Techniques and Results of Neutral Long-Lived echniques and Results - PowerPoint PPT Presentation

Techniques and Results of Neutral Long-Lived echniques and Results of Neutral Long-Lived Particle Sear Particle Searches in A ches in ATLAS and CMS in LHC Run 2 TLAS and CMS in LHC Run 2 Claudia-Elisabeth Wulz Institute of High Energy Physics, Vienna For the ATLAS and CMS Collaborations La Thuile, 18 March 2019 Rencontres de Moriond, Electroweak Session

Sour Sources of neutral long-l ces of neutral long-lived part ived particles icles Standard model (SM): neutral B, D and K mesons, neutrons, neutrinos Beyond standard model (BSM): plethora of different models • R-parity violating SUSY • Gauge-mediated SUSY breaking scenarios • Anomaly mediated SUSY breaking scenarios • Split SUSY • Stealth SUSY • Hidden valley scenarios • Dark QED (particularly dark photons) • Dark QCD (particularly dark hadrons) • Dark matter models • Left-right symmetric models (particularly heavy neutrinos) • Axion-like particles (ALPS) • Approximate symmetries Green: covered in this talk • ... Conditions for models with long-lived particles (at least one) • Small phase space - nearly degenerate mass spectra • Small couplings • Highly virtual intermediate states C.-E. Wulz 2 March 2019 Mar ch 2019

Mot Motivat ivation to sear ion to search for new long-l ch for new long-lived part ived particles icles Searches for new long-lived particles (c τ > 1 mm) ongoing for several years • ATLAS/CMS a priori designed/optimized for prompt particles, not new LLP‘s • Clever ideas for triggering / data acquisition / reconstruction / analysis have been and are being developed, in parallel with theory developments … but no signal observed so far! • Preparations for Run 3 and detector upgrades for HL-LHC have strong focus on new LLP‘s Figure by Kathryn Zurek C.-E. Wulz C.-E. W ulz 3 March 2019 Mar ch 2019

Signatur Signature-driven sear e-driven searches ches Figure adapted from Heather Russell Displaced decays • Displaced multitrack vertices photons photons mul multitrack itrack vert vertices ices • Displaced photons • Displaced jets • Emerging jets • Trackless jets, with low emerging emer ging dileptons ileptons, lepton-jets , lepton-jets electromagnetic energy fraction jets jets • Displaced dileptons and lepton-jets trackless, trackless, low-EMF jets low-EMF jets Delayed decays and trapped stable particles • Particles stopped in detector mul multitrack itrack vert vertices ices • Particles trapped in detector, in in muon muon system system e.g. magnetic monopoles • Out-of-time detection possible C.-E. Wulz 4 March 2019 Mar ch 2019

Chal Challenges lenges Physics • Unusual fractions of electromagnetic / hadronic energies in calorimeters • Decays outside usual detectors, e.g. jets in muon system • Unusual, not yet known signatures Trigger, reconstruction and data analysis • Inadequate triggers or triggers with low efficiency • Timing information not always available • Standard object reconstruction often inadequate • Secondary vertex finding algorithms not optimized • Interaction point constraint in triggering / reconstruction not usable • Systematic uncertainties need to be specially estimated • Simulation samples not readily available Backgrounds • In-time and out-of-time pileup • Long-lived standard model hadrons (K L , b, …) • Cosmic rays • Accelerator-related backgrounds (beam halo, satellite bunches) • Electronic noise • Material interactions C.-E. Wulz 5 March 2019 Mar ch 2019

Specific techniques Specific techniques Trigger and data acquisition • Simple requirements, without saturating bandwidth • Trigger on prompt particles in associated production • Trigger on subsequent bunch crossings, or during gaps in bunch trains • Scouting: store only reduced event information, but at high rate • Parking: store full raw data, without immediate processing • Non-standard information, such as timing, added to event record Reconstruction • Avoid that events get rejected at early stages of reconstruction -> check initial basic requirements • Track reconstruction optimized for prompt particles -> need dedicated tracking algorithms • Secondary vertices: b tagging algorithms extended to work better at distances beyond 1 cm • Particle flow reconstruction (CMS) needs to be adapted • Electrons and taus need further development • Spike cleaning in calorimeters must be checked and adapted • Instrumental and non-collision backgrounds from data • Pileup can be useful for low-p T displaced tracks, e.g. from sexaquarks C.-E. Wulz 6 March 2019 Mar ch 2019

Displaced jets Displaced jets arXiV 1811.07991 Benchmark models and interpretations CMS-EXO-18-007 • Jet-jet model: pp -> XX, X -> qq (X = neutral scalar) • SUSY models with LLP , e.g. GMSB model with long- ~ ~ lived gluino, decaying to gluon and gravitino (g -> g G) Signature • Jets with vertices displaced up to 55 cm from primary vertex in transverse plane, reconstructed from energy Displaced Di-Jet deposits in calorimeter towers, with or without MET Dedicated displaced jet trigger • H T > 350 GeV • ≥ 2 jets with p T > 40 GeV, I η I < 2 • ≤ 2 associated prompt tracks • ≥ 1 associated displaced track Background suppression • QCD multijets • Likelihood discriminant from track, jet and vertex information C.-E. Wulz 7 March 2019 Mar ch 2019

Displaced jets Displaced jets arXiV: 1811.07991, CMS-EXO-18-007 Gluino masses up to 2300 GeV excluded for proper decay lengths between 20 and 110 mm -1 35.9 fb (13 TeV) [fb] CMS 5 95% CL upper limits 10 σ Observed Median expected 4 Jet-Jet model m = 50 GeV 10 X m = 100 GeV X m = 300 GeV X m = 1000 GeV 3 10 X 2 10 10 1 − 1 10 2 3 4 1 10 10 10 10 c [mm] τ 0 C.-E. Wulz 8 March 2019 Mar ch 2019

Delayed jets Delayed jets Model assumption and dataset ~ ~ • GSMB SUSY model: g -> gG • Full Run 2 dataset: 137 fb -1 Trigger and signal selection • HLT trigger: MET > 120 GeV • MET + delayed calorimeter jet: 3 ns < t jet jet < 20 ns • Particle flow not used for jet reconstruction due to non-standard tracker component, calorimeter clustering only Jet timing in barrel ECAL CMS-EXO-19-001 • PbWO 4 crystals with Si APDs • Time resolution ≈ 200 ps • Cells within Δ R < 0.4 of jet • t jet jet defined by med defined by median cel ian cell t l time ime C.-E. Wulz 9 March 2019 Mar ch 2019

Delayed jets Delayed jets Backgrounds estimated from control regions with data • ECAL resolution tails • Direct APD hits Satellite bunches • In-time and out-of-time pileup (example profile) • Beam halo • Satellite bunches (2.5 ns steps) • Cosmic muon deposits in ECAL Rejection of main backgrounds through jet cleaning • number of ECAL cells > 25 • electromagnetic / total calorimeter energy fraction > 0.2 • fraction of tracks associated to primary vertex < 1/12 • RMS of t jet 6 10 2016 (35.5/fb) 5 10 2017 (41.8/fb) 10 7 After jet cleaning After jet cleaning 2016 (35.5/fb) 4 10 6 10 2017 (41.8/fb) 2018 (55.2/fb) 5 10 3 10 2018 (55.2/fb) 4 10 2 10 3 10 10 2 10 10 Befor Before jet cleaning e jet cleaning 1 1 − 1 10 20 15 10 5 0 5 10 15 20 − 1 − − − − 10 20 15 10 5 0 5 10 15 20 − − − − t (ns) t (ns) jet jet C.-E. Wulz 10 Mar March 2019 Mar March 2019 ch 2019 ch 2019

Delayed jets Delayed jets • Significantly extended reach in c τ compared to tracker based searches • Gluino masses up to 2500 GeV (2150 GeV) excluded for c τ of 1m (30 m) -1 CMS Preliminary L = 137 fb s = 13 TeV int (GeV) 10 10 (fb) (fb) ~ ~ ~ ~ 3500 -1 pp → g g , g → g + G GMSB NLO+NLL exclusion CMS Preliminary L = 137 fb s = 13 TeV int 4 σ σ 10 Events/0.5 ns 95% CL expected median 1 ± σ 95% CL upper limit on 95% CL upper limit on ~ g Beam halo background m 3 10 Core and satellite backgrounds 95% CL observed 3000 Cosmic background 2 10 Observation GMSB m = 2400 GeV, c = 1 m τ ~ 0 g 1 1 10 GMSB m = 2400 GeV, c τ = 10 m ~ g 0 GMSB m = 2400 GeV, c = 30 m τ ~ g 0 2500 1 − 1 10 − 2 10 2000 1 1 − − 10 10 − 3 10 − 4 10 1500 5 − 10 2 4 6 8 10 12 t (ns) jet 2 2 1000 − − 10 10 2.5 3 3.5 4 4.5 5 log (c /mm) τ 0 10 C.-E. Wulz 11 March 2019 Mar ch 2019

Trackless jets - scalars decaying in calorimeters rackless jets - scalars decaying in calorimeters Models and dataset • Hidden-sector models arXiV: 1810.12602 • 10.8 fb -1 and 33.0 fb -1 at √ s = 13 TeV • c τ range few cm to tens of m Signature and trigger • Displaced jets in HCAL or outer edge of ECAL • At least 2 trackless and low-EMF jets - CalRatio (CR) jets High-E Selection (m , m )=(600,150) GeV s T Φ 0.5 per-event BDT 0.002 • dedicated low- and high-E T triggers ATLAS Simulation A: signal region 0.0018 s =13 TeV 0.4 0.0016 Analysis strategy and background B A 0.0014 T High-E 0.3 • Machine learning techniques (neural 0.0012 network to determine jet origin, BDT 0.001 0.2 0.0008 classifier for jets) 0.0006 • Backgrounds: mainly multijet and 0.1 0.0004 D C beam-induced, estimated with ABCD m Φ = 600 GeV, m s = 150 GeV 0.0002 method 0 0 0 1 2 3 4 5 EXOT-2017-025, arXiV 1902.03094 ∑ ∆ R (jet, tracks) min C.-E. Wulz 12 March 2019 Mar ch 2019