Searches for Physics Beyond the Standard Model @ CMS Sung-Won Lee Texas Tech University for the CMS Collaboration Hadron Collider Physics Symposium, HCP-2010, Toronto, Canada Aug. 23~27, 2010 Sungwon Lee Physics at LHC 2006, Cracow 1
The Standard Model has been enormously successful, but it leaves many important questions unanswered: What is the origin of mass? Why are there three generations? Why the large difference between Plank and EWK scale? How can we incorporate gravity? Are fermions point-like or do they have substructure? What is the source of dark matter? Many theories/models attempt to address these issues. SUSY is most commonly invoked to address these, but there are many other models that seek to answer some or all of these questions The Standard Model of particle physics is a description of the known particles and their interactions Sung-Won Lee What can be (non-susy) BSM? Many possibilities! Extended Gauge Symmetries, Dynamical EWSB, Extra Dimensions, Compositeness, etc (see next slide) 2
{ } Roadmap: Beyond the S.M. H W 1 W 2 W 3 B { W ± Z γ } SUSY 2HDM h H A H ± Heavy, Excited, Composite States t q * u d ( ) ν ( e e ) ~ ~ ~ ~ χ ± χ 0 γ ~ ~ h H H ± u ν e ~ e ( ) H ±± (eg: TRIPLET) Technicolor, Black Holes ~ Extra Dimensions ( ) ( ) SUSY msugra GMSB AMSB RPV(?) Low Scale Gravity Non Commutative Geometry Presented by Rob McPherson (Univ. of Victoria) @ ICHEP 2002, Amsterdam 3 c s ν µ µ ( ) ( d ) ( c ~ ~ s ) ~ ~ ~ ν µ µ ( ) ~ t b ν τ ( ) ~ τ t b ( ~ ) ~ ν τ ( ) ~ τ b ν L Z W l * ν * LQ
{ } Roadmap: Beyond the S.M. H W 1 W 2 W 3 B { W ± Z γ } SUSY 2HDM Heavy, Excited, Composite States t q * u d ( ) ν ( e e ) ~ ~ ~ ~ χ ± χ 0 γ ~ ~ h H H ± h H A H ± u ν e ~ e ( ) H ±± (eg: TRIPLET) Technicolor, Black Holes ~ Extra Dimensions c s ν µ µ ( ) ( ) ( d ) ( c ~ ~ s ) ~ ~ ~ ν µ µ ( ) ~ t b ( ) ν τ ( ) ~ τ t b ( ~ ) ~ ν τ ( ) ~ τ b ν L Z W SUSY msugra GMSB RPV Split SUSY Large Extra Dimension (ADD) Warped Extra Dimension (RSG) Universal Extra Dimension Unparticles, Hidden Valleys, Little Higgs, HSCP, etc l * ν * LQ 4
This talk (Searches for New Physics @ CMS) describes Some of early new physics searches we ve performed via signature-based & compare how well the early LHC data agrees with the S.M. Some of the important theories we ve tested so far Excited Quarks, Axigluon/Colorons, Contact Int. String Resonances SUSY: stau, gloinos, stop, etc Some of new strategy & method we ve developed Sungwon Lee Physics at LHC 2006, Cracow 5
High-Mass Dilepton & Diphoton Resonances Z, e*, µ*, boosted Z, RS G ZZ, γγ, LV ff High-Mass Non-Resonant Signals W, C.I. µν(µµ), LED (γγ,µµ,ee,γ+met), UED, Unparticles LQ Multijets Final States Dijet: Mass, Centrality Ratio, Angular distribution BH, Multi-jet Resonance, Mono-jet, High mass resonance (ZZ/WW) Long-lived Particles HSCP, Stopped gluinos, GMSB γ, Top Pair BSM: Z ttbar Fourth Generation SUSY b b twtw (bzbz, cwcw, bzcw), t t tztz (bwbw), W tb, etc Sung-Won Lee 6
New Physics Searches with Dijet Sung-Won Lee 7
New Physics Search with Dijet events We study the inclusive dijet final state using dijet mass spectrum and the dijet centrality ratio observables. These provide a test of QCD and sensitivity to new physics BSM. Mass Spectrum simple test of cross section vs dijet mass from QCD and PDFs Centrality Ratio detailed measure of QCD dynamics from angular distribution provide most sensitive bump hunt for new particles decaying to dijets because of experimental uncertainties, less sensitive to quark compositeness less sensitive to dijet resonances, but important confirmation that bump is not QCD fluctuation sensitive search for quark compositeness Sung-Won Lee 8
Specific Dijet Resonance Models Parton resonances decaying to dijet are predicted by various theory models Axigluons / Colorons Excited Quarks, E 6 Diquarks RS Gravitons New vector bosons (Zʼ,Wʼ) Recent theoretical development: String Resonances Regge excitations of quarks and gluons cross section higher than q* models by factor of 25 (due to color, spin, chirality effects) String resonances would produce a spectacular bump in the dijet mass spectrum. Sung-Won Lee 9
Events accepted by single jet trigger one jet with E T > 50 GeV uncorrected trigger fully efficient at 220 GeV/c 2 Two Anti-Kt calo-jets (ΔR=0.7) with Δη < 1.3, η < 2.5 jet energy corrected for detector effects (from MC) + 10% systematic Dijet Mass Spectrum spectrum extends to 1.9 TeV with 836 nb -1 The data is in good agreement with the full CMS simulation of QCD from PYTHIA Sung-Won Lee 10
Highest Mass Dijet Events Sung-Won Lee 11
Cross Section Stability Stability vs. run is excellent RMS of cross section vs. run is 3%, constraining JEC stability to 0.5% of better Sung-Won Lee 12
Smooth Fit of Mass Spectrum Data well fit with four parameters χ 2 /ndf = 25.9/25 No indication of new physics String resonances have largest cross section and provide the highest search sensitivity Sung-Won Lee 13
Model-Independent Cross Section Limits We obtain generic cross section limits on qq, qg, gg resonances. The upper limits are compared to the predicted cross sections for 7 models. Model 95% C.L. Mass Limit (TeV) using CTEQ6L CMS 836 nb -1 CDF 1.13 fb -1 ATLAS** 315 nb -1 String 2.10 1.4 --- q* 1.14 0.87 1.20 Axigluon 1.06 1.25 --- E 6 Diquark 0.58 0.63 --- CMS now well beyond the Tevatron for both string resonances and q* CMS competitive with ATLAS q* limits** (arxiv:1008.2461) Sung-Won Lee 14
Model-Independent Cross Section Limits Observed/expected cross section limits on qg, and qq resonances are close!! Type of Limit 95% CL Mass Limit on qg & gg (TeV) String qg q* qg Axigluon qq E6 Diquark qq Observed 2.10 1.14 1.06 0.58 Expected 2.10 1.10 0.98 0.54 Sung-Won Lee 15
Dijet Centrality Ratio Quantifies the centrality of the dijet angular distribution at a given dijet mass Complimentary to the mass spectrum analysis Many experimental uncertainties cancel in the ratio (e.g. absolute jet energy scale, luminosity) roughly flat for t-channel QCD rises for quark C.I. bumps in dijet mass for resonances The data agree well with NLO + non-perturbative corrections Sung-Won Lee 16
New Physics with Dijet Centrality Ratio Comparison of data, QCD, C.I., q* models show no sign of new physics Sung-Won Lee 17
Long-Lived Heavy Particles Sung-Won Lee 18
LLH Particle Signatures @ CMS Long-Lived Heavy particles appear in many BSM Scenarios Some SUSY flavors predict LL gluino, stop, stau Hidden valley models, certain GUTs Life-times 10 2 ~10 3 s are of particular interested in cosmology Two main classes of LLH particles Lepton-like: no strong interactions Hadron-ike: hadronize to form R-Hadrons STOP MC If charged, slow moving particle will loss E more quickly than mip (higher de/dx) some of them will stop in the detector and eventually decay out-of-time w.r.t. beam crossing Two complementary analyses in CMS 1. Search for charged tracks with anomalously high de/dx 2. Search for stopped particles Simulated R-hadron stopping locations Both methods offer high sensitivity with early LHC data Sung-Won Lee 19
Search for Heavy Stable Charged Particles Signature based search look for high p T tracks with high de/dx track+muon (muon-like signature: e.g. mgmsb stau mass ~ 100-300 GeV) track-only (others: e.g. stop & gluino R-hadron, mass ~ 130-900 GeV) Event selection preselect track with p T > 7.5 GeV, relative uncertainty on p T > 0.15, I.P. < 2.5 mm, > 3 Si strip hits apply cluster cleaning split into subsample by η and N hits cut on track p T and de/dx discriminator Selection optimization tight selection for signal box loose selection to cross-check background estimate Sung-Won Lee 20
Results from HSCP Search in 198 nb -1 After loose selection, mass distribution: good agreement b/w data and MC, After tight selection, no events observed in signal region Set 95% C.L. limit on production cross-section for gluinos, stop, stau track + muon Sung-Won Lee 21
Search for Stopped Gluinos Gluinos, bound into R-Hadron, can decay either monojet or dijet Use dedicated calorimeter trigger for non-collision gap between filled bunches during LHC fills observation of signal during these periods will be a sign of BSM physics Q: where will they sop? A: ~20% gluinos can be stopped somewhere in CMS Performed detailed study of backgrounds cosmics & instrumental noise measured with 08/09 cosmic runs beam background observed in 900 GeV & 7 TeV data signal efficiency ~17% (of all R-hadrons) Counting Experiment Results perform counting experiment in lifetime (τ) bins no excess above expected background observed proceed to set 95% C.L. limits Sung-Won Lee 22
Stopping Model-Independent Results Limit on cross section x stopping probability independent of the model of interaction with matter 14 orders of magnitude covered in τ g decays during BX veto: τ < ~100 ns R-hadron decays within the orbit: τ < ~10-4 s decay over the full fill (~104s) slow decays after fill (loose sensitivity) Time-Profile analysis As well as counting experiment, we analyze the distribution of observed event times assuming a life-time < 100 µs, calculate a PDF for signal event time, using luminosity profile background PDF is flat in time, signal peaks at bunch crossing fit the data set 95% C.L. on the signal Sung-Won Lee 23
Stopping Model-Dependent Limits Result translated into a gluino cross-section limit use R-Hadron stopping probability for specific models cloud model, EM only, Neutral R-Baryon e.g. m(g) = 200 GeV & m(χ 0 ) = 100 GeV Included time-profile analysis (dot lines) to improve the sensitivity for τ(g) < 100 ns ~ ~ ~ excluded lifetime range 120 ns < τ(g) < 6 μs for m(g) = 200 GeV ~ also extend D0 result below 30 µs ~ Result translated into a gluino mass limit ~ ~ limit for fixed lifetimes, as function of m(g), m(χ 0 ) ~ ~ fixed m(g)-m(χ 0 ) = 100 GeV ~ ~ ~ no sensitivity below m(g) =150 GeV (efficiency drop) time profile analysis (τ = 200 ns): m(g) < 229 GeV counting exp. (τ = 2.6 μs): m(g) < 225 GeV Sung-Won Lee 24
Commissioning of SUSY Searches Early LHC data preparation for SUSY searches at CMS Sung-Won Lee 25
Discovery Potential & Search Strategy Fully Hadronic Signature SS dilepton 7 TeV data ~100 pb -1 should provide sensitivity to SUSY parameter space beyond current TEVATRON limits Achieving this, sensitivity strongly depends on how well we measure SM backgrounds e.g. Z(νν)+ Jets, W(lν) + X, QCD... how well we understand physics object (lepton,jets,met) Currently, SUSY commissioning focused on specific tools for searches background discriminating variables (e.g. α T, Δφ*, ΔΦ(MHT, MPT)) data-driven strategies for QCD background estimate huge effort going on to understand the SM backgrounds with data (see next few slides) Sung-Won Lee 26
Hadronic Analyses Suppressing QCD with α T α T is a powerful variable for suppress QCD Multijets to hadronic SUSY combines jet p T and angles; complementary to MET MC shows strong QCD suppression with α T > 0.55 in dijet/multijets, and improving with increasing H T. Validated α T behavior in data Suppressing QCD with Δφ* A complementary observable, Δφ*, to diagnose background events one jet mis-measured where confirm expected behavior in both di- and multi-jet: small Δφ* for QCD, more uniform for real MET Sung-Won Lee 27
Lepton+Jets+MET MET background to lepton+jets+met signatures from real MET (e.g. W/Z) & MET due to mis-measurements e+jet+met: select control sample by inverting selection cuts & perform template fit using relative isolation distributions. µ+jet+met: based on an extrapolation from a sideband fit to a functional form Entries / 0.05 250 200 CMS preliminary -1 s = 7 TeV, 53 nb Data Fit result Fit prompt Fit background MC prompt MC background 150 100 Prompt: 251.2 ± (17.9) (M control = 248) stat. T Background: 66.2 ± (11.3) (M control = 72) stat. T 50 e+jet+met predicted observed RI < 0.3 224 ± 13 263 RI < 0.3 && MET > 20 215 ± 13 215 0 0 0.5 1 1.5 2 2.5 Relative isolation µ+jet+met predicted observed Prompt µ MET < 20 251 ± 18 248 Background MET < 20 66 ± 11 72 Sung-Won Lee 28
Same-Sign Dilepton Data driven method for estimating SM background for SS ee, μμ and eμ channels Use a control sample (loose lepton-id & isolation) to measure efficiency of passing all analysis cuts as a function of lepton kinematics Monitor measured Tight-to-Loose-Ratios using different jet-triggered sample. TL ratio 0.4 0.35 0.3 0.25 0.2 0.15 CMS Preliminary -1 36 nb, s = 7 TeV HLT_L1Jet6U HLT_L1Jet10U HLT_Jet15U HLT_Jet30U TL ratio 0.7 0.6 0.5 0.4 0.3 CMS Preliminary -1 36 nb, s = 7 TeV HLT_L1Jet6U HLT_L1Jet10U HLT_Jet15U HLT_Jet30U 0.1 0.05 Electrons 0.2 0.1 Muons 0 10 15 20 25 30 35 (GeV/c) 40 p T 0 10 15 20 25 30 35 (GeV/c) 40 p T Predictions obtained using HLT_Jet15U No discrepancy observed Method in place, more statistics to come Sung-Won Lee 29
Conclusions CMS is searching for evidence of different models of new physics in several channels using the early LHC data & already exploring new territory beyond the Tevatron. Only recent results shown here. Many new physics searches are underway. No signals of the new physics observed in the early LHC data yet. More LHC data on the way; Analyses of 1pb -1 data samples just beginning. New exciting results are in the pipe-line. Stay tuned! Sung-Won Lee 30
Extra Slides Sung-Won Lee 31
Future Prospects Conservative expectations for limits as function of integrated luminosity CMS is exploring new territory beyond the Tevatron string, q*, Axigluon, E6 Diquark resonance mass limits. We expect to surpass Tevatron limit of Λ > 2.8 TeV at 95% C.L. with 4 pb -1. Sung-Won Lee 32
Dijet Resonance Mass Limits 95% C.L. Mass Limit (TeV) using CTEQ6L Model CMS observed 836 nb -1 CMS expected 836 nb -1 CDF 1.13 fb -1 ATLAS observed 315 nb -1 ATLAS** expected 315 nb -1 String 2.10 2.10 1.4*** --- --- q* 1.14 1.10** 0.87 1.20 0.98 Axigluon 1.06 0.98 1.25 --- --- E 6 Diquark 0.58 0.54 0.63 --- --- ** CMS expected limit with 315 nb -1 is 0.93 TeV *** CMS evaluation of string resonance cross section Sung-Won Lee 33