Search for supersymmetry with displaced dileptons at the ATLAS experiment Dominik Krauss Max-Planck-Institut f¨ ur Physik March 30, 2017 Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 1 / 11
Search for displaced dileptons Search for massive long-lived particles decaying to two charged leptons ( e or µ ) Experimental signature: Displaced vertices with two lepton tracks Sensitive to lifetimes of about 1 ps to 1 ns Model independent search interpreted in supersymmetric models q ℓ p ℓ ˜ q λ ν χ 0 ˜ 1 χ 0 ˜ ν 1 λ q ˜ p ℓ q ℓ Example of a R -parity violating model Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 2 / 11
Reconstruction of displaced vertices Standard tracking reconstructs tracks up to | d 0 | = 10 mm Additional tracking optimised for tracks up to | d 0 | = 300 mm Secondary vertices reconstructed by standard ATLAS vertexer Tracking and vertexing very resource-intensive → Event preselection based on photon and muon spectrometer triggers 1 Vertex reconstruction efficiency x 0.9 ATLAS Work in progress Local z axis ∼ ~ 0 χ τ 0.8 m( q ) = 700 GeV, m( ) = 500 GeV, c = 100 mm 1 z ∼ 0 χ → µ µ ν 0.7 1 |d 0 | 0.6 y q/p, θ , φ 0.5 z 0 0.4 Reference 0.3 0.2 Point 0.1 Particle 0 Track 0 20 40 60 80 100 120 140 Transverse radius [mm] Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 3 / 11
Displaced vertex selection Displaced vertex with at least two oppositely charged leptons Lepton tracks: p T > 10 GeV and | d 0 | > 2 mm Displacement: 4 mm in transverse plane to all PVs Fiducial volume: Vertices inside detector material are vetoed m DV > 10 GeV Vertex has to pass at least one criterion used to preselect data events Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 4 / 11
Background sources of displaced vertices Number of vertices ATLAS Work in progress Conversions 4 10 s = 13 TeV, t t MC Hadrons Crossings (PV+PV) Crossings (PV+PU) 3 10 Crossings (PU+PU) 2 10 10 1 0 5 10 15 20 25 30 35 40 45 50 Vertex invariant mass [GeV] Plot shows origin of displaced vertices with two tracks in a t ¯ t Monte Carlo sample No leptons required and p T cut on tracks lowered to 1 GeV Random crossing of tracks dominant background for m DV > 10 GeV Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 5 / 11
Hadron decays vertices / 0.2 GeV Number of ee vertices / 0.2 GeV 7 ATLAS Work in progress ATLAS Work in progress 3 10 -1 6 -1 s = 13 TeV, 32.7 fb s = 13 TeV, 32.7 fb 5 2 10 4 µ µ Number of 3 10 2 1 1 0 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 Vertex invariant mass [GeV] Vertex invariant mass [GeV] µµ vertices ee vertices Validation region on data with inverted mass cut and loosened vertex selection Most vertices originate from displaced J /ψ particles of B -hadron decays No dilepton vertex with m DV > 5 . 5 GeV observed → Background from hadron decays negligible Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 6 / 11
Random crossings Unrelated lepton tracks can randomly cross and form a vertex Dominant background of this search Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 7 / 11
Random crossings Unrelated lepton tracks can randomly cross and form a vertex Dominant background of this search Estimation procedure for signal regions: Collect all electrons and muons in data passing our track selection criteria Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 7 / 11
Random crossings Unrelated lepton tracks can randomly cross and form a vertex Dominant background of this search Estimation procedure for signal regions: Collect all electrons and muons in data passing our track selection criteria Randomly select a given number of lepton pairs (“seed pairs”) Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 7 / 11
Random crossings Unrelated lepton tracks can randomly cross and form a vertex Dominant background of this search Estimation procedure for signal regions: Collect all electrons and muons in data passing our track selection criteria Randomly select a given number of lepton pairs (“seed pairs”) Run vertex algorithm on each pair Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 7 / 11
Random crossings Unrelated lepton tracks can randomly cross and form a vertex Dominant background of this search Estimation procedure for signal regions: Collect all electrons and muons in data passing our track selection criteria Randomly select a given number of lepton pairs (“seed pairs”) Run vertex algorithm on each pair Count number of vertices passing vertex selection Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 7 / 11
Random crossings Unrelated lepton tracks can randomly cross and form a vertex Dominant background of this search Estimation procedure for signal regions: Collect all electrons and muons in data passing our track selection criteria Randomly select a given number of lepton pairs (“seed pairs”) Run vertex algorithm on each pair Count number of vertices passing vertex selection Calculate crossing probability p xing = Number of vertices found in this procedure Number of seed pairs Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 7 / 11
Random crossings Unrelated lepton tracks can randomly cross and form a vertex Dominant background of this search Estimation procedure for signal regions: Collect all electrons and muons in data passing our track selection criteria Randomly select a given number of lepton pairs (“seed pairs”) Run vertex algorithm on each pair Count number of vertices passing vertex selection Calculate crossing probability p xing = Number of vertices found in this procedure Number of seed pairs Estimate: Number of lepton pairs in data × p xing Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 7 / 11
Validation of random crossing estimation Validation region: Vertices with two tracks that fail lepton identification Enlarge statistics: No trigger and opposite charge requirements All other vertex selection criteria applied Vertices / 10 GeV 4500 Observation ATLAS Work in progress 4000 -1 Prediction s = 13 TeV, 32.7 fb 3500 VR 3000 2500 1 . 1 × 10 8 Number of pairs 2000 1500 1 . 6 × 10 − 4 Avg. crossing prob. 1000 500 Predicted vertices 17947 0 Data/Estimate 1.5 1 Observed vertices 14775 0.5 0 20 40 60 80 100 120 140 160 180 200 Vertex invariant mass [GeV] Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 8 / 11
Background estimate for signal regions p xing / 10 − 5 N est vx / 10 − 4 SR N ℓℓ 22 +0 . 6 1 . 2 ± 0 . 1 (stat.) +0 . 3 − 8 . 9 (syst.) 0 . 52 ± 0 . 05 (stat.) ± 0 . 13 (syst.) − 0 . 6 (syst.) ee 11 +0 6 . 9 ± 0 . 2 (stat.) +1 . 6 e µ − 2 . 7 (syst.) 6 . 2 ± 0 . 2 (stat.) ± 1 . 4 (syst.) − 2 . 3 (syst.) 5 +0 4 . 9 ± 0 . 1 (stat.) +1 . 1 µµ − 2 . 6 (syst.) 9 . 7 ± 0 . 3 (stat.) ± 2 . 2 (syst.) − 2 . 8 (syst.) Random crossing background is of the order 10 − 4 for all SRs p xing larger in VR due to missing trigger requirement p xing smaller for ee than for µµ (also observed on MC) Total uncertainties on the estimates not larger than 60% Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 9 / 11
Cosmic muons Cosmic muons sometimes reconstructed as a back-to-back muon pair ( η 1 + η 2 ) 2 + ( | ∆ φ | − π ) 2 � Back-to-backness: ∆ R cosmic = Veto cosmic muons in signal regions by requiring: ∆ R cosmic > 0 . 04 Invert cosmic veto to study back-to-backness of cosmic muons: Dimuon pairs 2 10 ATLAS Work in progress -1 s = 13 TeV, 32.7 fb Cosmic rays control region 10 All dimuon pairs (scaled) DV matched dimuon pairs 1 − 1 10 0 0.01 0.02 0.03 0.04 0.05 0.06 ∆ R cosmic Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 10 / 11
Summary Search for displaced vertices with at least two lepton tracks Interpreted in supersymmetric models Dominant background from random crossings of leptons Data-driven estimate of random crossings Background is of the order 10 − 4 for all SRs Potential signal can be identified very clearly in data Dominik Krauss (MPI Physik) Displaced dileptons March 30, 2017 11 / 11
Recommend
More recommend