Supersymmetry at the LHC: Searches, Discovery Windows, and Expected Signatures Darin Acosta representing ATLAS & Outline Introduction to SUSY, LHC, and the Detectors Non-Higgs sparticle searches: Î Trigger Strategies Î mSUGRA ÎInclusive squark/gluino searches ÎExclusive sparticle mass reconstruction – neutralino – sbottom, gluino Î GMSB Îstau (heavy lepton) search ÎRadiative neutralino decay and lifetime ÎCascade reconstruction Summary SUSY at the LHC, HCP 2002 2 D. Acosta, University of Florida Minimal SuperSymmetry SUSY Î Symmetry between bosons and fermions ~, ~l q Î Squarks/sleptons: scalar counterparts to the fermions ~± , χ ~0 ~ χ ÎCharginos/neutralinos/gluinos: 1,2 1,2,3, 4 , g fermion counterparts to SM gauge bosons ÎAt least two Higgs doublets (5 scalars): h, H 0 , A, H ± Î Avoids fine-tuning of SM, can lead to GUTs MSSM Î Usually consider RP ≡ (-1)3(B-L)+2S conserved ⇒ LSP is stable Î 105 new parameters mSUGRA: Î Require SUSY to be a local symmetry Î Universal gravitational interactions break SUSY at scale F ~ (1011 GeV)2 Î 5 free parameters Îm0 : Common scalar mass Îm1/2 : Common gaugino mass ÎA0 : Common scalar trilinear coupling Îtan β : Ratio of v.e.v. of Higgs doublets ÎSign(µ) : sign of Higgsino mixing parameter d i d i d i ~f M a g~f > M aq~f > M a χ ~± ≈ M χ ~ 0 ≈ 2M χ ~0 Î Typically: M χ 1 2 1 SUSY at the LHC, HCP 2002 3 D. Acosta, University of Florida Large Hadron Collider (LHC) CMS ATLAS R = 4.5 km E = 7 TeV Two proton rings housed in same tunnel as LEP Design luminosity: L = 1034 cm–2s–1 = 100 fb-1/year (Pile up: ~20 collisions/crossing) Start-up luminosity: L ~ 1033 cm–2s–1 = 10 fb-1/year Completion: mid 2007 SUSY at the LHC, HCP 2002 4 D. Acosta, University of Florida mSUGRA Cross Sections @ LHC q~, g~ m1/2 (GeV) Total cross section 1400 1 fb 1200 1000 10 fb TH 800 100 fb 600 1 pb 400 10 pb 200 EX A0 = 0 , tan β = 35 , µ > 0 0 0 500 1000 1500 2000 m0 (GeV) Î Squark/gluino production dominates total SUSY cross-section for low to moderate m1/2 Î Cross sections don’t vary much with µ, tanβ SUSY at the LHC, HCP 2002 5 D. Acosta, University of Florida Detectors Tile CAL toroids 4T solenoid ATLAS LAr CAL 2T solenoid TRT and Si tracker muon Cu/Scin HCAL PbWO4 ECAL Full Si tracker Compact Muon Solenoid (CMS) SUSY at the LHC, HCP 2002 6 D. Acosta, University of Florida SUSY Signatures Complex squark/gluino decay chains Î Many high-ET jets Î Heavy-flavor (τ and b, especially at large tanβ) Î Leptons Î From sleptons, charginos, W/Z, and b-jets Î Missing transverse energy (MET) Î From LSP and neutrinos from taus, sneutrinos Example: m0 = 1000 GeV m1/2 = 500 GeV tan β = 35 µ>0 A0 = 0 CMS event simulation SUSY at the LHC, HCP 2002 7 D. Acosta, University of Florida Trigger Challenge Reduce 40 MHz bx rate (1 GHz pp) → O(100 Hz) Î Inclusive Jet Rate (cone algorithm, R=0.5): Full GEANT-based detector simulation on QCD background (used in preparation of CMS DAQ Technical Design Report) L=2×1033 Î Expected MET Rate: Recon. MET (hi lumi) Recon. MET (low lumi) Gen. MET (hi lumi) Gen MET (low lumi) Requiring a rate to tape of a ~few Hz implies an inclusive single jet threshold of 400–600GeV, and an inclusive MET threshold of 100–200 GeV Reconstructed MET rate below 100 GeV mainly from calorimeter energy resolution SUSY at the LHC, HCP 2002 8 D. Acosta, University of Florida SUSY Trigger Exercise (CMS) Î Consider several points in the m0-m1/2 plane near the Tevatron reach (most difficult for LHC) Î Consider points with and without Rp conservation ~0 → 3j χ Î For Rp choose most difficult case: 1 Î Run full GEANT-based detector simulation on SUSY signals and SM backgrounds to evaluate trigger performance Î Optimize efficiency for a rate to tape of 3 Hz SUSY at the LHC, HCP 2002 9 D. Acosta, University of Florida Example Trigger Strategy (CMS) Possible jet and MET triggers (at Level-2): (for L=2×1033) • MET >170 GeV • 3 jets > 60 GeV and MET > 110 GeV • 4 jets > 120 GeV • 1 jet > 190 GeV, MET>90 GeV, and ∆φ(j1,j2) < π–0.5 • 2 jets>40 GeV, MET>100 GeV, and ∆φ(j1,j2) < π–0.5 • 4 jets>80 GeV, MET>60 GeV, and ∆φ(j1,j2) < π–0.5 Efficiency for SUSY points: Î ε=0.78, 0.74, 0.54, 0.38, 0.27, 0.17 4 5 6 4R 2nd jet 5R 6R With RP 1st jet Background rate of ~3Hz dominated by QCD For completeness, inclusive lepton triggers are: Î PT(electron)>25–30 GeV , PT(muon)>20 GeV (for L=2×1033) SUSY at the LHC, HCP 2002 10 D. Acosta, University of Florida Fast Simulation Full simulation of signals and backgrounds like that shown for trigger exercise is too CPU intensive for complete SUSY reach determination Î Require O(108) events, but full GEANT simulation takes tens of minutes per event Î Use physics generators + parameterized detector performance ATLFAST: Î Tracks (µ): ∆PT/PT = 0.4PT ⊕ 1% (PT in TeV) Î EM resolution: σ/E ~ 10%/√E ⊕ 0.3% (E in GeV) Î Jet resolution: σ/E ~ 60%/√E ⊕ 2% (E in GeV) CMSJET: Î Tracks (µ): ∆PT/PT = 0.15PT ⊕ 0.5% (PT in TeV) Î EM resolution: σ/E ~ 5%/√E ⊕ 0.5% (E in GeV) Î Jet resolution: σ/E ~ 100%/√E ⊕ 5% (E in GeV) SUSY at the LHC, HCP 2002 11 D. Acosta, University of Florida ~ g ~ Search Inclusive q, Counting excess events over SM background Î Discovery mode SUSY search at LHC Î Explicit mass reconstruction not done 6 Analyses: Î ETmiss: Î Ol: Î 1l: Î 2lOS: Î 2lSS: Î 3l: jets+MET, no lepton requirements no leptons 1 lepton 2 leptons, opposite sign 2 leptons, same sign 3 leptons CMS Study: Lepton identification Î Electron: PT>20 GeV, isolated, |η|<2.4 Î Muon: PT>10 GeV, isolated or not, |η|<2.4 Vary cuts in 6 categories (~104 combinations) Î #Jets, MET, Jet ET, ∆φ(l,MET), Circ., µ Iso. Î Common cuts: ÎMET>200 GeV, ≥2 jets, ETjet > 40 GeV, |η|<3 Optimize S/√(S+B) in a counting experiment Î Probe 500 (m0, m1/2) points Î ~106 signal events, ~108 QCD, tt, W/Z+jets Plot 5σ sensitivity contours SUSY at the LHC, HCP 2002 12 D. Acosta, University of Florida -1 1400 L dt = 100 fb A0 = 0 , tan β= 35 , µ > 0 ~ g(3000) miss ET 0l + 1l + 2l OS -1 (300 fb ) ) h(123 miss ET 1200 ISAJET 7.32 + CMSJET 4.5 m1/2 (GeV) ~ g ~ Reach CMS q, ~ g(2500) 0l 1l 1000 ~ ~ 25 q( q(2 0 ) 00 00) TH ~ g(2000) 2l OS 800 2l SS 3l ~ g(1500) 600 ~ q( ) 00 15 2 h =1 2 4 h = 0. ~ 400 g(1000) 0) 00 q(1 ~ 5 0.1 2 h = ~ ~ q(5 0 200 0) g(500) h(110) EX 0 0 500 1000 1500 2000 m0 ( GeV) Î Jets+MET search gives greatest sensitivity Nucl. Phys. B547 (1999) 60 SUSY at the LHC, HCP 2002 13 D. Acosta, University of Florida m1/2 (GeV) Jets+MET Reach vs. Luminosity 1400 L dt = 1, 10, 100, 300 fb-1 A0 = 0 , tan β= 35 , µ > 0 ~ g(3000) miss ET miss ET 1200 CMS -1 (300 fb ) ) h(123 -1 (100 fb ) ~ g(2500) 1000 ~ ~ 2 q( q(2 50 000 0) ) TH ~ g(2000) miss ET 800 -1 (10 fb ) ~ g(1500) 600 ~ miss 0) 50 1 q( 2 h =1 2 4 h = 0. ET -1 (1 fb ) ~ 400 g(1000) q(1 ~ 0) 00 5 0.1 2 h = ~ ~ q(5 200 00) g(500) h(110) EX 0 0 500 1000 1500 2000 m0 ( GeV) Î Squarks/gluinos probed to ~1.5 TeV with 1 fb-1 Î Up to 2.5 TeV at design luminosity (100 fb-1) SUSY at the LHC, HCP 2002 14 D. Acosta, University of Florida Other Parameter Choices ATLAS TDR 15 CERN/LHCC 99-15 L=10 fb–1 1l tan β = 10, µ < 0 0l 800 SS 600 3l OS 2l,0j 400 3l,0j 200 0 1l 0l 800 tan β = 10, µ > 0 3l SS 600 OS 400 2l,0j 3l,0j 200 0 0 500 1000 1500 2000 m0 (GeV) Î Similar cuts and optimization as for CMS study Î Sensitivity for lower tanβ also derived, but lower mass Higgs inconsistent with present limits SUSY at the LHC, HCP 2002 15 D. Acosta, University of Florida Exclusive Di-Lepton Reconstruction Measure invariant mass distribution of same flavor leptons as evidence for ~0 → χ ~ 0 l + l − or χ ~ 0 → ~l + l − → χ 2 1 2 OS ~ 0l+l− χ 1 ~ 0 can be produced via Drell-Yan χ ~±χ ~0, χ 2 1 2 ~, g~ but more prevalent in cascade decays of q Endpoint in mass spectrum exhibits sharp edge: ~0 → χ ~ 0l+ l− χ 2 1 SUSY e+e− + µ+µ− SM background Events/4 GeV/30 fb −1 ATLAS “Point 4”: m0=800GeV m1/2=200GeV tanβ=10 µ>0 A0=0 L=30 fb–1 M( l+l−) (GeV) This point selected by: Some Z0 from other gauginos SM background is small Î Two OS leptons, PT>(20,10) GeV, |η|<2.5 Î MET>200 GeV, 4 jets: PT1>100 GeV, PT234>50 GeV d i d i c hie c h d ii c h max ~0 − m χ ~0 =m χ 3-body decay endpoint: mll 2 1 2-body: m max = m 2 χ ~ 0 − m 2 ~l m 2 ~l − m 2 χ ~ 0 / m ~l ll 2 1 e d i SUSY at the LHC, HCP 2002 16 D. Acosta, University of Florida ~ ~ Exclusive b, g Reconstruction Completely reconstruct a SUSY decay chain: p l b ~ χ 20 ~ g m ~± l ~ b ~ χ10 l± b p ATLAS Study Î “Point 3” of TDR Îm0=200 GeV, m1/2=100 GeV, tanβ=2, µ<0, A0=0 CMS Study Î Investigate “Point B” of “Proposed Post-LEP Benchmarks for SUSY” Eur.Phys.J.C22 (2001) 535 Î m0=100 GeV, m1/2=250 GeV, tanβ=10, µ>0, A0=0 af c h ~ = 174 GeV md χ i m g~ = 595 GeV, m bL, R = 496,524 GeV 0 2 ~0 Start with reconstructing χ 2 : Î Two OS isolated leptons, PT>15 GeV, |η|<2.4 Î MET>50 GeV SUSY at the LHC, HCP 2002 17 L=10 fb–1 CMS SUSY bkgnd M(e + e - ) + M(µ + µ − ) D. Acosta, University of Florida ~ b Reconstruction sbottom reconstruction: Î Select window around di-lepton endpoint (16 GeV) ~ 0 rest frame with χ ~ 0 at rest Î χ 2 1 F d i GH d i IJ pvdl l i d iK ~ from di - lepton endpoint and Can estimate md χ i ~ ≈ 2m χ md χ i d~ i ~0 m χ 1 v ~0 p χ2 = 1+ m l+l− + − 0 1 0 2 0 1 but analysis not too sensitive to details Î Add most energetic b-jet to reconstruct ~ b Î Eb-jet>250 GeV, |η|<2.4 Î b-jet: ≥2 tracks with IP significance > 3σ Î Require MET>150 GeV Î Require E(ll)>100 GeV Reconstructed mass in reasonable agreement with input (480 vs. 510 GeV) L=10 fb–1 CMS Resolution <10% with assumption on LSP mass (can’t resolve L/R mass splitting, however) Mass (GeV) SUSY at the LHC, HCP 2002 18 D. Acosta, University of Florida ~ g Reconstruction Add another b-jet closest in φ to reconstruct ~g CMS Point B Reconstructed mass in reasonable agreement with input (585 vs. 595 GeV) L=10 fb–1 af ch d i ~ ~0 : m g~ − m b is independent of m χ 1 ATLAS Point 3 Events/4 GeV/10 fb-1 20000 Mass (GeV) CMS Point B Expect 20 GeV Expect 85 GeV 15000 10000 5000 0 0 20 40 60 M(χ2bb)-M(χ2b) GeV 80 100 Mass (GeV) ~ mass Cut around g SUSY at the LHC, HCP 2002 19 D. Acosta, University of Florida Minimal GMSB Gauge Mediated Symmetry Breaking Î Uses SM gauge interactions instead of gravity to break SUSY Î Solves FCNC problem Î SUSY breaking scale much less than mSUGRA scale √F << 1011 GeV Î Particles get mass from SM gauge interactions at a messenger scale Mm ~ O(1000 TeV) << MPl Î n = number of SU(5) messenger fields Λ = F / Mm ~ 100 TeV ~ G is LSP ( m << 1 GeV) ~ ~0 → G γ ( n = 1, low tan β ) NLSP: χ 1 ~ ~ l → Gl (n > 1, high tan β ) Î NLSP lifetime: F 100 GeVIJ FG F IJ cτ ~ 1.3 mG H M K H 1000 TeVK 5 4 NLSP Î cτ >> detector size Î slepton ( ~ τ ) is a long-lived heavy lepton (like µ) Î neutralino leads to MET, like MSSM Î cτ ~ detector size Î Measure NLSP lifetime Î Estimate F Î cτ << detector size Î radiative decay with γ SUSY at the LHC, HCP 2002 21 D. Acosta, University of Florida ~ GMSB Heavy Lepton (τ) Search 1/β Use drift-tube muon systems of ATLAS and CMS to measure time-of-flight for heavy leptons 2 (σ ~ 1ns) 1.8 1.6 1.4 1.2 1 0.8 0.6 0 200 400 600 800 1000 1200 1400 momentum (GeV) Î Measure 1/β and p ⇒ reconstruct mass CMS: particles / 20 GeV Î Require 2 “muons” with PT>45 GeV, M>97 GeV Î |η|<1 for CMS drift-tube system Î Can measure stau mass from 90–700 GeV: 70 GMSB scenario: n=3, tanβ=45, Λ=50-300 TeV, M/Λ=200 114GeV; L=1/fb; eff=5% 60 50 303GeV; L=10/fb; eff=15% 40 30 636GeV; L=100/fb; eff=26% 20 10 0 0 100 200 300 400 500 600 700 800 900 1000 reconstructed mass (GeV) SUSY at the LHC, HCP 2002 22 CMS CR 1999/019 D. Acosta, University of Florida ~ GMSB N1 Lifetime Measurement ~ Look for N1 decays inside detectors: 1) Electromagnetic showers not pointing to vertex ATLAS vertex resolution for H→γγ Events/0.2cm Î Use fine angular resolution from LAr EM calorimeter (ATLAS) and PbWO4 crystals (CMS) 300 σ =1.33 cm 200 100 Î ATLAS: if no non-pointing γ’s in 30 fb-1 ⇒ cτ > 100km 0 -10 -5 0 5 10 z cal o-ztrue (cm) (Λ=90 TeV, M = 500 TeV, n=1) 2) Showers in muon system Î Identify showers with high hit multiplicity σ+(cτ)/cτ CMS: Overall sensitivity to measure cτ: L=143/fb; effkin=10% 10 m(N1)=291GeV ECAL counting 1 µ CAL counting ECAL/µ CAL ECAL impact µ -1 CAL slope COMBINED 10 10 -2 10 -1 SUSY at the LHC, HCP 2002 1 10 23 10 2 10 3 cτ (m) D. Acosta, University of Florida GMSB Cascade Search (CMS) Consider the cascades ~ ~ N2 → lR + l ~ → N1 + l + + l − ~ ~ N1 → lR + l ~ → G + l+ + l− or (NLSP is neutralino) (NLSP is slepton) 100 Λ = 30–90 TeV M =120–300 TeV n=3,4,5 SM GMSB 1 GMSB 2 GMSB 3 GMSB 4 GMSB 5 75 50 Λ=30-200 TeV, M=40-30000 TeV, tanβ=1.5-55, signµ= ±1 25 0 Ñ1 Mass (GeV) Number of lepton pairs/10 fb-1 which will lead to a detectable sharp endpoint in the di-lepton invariant mass as in mSUGRA 1500 -25 -50 0 d 200 400 600 (ll) + − (GeV/c) M e + e − + µ + µ − M− µ e − e+ µ − ee+mm-em inv i edge>0, M/ < 1.8 edge>0, M/ > 1.8 eff=1 1000 eff=0.15 Î MET>200 GeV Î 2 OS same flavor leptons, PT>40 GeV 500 Î PT (jet4)>80 GeV ss s= Ñ1 ˜1 Ma s Ma GMSB 1 GMSB 2 GMSB 3 Probe in 10 fb–1 : m τ~1 < 350 GeV and a f ~ mc N h < 500 GeV GMSB 4 GMSB 5 0 0 200 1 SUSY at the LHC, HCP 2002 24 400 600 800 τ˜1 Mass (GeV) D. Acosta, University of Florida LHC Summary Discovery of SUSY, if it exists, is almost assured at the LHC Î Inclusive mSUGRA squark/gluino discovery reach to 1.5 TeV with 1 fb–1, 2.5 TeV with 100 fb–1 Î Difficult part will be untangling decay chains and measuring masses Possibility to reconstruct squark/gluino/neutralino decays in mSUGRA and GMSB in several prototype analyses Trigger strategies identified for efficient coverage Ability exists to identify heavy leptons in GMSB scenarios, as well as NLSP lifetime in radiative decays Many more exhaustive SUSY studies at the LHC experiments are available: Î ATLAS TDR 15 CERN/LHCC 99-15 Î CMS Note 1998/006 Looking forward to studying SUSY spectroscopy before the end of the decade ! SUSY at the LHC, HCP 2002 25 D. Acosta, University of Florida Can mSUGRA Escape LHC? M.Battaglia et al., Eur.Phys.J. C22 (2001) 535 proposed several SUSY benchmark points in the post-LEP era Two of them would lead to sparticles beyond the reach of the LHC except for a light Higgs Î squark/gluino masses > 2.5 TeV But most other points covered well If SUSY exists, prospects at LHC look favorable SUSY at the LHC, HCP 2002 20 D. Acosta, University of Florida R-Parity Violation Non-conservation of RP ≡ (-1)3(B-L)+2S leads to 3 new terms in SUSY superpotential: W = λ ijk Li L j Ekc + λ ijk ′ Qi L j Dkc + λ ijk ′′ Uic Lcj Dkc Choose most challenging last case of baryon ~0 → 3j number violation: χ 1 ATLAS Study of “Point 5” Î m0=100 GeV, m1/2=300 GeV, tanβ=2, µ>0, A0=300 Î MET is reduced, but still substantial Î Number of jets increases Î Leptons from neutralino decays 0.16 0.14 3 10 0.12 Probability Events/10 GeV/30 fb -1 Should still be able to explore much of the parameter space as with mSUGRA 2 10 0.10 0.08 0.06 0.04 0.02 10 0 0 200 400 600 800 1000 ETmiss (GeV) Figure 20-85 E Tmiss distribution for SUGRA Point 5 in the case of R-parity conservation (shaded histogram) and R-parity violation (empty histogram). SUSY at the LHC, HCP 2002 26 0 2 4 6 8 10 12 14 16 18 20 Njet Figure 20-86 Total jet multiplicity ( p Tjet > 15GeV ) distribution for R-parity conservation (shaded) and Rparity violation at SUGRA Point 5. The jets are reconstructed using a topological algorithm based on joining neighbouring cells. D. Acosta, University of Florida