Searches for new physics in same-sign leptons final states with the ATLAS experiment Otilia Ducu, supervisors : Calin Alexa, Pascal Pralavorio IFIN-HH Bucharest, CPPM/Universite d'aix-marseille and CNRS/IN2P3 IFIN-HH days 2014, 18 19 December
Same - sign leptons final state 1 / 11 Highly sensitive to Beyond Standard Model physics SUSY, 4th quark generation, same sign top pairs, black holes Searches including leptons smaller branching ratio but it can reach uncovered regions of the phase space or compressed spectra Origin of same sign leptons final state? W / Z / Higgs bosons produced in association with top quarks OR [pairs] of new particles (... )
Analyzes motivation 2 / 11 SUSY able to solve the hierarchy problem in the Standard Model Heavily studies in ATLAS Predicts superparteners for all SM particles ( heavier, spin differs by ½ ) Produced/annihilated in pairs, LSP in the final state, R-parity conserved,.. top Higgs coupling measurement it can reveal hints of new physics Largest coupling : Direct measurement only through tt + Higgs production
Signal regions (SR) definition 3 / 11 SUSY 2 same sign or three leptons, ( b ) jets and missing transverse energy Define five SRs depending on b jet and lepton multiplicity to analyze 8TeV data SR 3b highly sensitive to gluino mediated and RPV scenarios SR 1b targets signatures with 1 or 2 b jest in the cascade decays SR 0b and 2 SR 3 leptons direct squark or models with high lepton multiplicity ATLAS-CONF-2012-105, ATLAS-CONF-2013-007, JHEP 06 (2014) 035
Signal regions (SR) definition 4 / 11 tt + Higgs signal 2 SS leptons, 2 b jets and up to 4 light jets expected nr. events with SS leptons ( H WW ) ~20 @8TeV 20.3fb-1, ~500 @14TeV 100fb-1 Goal l+ define potential SRs for Run-2 optimizing : n Lepton definition pt, track and calo isolations Kinematic cuts meff, HT, ETmiss, mt, # (b) jets,.. Finally two SRs were defined :. 2 q q' and == 1 b-jets relaxed cuts tighter cuts
Signal regions (SR) definition 4 / 11 tt + Higgs signal 2 SS leptons, 2 b jets and up to 4 light jets expected nr. events with SS leptons ( H WW ) ~20 @8TeV 20.3fb-1, ~500 @14TeV 100fb-1 Goal define potential SRs for Run-2 optimizing : Lepton definition pt, track and calo isolations Events Kinematic cuts meff, HT, ETmiss, mt, # (b) jets, Finally two SRs were defined :. 2 and == 1 b-jets relaxed cuts tighter cuts
Background classification 5 / 11 Three main categories similar for SUSY and tt + Higgs Prompt leptons background: tt + W, t + Z, ZZ, WZ, W±W± MC estimation tt+ W σ = 0.23 pb tt + Z σ = 0.21 pb SS leptons b-jets jets ɆT WZ 3l + ν σ = 10.2 pb
Background classification 6 / 11 Three main categories similar for SUSY and tt + Higgs Fake lepton bkg : hadrons misidentified as leptons, leptons from heavy flavor decays, electrons from photon conversions Data driven estimation using a Tight to Loose matrix method Charge flip background : reconstructed charge flipped wrt. original electron data driven estimation
Background validation Compare the agreement between data and background estimation 7 / 11
Results 8 / 11 SUSY final results published in one paper No excess above SM expectation in SR3b and SR3Lep Small excess ( 2.1s ) in SR0b and SR1b A total of 14 SUSY simplified models and one mued model were used to place new or significantly improved ATLAS exclusion limits
Results 9 / 11 tt + Higgs preliminary results obtained for different luminosities at 14 TeV Significance, with no CR SR2b is the most sensitive, but 20% improvement brought by SR1b Several control regions are added to decrease the systematic uncertainty Still room for improvements work in progress
Discussion (1) 10 / 11 Search for new physics in same sign leptons final states Involved in SUSY since ~3 years 2 CONF notes and 1 paper using 8 TeV data Prospects for tt + Higgs coupling measurement @ 13 TeV INT note in preparation IFIN team : Otilia Ducu, Julien Maurer, Albert Olariu My current working plan : Write my thesis I must be ready to analyze first Run -2 data Goal : publish results with first 13 TeV data for summer conferences Complete tt + Higgs / SUSY prospects studies
Discussion (2) 11 / 11 Public talks during 2014 : Pheno 2014 conf. Inclusive searches for squarks and gluinos with the ATLAS detector ( on behalf of the ATLAS collaboration ) GDR Terascale Search for new physics in events with 2 SS leptons, jets and E Tmiss with ATLAS and CMS detectors Egamma workshop Electrons and photons : feedback from SUSY group ( on behalf of ATLAS SUSY group ) Workshop physique ATLAS FR Search for strongly produced SUSY particles with two same sign leptons and jets Posters during 2014 : ATLAS week ( Sibiu, RO ) Search for new physics in events with two same-sign or three leptons using 20.3 fb-1 of pp collisions at 8TeV
Backup
Signal regions (SR) definition 11 tt + Higgs : Finally two signal regions are defined # bjets (pt>20) leptons pt #jets (low pt) SR1b SR2b ==1 2 pt 30 pt 25 5 (pt>30) 5 (pt>25) #jets (high pt) Z veto Orthogonality with BSM searches 2 (pt>40) ee channel: veto 80 < mee < 100 20 < MET< 200, 450 < meff < 1800 20 < MET< 200, mt < 200, meff > 350
Signal regions (SR) definition 12 SUSY 2 same sign or three leptons, ( b ) jets and missing transverse energy
Instrumental backgrounds Reconstructed electron charge flipped with respect to original electron When wrong track used to reconstruct electron Transforms opp-sign into same-sign event Estimated by weighting OSData events by charge flip rate (~ 1%) fake leptons produced in the weak b-hadron decay, conversions, meson decay in-fligh, hadrons faking electrons. Tag a quark in the final state not wanted Estimated from data using mainly lepton isolation (energy of the tracks or calorimeter cells around lepton) 13 Not always well simulated using data-driven methods
Generalities 14 Search for SUSY with same-sign leptons, (sometimes b-) jets and (sometimes) MET Low SM background, fake+non-prompt leptons bkg plays important role in SR NB in the following, using indifferently fake or non-prompt ; we are anyway not making a difference between these 2 in the analysis, being dominated by non-prompt leptons Fully data-driven method used for estimation: the well-known matrix method Start from sample of leptons with loose identification cuts, then split in subsamples depending whether leptons satisfy or not reinforced (=tight) identification+isolation cuts Knowing e/f the probabilities for real/non-prompt loose leptons to satisfy tight cuts, one can relate number of Real/Non-prompt leptons in the SR with number of leptons Passing/Failing tight cuts, by a set of linear equations example for a di-lepton final state (leptons indexed by 1 and 2), build 4 equations: n{p1p2} = e1e2 n{r1r2} + e1f2 n{r1n2} + f1e2 n{n1r2} + f1f2 n{n1n2} n{p1f2} = e1(1-e2) n{r1r2} + e1(1-f2) n{r1n2} + n{f1p2} = (1-e1)e2 n{r1r2} +... n{f1f2} = (1-e1)(1-e2) n{r1r2} + can be written under matrix form n[p/f] = M(e1,e2,f1,f2). n[r/n] Invert matrix M to get n[real/non-prompt] from n[pass/fail] We're using a generalized method that adapts to the number of leptons in the event Parameters e,f are separately measured in dedicated samples enriched in real/fake leptons
Dynamic matrix method 15 Fully data-driven method dynamic matrix method Take an N loose leptons event in data Assign weights for each tight / loose combination separately Event selection: exactly 2 signal SS leptons > 2 signal leptons (no charge requirement) Example: e+e-μ+ an event with 3 baseline leptons Dynamic matrix method output: e+t e-t μ+l TTL fails cuts e+t e-l μ+t TLT 2 leptons SS e+l e-t μ+l LTL fails cuts e+t e-t μ+t TTT > 2 leptons Each event can be used more than once eg. a TLT event TLT signal SS (if all signal cuts are passed) TLT loose leptons for 3Lep MxM