statistic : 12M events

Deformation of the forward backward asymmetry down to ≈ 600 GeV

AFB is a useful observable

AFB is defined with the angle q* between the quark and electron directions

In a pp collider we don’t know the quark direction.

We assume that the Z’ and the quark are in the same direction

At high rapidity : The assumption is good

At low rapidity : We are wrong once out of two

The forward-backward asymmetry is diluted due to this effect

Probability to be wrong when taking the Z’ direction as the quark one

We can correct the diluted forward-backward asymmetry ( )

ATL-PHYS-PUB-2005-010

e = probability to be wrong

Typical spin 1 particle behavior

Detector independent

We lose the angular information

An attractive method consisted in fitting the cos(q) evolution of the AFB

The theoretical behavior is :

Example for the c model ( MZ’=1500 GeV, 1.48 TeV < Mll < 1.52 TeV) :

Diluted

For the c model ( MZ’=1500 GeV) :

- The correction method gives good results
- We are able to reconstruct with good accuracy the forward-backward asymmetry

Diluted

- Introduction and motivations
- The different theoretical Z’ models
- The LHC and the ATLAS experiment
- The ATLAS Z’ discovery potential
- How can we infer the underlying theory ?
- Conclusions and outlook

- We study Z’ from different kinds of models

- Grand Unified Theory

Model independent parameterization

ADD like

RS like

- Extra-Dimension Theory

- The ATLAS discovery potential is high

- Computed using a model independent method to take into account the detector efficiency

- We are able to reconstruct properly useful observables
- for the model discrimination

- The total decay width
- The forward-backward asymmetry

- For the Z’ study

Study other realistic points for the RS model

- For the ATLAS discovery potential

Improve the high energy electron identification

Study the systematic uncertainties due to :

energy scale and linearity

parton distribution functions

radiative corrections

…

- For the model discrimination

Study other observables : Z’ rapidity, BR, …

Study other particles : W’, 2nd KK excitation, …

Backup

If we observe a signal

- We can study :
- The total decay width
- The forward-backward asymmetry

Toward a model discrimination

This implies a lot of statistics

Effect due to resonance

Effet due to destructive interference

ISR ON

Z’RS discovery : Point B

ATLAS

CMS

In our simulations we take into account detector acceptance …

selection efficiency for fully simulated Z’e+e-

Selection criteria:

- 2 e± with |h|<2.5
- 2 identified e±
- Opposite charges
- back to back

Geometrical acceptance increases with mass (boost effect).

Opposite charge selection efficiency decreases with mass.

We have to optimizeelectron identification at very high pT .

Reconstructedevents

Reconstructedevents

Not enough statistic

Need a low luminosity study

Fit of the DY invariant mass between 150 and 600 GeV

(GeV)

Good fit for luminosity equal to few fb-1

A study of the fit parameters may give us informations even at low luminosity

Signal = Z’

background = Drell-Yan ( g/Z MS )

Pythia

Z’RS

We can generate Z’RS events

- We have to be careful with the Z’RS pT and rapidity when the ISR is switched ON.
- The two generators give compatible results for the standard model process.

ATLAS discovery potential goes beyond the LEP limits in most scenarii, already with 400 pb-1

Z’RS discovery : the two channels and the two analyses are combined

≈3 TeV

≈4 TeV

≈6 TeV

Point A

≈4.2 TeV

≈9.5 TeV

> 10 TeV

Point B