Semi-leptonic and Dileptonic Top-Quark Decays at ATLAS Raphael Mameghani IMPRS/GK Young Scientist Workshop at Ringberg 23rd July 2007 1/ 25
Outline 1 Top pair production at LHC & ATLAS 2 Ratio of semileptonic and fully leptonic decays Semileptonic channel 3 Dileptonic channel 4 Summary & outlook 5 2/ 25
Large Hadron Collider pp accelerator with 27 km circumference superconducting magnets √ s = 14TeV design luminosity: 10 34 cm − 2 s − 1 (start: 10 33 cm − 2 s − 1 ) bunch crossings every 25 ns 3/ 25
ATLAS width: 44m, height: 22m, weight: 7000t together with CMS the universal detectors of LHC toroid magnet containing the muon system solenoid magnet for the inner detector 4/ 25
Top Pair Production at the LHC Mainly (87%) gg → t ¯ t : q → t ¯ But also (13%) q ¯ t : Expectation: σ ( pp → t ¯ t ) ≈ 830 pb (NLO) � L dt = 10 fb − 1 8 · 10 6 events per year at 5/ 25
Top Pair Decays t → W + b ( ≈ 100%) W → e, µ , τ + ν (each 1/9) W → q ¯ q (2/3) combinatorics ⇒ 5%: 30%: signature 2 jets 4 2 thereof b-jets 2 2 char. leptons 1 2 missing E T 1 6/ 25
Reasons for a Ratio Measurement A cross section ratio might compensate Experimental Uncertainties � L dt ) luminosity (as N = σ · (energy and momentum scale uncertainties which might affect counting efficiencies) Theoretical Uncertainties parton density functions unknown effects of higher order But: two channels 7/ 25
Branching Ratio Fully Leptonic / Semileptonic Standard Model expectation R ℓℓ/ℓ = BR ( t ¯ t → 2 ℓ + 2 ν + 2 q ) t → ℓ + ν + 4 q ) = N ℓℓ / N ℓ = 1 / 6 BR ( t ¯ discrepancies might occur due to rare top decays, e.g. H + + b t → ⇒ deficit of electrons and muons H + → τν , ¯ cs Previous Examination (ATLAS Design Report) � L dt = 10 fb − 1: statistical precision for 1 year with ∆ R ℓℓ/ℓ / R ℓℓ/ℓ ≈ 0 . 5 % Smeared 4-vectors with the cuts p T ( ℓ ) > 20GeV, E / T > 20GeV, min. 2 b-jets with p T > 20GeV Here with full detector simulation and no b-jet identification 8/ 25
Semileptonic Channel Semileptonic Channel 9/ 25
Signal & Background Semileptonic Signal ( ≈ 250pb) t ¯ t → ℓ ν + 4 jets , with ℓ = e , µ (MC@NLO) Other No All Hadronic t ¯ t Background ( ≈ 210pb) t ¯ t → ℓ ν ℓ ν + 2 jets , τ + X + 2 jets (MC@NLO) hadronic τ decays look quite similar to jets in the detector, in the following assumed that τ identification is not working W + jets Background ( ≈ 790pb) q ¯ q → W → ℓ ν + QCD jets ALPGEN & HERWIG irreducible contribution (apart from kinematics) 10/ 25
QCD Background I Only ATLFAST simulation available (Generators: Alpgen + Pythia) # Alpgen partons cross sections [pb] 3 4 766 000 Cross sections: 4 549 000 5 64 000 6+ 30 000 Fake Leptons/Electrons A jet may be reconstructed as an electron / T from limited energy resolution Together with E ⇒ Background for semileptonic channel Would need 100 million fully simulated events! Fake jets manually: P(jet → e) = 10 − 3 = const. assumed 11/ 25
QCD Background II partons events fake electron events factor 3 225 834 402 076 1.78 4 200 000 508 064 2.54 5 430 106 1 466 860 3.41 6+ 446 900 2 249 220 5.03 σ eff = σ · P ( j → e ) · factor No energy dependence of partons σ eff [pb] P(j → e) considered here 3 8483 4 1394 Correlations as each jet in 5 218 event is used once as an 6+ 151 electron 12/ 25
Cut Flow n events 1: 1 ℓ P T > 20 GeV, | η | < 2 . 5 t t → l ν 4j 10 10 7 7 2: 3 jets > 40 GeV + 1 jet > 20 GeV t t → l ν l ν 2j, τ X W → l ν , τ ν + 4j (all | η | < 2 . 5) 6 6 10 10 QCD fake l 3: E / T > 20 GeV 5 5 10 10 in 3 jets with max. ∑ � 4: P T : | m 2 jets − m W | < 10 GeV 4 4 10 10 5: m total < 900 GeV 3 3 lead. 3 jets | cos θ ∗ | ’s < 0 . 7 10 10 6: 0 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 t ¯ red: t semileptonic 9: without cuts 5 & 6 brown: other no all hadronic t ¯ t Cuts on missing transversal energy blue: W + jets and number of jets are the most yellow: QCD with fake important ones electrons 13/ 25
Hadronic Top Hadronic top mass: Mass of Hadronic Top after Preselection Events Events 7000 7000 t t l 4j → ν 3 jets with highest vec- t t → l ν l ν 2j, τ X 6000 6000 tor sum P T W → l ν , τ ν + 4j ( t ¯ t tend to recoil) 5000 5000 QCD fake l But do not take the 4000 4000 event if no pair within 3000 3000 the 3 jets has an invari- 2000 2000 ant mass less than 10 GeV away from the W 1000 1000 mass! 0 0 0 0 50 100 150 200 250 300 350 400 450 500 50 100 150 200 250 300 350 400 450 500 M M / GeV / GeV top top 14/ 25
Leptonic Top With mass constraint two solutions for neutrino p z W = ( E ℓ + E ν , T ) 2 − ( p ℓ, x + p ν , x ) 2 − ( p ℓ, y + p ν , y ) 2 − ( p ℓ, z + p ν , z ) 2 M 2 Take smaller solution Mass of Leptonic Top after Preselection Top Mass Difference after Preselection 14000 14000 Events Events Events Events 6000 6000 t t → l ν 4j t t → l ν 4j t t l l 2j, X 12000 12000 → ν ν τ t t → l ν l ν 2j, τ X 5000 5000 W → l ν , τ ν + 4j W → l ν , τ ν + 4j 10000 10000 QCD fake l QCD fake l 4000 4000 8000 8000 3000 3000 6000 6000 2000 2000 4000 4000 1000 1000 2000 2000 0 0 0 0 0 0 50 100 150 200 250 300 350 400 450 500 50 100 150 200 250 300 350 400 450 500 -200 -150 -100 -200 -150 -100 -50 -50 0 0 50 50 100 100 150 150 200 200 M M / GeV / GeV (M (M - M - M ) / GeV ) / GeV top,lept top,lept top top top,had top,had Invariant mass from neutrino vector + Difference between hadronic and lep- lepton + remaining jet with highest P T tonic top mass 15/ 25
Missing Transverse Energy before & after all Cuts Missing E / GeV Missing E (after cuts) / GeV T T Events Events Events Events 600 600 t t l 4j t t l 4j → ν → ν t t l l 2j, X t t l l 2j, X → ν ν τ → ν ν τ 6 6 10 10 500 500 W → l ν , τ ν + 4j W → l ν , τ ν + 4j QCD fake l QCD fake l 400 400 5 5 10 10 300 300 200 200 10 10 4 4 100 100 0 0 0 0 20 20 40 40 60 60 80 80 100 100 120 120 140 140 0 0 20 20 40 40 60 60 80 80 100 100 120 120 140 140 missing E missing E missing E missing E T T T T 16/ 25
Top Masses after all Cuts Mass of Hadronic Top Mass of Leptonic Top Events Events Events Events t t l 4j t t l 4j → ν → ν 2500 2500 t t l l 2j, X t t l l 2j, X → ν ν τ → ν ν τ 2000 2000 W → l ν , τ ν + 4j W → l ν , τ ν + 4j 2000 2000 QCD fake l QCD fake l 1500 1500 1500 1500 1000 1000 1000 1000 500 500 500 500 0 0 0 0 0 0 50 100 150 200 250 300 350 400 450 500 50 100 150 200 250 300 350 400 450 500 0 0 50 100 150 200 250 300 350 400 450 500 50 100 150 200 250 300 350 400 450 500 M M / GeV / GeV M M / GeV / GeV top top top top 17/ 25
Without Cut on Missing Transverse Energy Missing E (after cuts) / GeV Mass of Hadronic Top T 1000 1000 Events Events Events Events 3500 3500 t t → l ν 4j t t → l ν 4j t t → l ν l ν 2j, τ X t t → l ν l ν 2j, τ X 3000 3000 800 800 W → l ν , τ ν + 4j W → l ν , τ ν + 4j 2500 2500 QCD fake l QCD fake l 600 600 2000 2000 1500 1500 400 400 1000 1000 200 200 500 500 0 0 0 0 0 0 50 100 150 200 250 300 350 400 450 500 50 100 150 200 250 300 350 400 450 500 0 0 20 20 40 40 60 60 80 80 100 100 120 120 140 140 missing E missing E M M / GeV / GeV top top T T Hadronic top mass & missing E T without E / T cut QCD background increases by ≈ 1 order of magnitude 18/ 25
Dileptonic Channel Dileptonic Channel 19/ 25
MC Samples Leptonic t ¯ t Signal ( ≈ 40pb) t ¯ t → ℓ ν ℓ ν + 2 jets , with ℓ = e , µ Again no b quark identification info used in my analysis Other No All Hadronic t ¯ t Background ( ≈ 420pb) t ¯ t → ℓ ν + 4 jets , τ + X + 2 jets (5200 sample) Assume that τ identification is not available Z Background ( ≈ 5000pb) Z → ℓℓ + jets from hard interaction Diboson Background ( ≈ 35pb) ZZ , WZ , WW → leptons + jets from hard interaction 20/ 25
W + jets Background I W decay cross sections [pb] 17 740 e ν Cross sections: 17 740 µν τν 17 170 Fake Electrons A jet may be reconstructed as an electron E / T from neutrino ⇒ Background for di-leptonic channel Manually: P(jet → e) = 10 − 3 = const. assumed Random charge ± e assigned to fake electrons 21/ 25
W + jets Background II W decay events fake electron events factor e ν 347 500 342 285 0.985 215 200 153 149 0.712 µν τν 98 007 85 976 0.877 σ eff = σ · P ( j → e ) · factor No energy dependence of W decay σ eff [pb] P(j → e) considered here e ν 17.2 Correlations as each jet in 12.4 µν event is used once as an 15.1 τν electron 22/ 25
Recommend
More recommend
