Adrien Renaud (LAL-Orsay IN2P3/CNRS and Université Paris-Sud) for the Atlas collaboration. Search for supersymmetry via resonant final states with the.

Adrien Renaud (LAL-Orsay IN2P3/CNRS and Université Paris-Sud)for the Atlas collaboration.

Search for supersymmetry via resonant final states with the

ATLAS detector

HEP2012 – Valparaiso, Chile January 6, 2012

• PRL (arXiv:1103.5559)• EPJC (arXiv:1109.3089)• EPJCL (arXiv:1110.2693)

2

Overview

1) Introduction 2) RPV tau sneutrino search in the eμ final state3) Scalar gluon search in the four jets final state4) Conclusion

3

Susy via Resonances ??MSSM:- particles with L or B numbers- renormalizable terms violate L or B

Potential disaster:- much too fast proton decay

RP conservation:- PR = (-1)2S • (-1)3(B-L)

- forbids L and B violating terms- PR= +1 (-1) for SM (SUSY) particles- LSP stable, in all cascade decays, DM

No resonances--> look for an excess in SM tail

4

Susy via Resonances ??Beyond MSSM with RP conservation:

Extended supersymmetry: -- minimal ? new pheno ? flavor-violation ? dirac gauginos ? … -- new particles with PR = +1 --> resonances in RP conserved susy !

RP violated susy: -- B and L conservation ? neutrino masses and mixing ? …

-- some of the couplings have to be small --> stable proton

--> Scalar-gluon search

--> RPV tau sneutrino search

5

Introduction to RPV Sneutrino search

Search for RPV sneutrino with lepton number violation decay:

- eμ clean signature with low SM background- previous limit from low energy tau branching ratio:

λ’311 < 0.11 and λ312 < 0.07 for Mslepton = Msquark =

100 GeV

λ’311 ≠ 0 and λ312 ≠ 0

6

Event Selection

Using L=1.07 fb-1: 2011 dataSingle lepton (e, μ) trigger (100±1%)

Electron:-- pT > 25 GeV-- |η| < 1.37 or 1.42 < |η| < 2.47-- isolated in EM calorimeter-- shower shape requirements

Muon:-- pT > 25 GeV-- |η| < 2.4-- reconstructed in ID and MS-- isolated in the ID

- Exactly one muon and one electron with opposite-sign charge

- No requirements on #jets and ETmiss

Signal generated with HERWIG + JIMMY + NLO k-factor

7

Highest invariant mass (662 GeV) eμ event

8

Backgrounds Physics backgrounds (real leptons) -- Z/γ*->tau,tau top pair single top WW, WZ, ZZ -- Estimated using MC corrected for data/MC differences

Instrumental background (lepton faked by photon or jet) -- W/Z+γ estimated using MC -- SM multijet and W/Z+jets estimated using data-driven matrix method: 1) Define loose and tight lepton definitions 2) Apply on all events to get NTT,NTL,NLT,NLL

3) Determine efficiency (r) and fake rate (f) for a lepton that has passed the loose definition to also pass the tight definition 4) Solve the 4*4 matrix:

9

Data / Background

Kolmogorov-Smirnov test probability: 56%Data consistent with absence of new physics

Final observable:

10

Limits

Excluded @ 95% CL: - 135 fb @ 100 GeV - 4.5 fb @ 1 TeV

Excluded @ 95% CL: - 1.32 TeV for λ′311 = 0.10 and λ312 = 0.05 - 1.45 TeV for λ′311 = 0.11 and λ312 = 0.07

Bayesian method (uniform prior for signal cross section):

Best limits for: - Msneutrino > 270 GeV

11

Introduction to scalar-gluon search

Non minimal realisations of SUSY:• Extended SUSY: hybrid N=1/N=2• Extended R-Parity: MRSSM

Sgluon = Scalar color-octet with SM-like PR

Production:

Decay:

Wojciech Kotlarski @7TeV

Pair production

Single production

A 4-jets final state:

Diff cross section from PL.B672,2009implemented as external process to PYTHIA

12

Analysis strategy

4 jets pT > 0.55 * Msgluon

ΔRjj < 1.6

|Cos(θ*)| < 0.5 |M1–M2|/(M1+M2) <

0.075Look for an excess in the (M1+M2)/2 distribution.

Combinatory: Event selection:

Using L=34 pb-1: 2010 dataMultijet trigger: 4 jets pT > 55 GeV --> Low threshold = low mass~same amount recorded in 2011 for this threshold

Scalar production at rest suppressed by factor β -- > “ Slightly Boosted ” regime

Minimize :

| ΔRij – 1 | + |ΔRkl - 1|

Where ijkl are the 4 leading jets

13

Multijet Event

14

DATA-driven BGR estimation

ABCD method:• NA = NB * NC / ND

• Take shape of final

observable in region B via a fit:

JHEP09(2011)074

Only Simulation

BackgroundBackground estimationSignalBackground + Signal

|M1

–M2

|/(M

1+

M2

)

|Cos(θ*)|

C D

BA

0.5 0.7

7.5%

15

ABCD results (1/2)

Prediction within1-2 sigma stat

Reasonably good fit

Good agreement A/B

Agreement between data and background prediction:

16

ABCD results (2/2)

Systematic uncertainties

17

Limits

Profile likelihood ratio and CLS approach :

Likelihood is the Product of : -- Poisson for each bin (shape analysis) -- Gaussian for each systematic uncertainty

Contamination in ABCD method. Correlations between systematic.

Excluded σ @ 95% CL :1 nb @ 100 GeV200 bp @ 200 GeV

Excluded @ 95% CL:• Sgluon : 100-185 GeV (except 5 GeV around 140 GeV)

L = 34 pb-1:

18

ConclusionSearch for tau sneutrino decaying to eμ: -- limits for sneutrino mass vs coupling in RPV (0.1 – 2. TeV)

Search for pair-produced sgluon decaying to four-jets: -- limits for low mass colored scalar (100 – 200 GeV)

Other interpretations already available (LFV Z’, hypercolor).More to come with higher Luminosity, Energy, refined analysis

Already 3 published papers

Funny link between 2 analysis:

- slightly modified analysis to search for RPV stops !

qq

Λ’’313

19

Spare

20

The ATLAS detector

21

Jet reconstruction

Jets : -- Anti-kt R=0.6

-- EM+JES calibration

-- pT > 20 GeV && |η| < 2.8

JER

JES

22

RPV sneutrino search ET

miss not used in the analysis to make the search more generic

23

Other interpretations

Z′ gauge bosons with lepton flavor

violating (LFV) interactions

--> Same analysis as for stau

sneutrino

Hypercolor model:

Hyperpion, scalar color octet

--> Same as analysis as sgluon

24

Data / MC (1/2)

ALPGEN+HERWIG+JIMMYSM-multijet production

after pT cut(4 jets pT>55 GeV -> M=100GeV)

1) 4 jets pT > 0.55 * Msgluon

--> sliding cut

2) ΔR(jj) < 1.6

25

Data/MC (2/2)

After pT cut and ΔR cuts

After all cuts but cos(θ*)

- Reasonably good description- Ratio compatible with 1 even without JES uncertainty - large MC stat uncertainty

3) |M1–M2|/(M1+M2) < 0.075

4) |Cos(θ*)| < 0.5

26

Data / Background

Good agreement with SM prediction