La fisica del quark top Fabrizio Margaroli The last discovered quark • Top quark was needed to complete the third generation of quarks • 20 years long search, spanning 2 generations, 3 colliders, 2 continents • Discovered in 1995 at Fermilab by CDF and D0 collaborations Fabrizio Margaroli Getting into XXI century physics! Sapienza Università di Roma-INFN Roma 1 2 The last discovered quark • The Standard Model is a highly predictive theory: we knew (approximately) the top quark mass far in advance Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 3 The last discovered quark • The Standard Model is a highly predictive theory: we knew (approximately) the top quark mass far in advance Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 4 The last discovered quark Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 5 The top quark is special • Since 1995, the most massive elementary particle, almost as massive as a gold atom (gold atom = 79 protons + 118 neutrons + 79 electrons) ! ! ! ! ! ! ! ! ! • Yes even after the Higgs discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 6 The top quark is special • Since 1995, the most massive elementary particle, almost as massive as a gold atom (gold atom = 79 protons + 118 neutrons + 79 electrons) • Thanks to its large mass, the top quark has the largest coupling to the Higgs boson ! ! ! ! ! ! • Yukawa(top) = √2 Mtop/vev (vev=246GeV) = 0.996±0.006 • closeness to one leads to speculation about possible special role of the top quark in the electroweak symmetry breaking mechanism ! Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 7 How to produce it • Top quarks produced once in 10^10 collisions at Fermilab (1/10 000 000 000) • Top quarks produced once in 10^8 collisions at LHC Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 8 Top quark decays • top lifetime τ t = 1 / Γ t 5 × 10 −25 s τ had = 1 / ΛQCD 3 × 10 −24 s • only quark that decays before hadronizing → no top hadron spectroscopy • top quark width = 1.3 GeV much smaller than experimental resolution • BR(t→Wb)=99.8%∝|Vtb|2 (where Vtb is inferred from the knowledge of the other elements of the CKM matrix, and make use of unitarity) Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 9 Identifying top quarks final states • I assume you know from earlier classes how to identify a charged and neutral leptons, quarks (jets). Here you need one more ingredient: identifying jets originating from b-quarks • b-quarks form B mesons, that have lifetime long enough to be able to detect its decays through the reconstruction of displaced tracks with large impact parameter (IP), giving rise to a secondary vertex (SV) ! ! ! ! ! ! ! • typical efficiencies: 40% per Tevatron, 80% LHC. Mistags rates can be <<1% Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 10 Things are just getting complicated • Reminder: BR(top→bW)=99.8% • BR(W→qq)=66.6%, BR(W→leptons)=33%, so final state depends on how the two W decay. • If both W bosons decay leptonically “dilepton” channel Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 11 Things are just getting complicated • BR(top→bW)=99.99% • BR(W→qq)=66.6%, BR(W→leptons)=33%, so final state depends on how the two W decay. • If only one W decays leptonically and the other hadronically, “semi-leptonic decays” Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 12 Things are just getting complicated • BR(top→bW)=99.99% • BR(W→qq)=66.6%, BR(W→leptons)=33%, so final state depends on how the two W decay. • If both W bosons decay hadronically, then “all-hadronic” final state Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 13 In case you are not good at math... Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 14 How to identify a collision giving a top pair t→Wb→qq/lnu b thus for each top quark in the event: • identify the b jet • identify the other jets/identify a lepton and/or neutrino Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 15 How to identify a collision giving a top pair t→Wb→qq/lnu b thus for each top quark in the event: • identify the b jet • identify the other jets/identify a lepton and/or neutrino Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 16 How to identify a collision giving a top pair t→Wb→qq/lnu b thus for each top quark in the event: • identify the b jet • identify the other jets/identify a lepton and/or neutrino Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 17 How to identify a collision giving a top pair t→Wb→qq/lnu b thus for each top quark in the event: • identify the b jet • identify the other jets/identify a lepton and/or neutrino Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 18 Backgrounds • Being a “small” signal -> σ(ttbar)=10-10 σ(total), so anything end up looking just as top • Of course, it depends on what signature you are looking to Dilepton: 2 opposite sign leptons, jets, missing transverse energy Dilepton Physics background Instrumental background Fabrizio Margaroli Drell-Yan+jets fake leptons(QCD) Sapienza Università di Roma-INFN Roma 1 19 Backgrounds • Being a “small” signal -> σ(ttbar)=10-10 σ(total), so anything end up looking just as top • Of course, it depends on what signature you are looking to Lepton+jets: 1 charged lepton, large missing transverse energy, jets Physics background Instrumental background Fabrizio Margaroli Dilepton Lepton+jets All-hadronic Drell-Yan+jets W+jets QCD bb+jets fake leptons(QCD) fake leptons(QCD) Sapienza Università di Roma-INFN Roma 1 QCD qq+jets 20 Backgrounds • Being a “small” signal -> σ(ttbar)=10-10 σ(total), so anything end up looking just as top • Of course, it depends on what signature you are looking to All-hadronic: many jets, at least one b-jet Physics background Instrumental background Fabrizio Margaroli Dilepton Lepton+jets All-hadronic Drell-Yan+jets W+jets QCD bb+jets fake leptons(QCD) fake leptons(QCD) Sapienza Università di Roma-INFN Roma 1 QCD qq+jets 21 How to distinguish signal from background • After you require jets, leptons and missing transverse energy, S/B 1/5 (lepton+jets) • remember, top is heavy, you need 2XMtop = 350GeV to produce a pair • backgrounds have lower (MW,MZ) or none (QCD) thresholds - low objects Pt ! ! ! ! ! • Ht=scalar sum of transverse momentum of all final state objects • etamax=max eta of the jets • b-jets are much rarer than light flavor jets Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 22 One more step in complexity • Sometime life is harder (all-hadronic channel) and you have to combine the statistical power of several observables into one - likelihood ratios, neural networks, boosted decision tree, etc. Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 23 Ok let’s assume you have good S/B now... • ...how do you know it’s really a top quark? • start pairing things in order to get the right resonances. How to do this? • Depends on the final state • Remember the flow: tt→bWbW→bqq(lnu)bqq(lnu) • Let’s pick tt→bWbW→blnubqq • One charged and a neutral lepton uniquely identify a W boson • If you identify both b quarks, you have two b’s, one matches with the leptonic W. • If you identify only one b quark, larger number of combinations • What to do if you have so many combinations? Pick the one that seems most “reasonable” Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 24 How do you know it’s really a top quark? • Choose the combination where the dijet invariant mass most reasonably match a W boson invariant mass, the lnu system most reasonably matches the other W, and the b quarks rearrange in such a way as to have 2 Mtop distributions reasonably similar. In other words: • choose the value of the top mass that minimizes a chi square Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 25 Some W and top invariant masses • From O(10) evts for discovery, [email protected] Fabrizio Margaroli to O(1000) for precision, [email protected] Sapienza Università di Roma-INFN Roma 1 to O(10000) pp@7TeV 26 Anatomy of a peak • Peak is a Breit Wigner. Then distortion by PDF. Add instrumental resolution effects. Top with combinatorial background, and extra jets coming from initial and final state radiation ! ! ! ! ! ! ! ! • Question, if you approximate the peak with a gaussian, sigma is 27GeV, you have 30 evts with negligible background, what uncertainty do you expect to measure on Mtop? Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 27 Mtop summary Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 28 Mtop summary Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 29 Systematics • Any idea? • Let’s keep it simple: everything you measure has an uncertainty. • There are experimental uncertainties: how often you identify a lepton, jet, b-jet, etc. Also, their energy/momentum. • There are theoretical uncertainties: how much you (the theoretician?) understand the underlying physics. • Factorize problems: uncertainty on the perturbative part of the process (qq→tt) and uncertainty in the non-perturbative part (q→jet) Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 30 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 31 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 32 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 33 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 34 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 35 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 36 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 37 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 38 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 39 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 Range on x/y axes now 25% smaller! 40 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 Range on x/y axes now 25% smaller! 41 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements Continuous improvements in Mt and Mw lead to increasing constraints on MHiggs. ! Also, SM consistency check in case of discovery ! Mhiggs ∼90±30GeV Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 Range on x/y axes now 25% smaller! 42 From the top (and W) mass, to the Higgs • Just as the value of Mtop was predicted from other SM observables, Mhiggs can be predicted by using existing precision measurements ! ! ! Continuous improvements in Mt and Mw lead to increasing ! constraints on MHiggs. ! ! SM consistency check in Also, case of discovery ! ! Mhiggs ∼90±30GeV ! Range on x/y axes now 25% smaller! ! ! • Same game can be played in non-SM scenarios. Discovered particle still consistent with BSM Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 43 A lot to be measured about the top Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 44 Up in energy: from Tevatron to LHC • Fabrizio Margaroli why W and Z boson cross section increases by a factor of 5, while ttbar a factor of 25 in going from [email protected] to pp@7TeV? Sapienza Università di Roma-INFN Roma 1 45 Top rediscovered at the LHC • Comparison in top production rates at the Tevatron (10years [email protected]) and LHC (1 year pp@8TeV) ! TeV LHC@8TeV Lumi/hour (pb-1/h) 0.4* 14* ! sigma(pb) 7.5 220 ! Nevts top/h 3* 3000* Lumi total 10 16 Nevts top total ? ? ! ! ! ! ! *numbers refer to Tevatron and LHC best stores respectively ! The LHC is a top factory! plus b-tag efficiency is twice larger, and ATLAS/CMS have a factor of two larger eta coverage for leptons/jets Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 46 What to do with all those tops? Differential cross section measurements Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 47 Interesting processes involving top Top (@7 TeV) σ 1,000,000 fb 100,000 fb 10,000 fb 1,000 fb 100 fb tt t ttɣ ttW ttZ ttH 10 fb • Probe strong interactions in the top sector (qq→gluon→ttbar) • Probe electromagnetic intearactions in the top sector (qq→ttɣ) Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 48 Single top quark production • Three main categories of electroweak single top quark production, each one is interesting in its own way s-channel t-channel Wt channel Collider s-ch t-ch Wt Tevatron LHC@7TeV 1 5 2 65 0.05 16 Do you know why the ratios are so different? Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 49 Single top observation in 2009 Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 50 Single top observation in 2009 Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 51 The measurements today Tevatron LHC Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 52 Other nice measurements: ttW, ttZ • ttW and ttZ are very rare SM processes. Interesting per se, they are also backgrounds to new physics searches. • How to look for such a rare and complex final state? Same sign leptons! W Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 53 Where do we stand • ttH is difficult: cross section 130±20fb (@8TeV) approx 1/100 of gg→H ? • CDF/CMS/ATLAS investigated semileptonic and dileptonic decays of ttbar, and bbar decays of H • Limits ranging from 4.5xSM(CMS) to 10.5xSM(ATLAS) using 5fb @7TeV ? Ttbar in lepton+jets, Plus dilepton Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 54 Conclusions • • • • Top quark physics is a fascinating topic since 1995 Entered the LHC era, top stats have been increasing one order of magnitude each year! Interesting connections to electroweak symmetry breaking Pivotal for new physics exploration at higher and higher mass scales ! ! more on new physics and top quarks next time Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 55 Some useful reading • Top is recent so you won’t find much on textbooks • An exception: you can find the discovery papers on the Goldhaber Cahn • Review papers are very useful at graduate level (graduate = MS, PhD) if you want to know more on a specific topic • A top mass review: L.Galtieri, I.Volobuev, myself - Rept.Prog.Phys. 75 (2012) 056201 arXiv: 1109.2163 • Recent reviews on top quark physics: • W.Bernreuther, Top quark physics at the LHC. J.Phys. G35 (2008) 083001 http:// arxiv.org/pdf/0805.1333.pdf • Particle data group http://pdg.lbl.gov/2012/reviews/rpp2012-rev-top-quark.pdf Fabrizio Margaroli Sapienza Università di Roma-INFN Roma 1 56