Higgs Boson Production in Association with a Top Quark Pair at √ S = 13
Total Page:16
File Type:pdf, Size:1020Kb
Higgs boson production in association with a top quark pair p at s = 13 TeV with the ATLAS detector Robert Wolff on behalf of the ATLAS Collaboration CPPM, Aix-Marseille Universit´eand CNRS/IN2P3, 163, avenue de Luminy - Case 902 - 13288 Marseille, France A search for the associated production of the Higgs boson with a top quark pair (ttH¯ ) is performed in multileptonic final states using a dataset corresponding to an integrated lumi- nosity of 36.1 fb1 of proton-proton collision data recorded by the ATLAS experiment at a p center-of-mass energy of s = 13 TeV at the Large Hadron Collider. Higgs boson decays to WW ∗, ττ and ZZ∗ are targeted. Seven final states, categorized by the number and flavour of charged-lepton candidates, are examined for the presence of the Standard Model Higgs boson with a mass close to 125 GeV and a pair of top quarks. An excess of events over the expected background from Standard Model processes is found with an observed significance of 4.1 standard deviations, compared to an expectation of 2.8 standard deviations. Furthermore, the combination of this result with other ttH¯ searches from the ATLAS experiment using the Higgs boson decay modes to b¯b, γγ and ZZ∗ ! 4` is briefly reported. 1 Top quark Yukawa coupling at the LHC The Higgs boson has been discovered by the ATLAS and CMS experiments at the Run 1 of the Large Hadron Collider (LHC) with a mass close to 125 GeV1;2. Its measured properties like spin and interactions are consistent with those predicted for a Higgs boson by the Standard Model (SM) of particle physics. The coupling of the Higgs boson with the heaviest SM particle, the top quark, is of particular interest. The value of this top Yukawa couplingp λt is predicted by the SM from the top quark mass mt = 173 GeV via the formula λt = 2mt=v ≈ 1 with the vacuum expectation value (VEV) of the Brout-Englert-Higgs field v = 246 GeV. Any deviation from the SM value might be a hint for new physics. Fig.1 shows the Feynman diagrams for the ATL-PHYS-PROC-2018-023 07 May 2018 decay and production channels in the measurement of λt. The indirect measurement uses the g γ g t • t,b • H H •t,b,W λt H g γ g t¯ Figure 1 { Feynman diagrams for (left) gg ! H production,. (centre) H ! γγ decays and (right) ttH¯ production. The top quark Yukawa. coupling (entering via loops in gg ! H and H ! γγ) is indicated by a black circle. Higgs boson production by gluon-gluon fusion (gg ! H) and the Higgs boson decays to pairs of photons (H ! γγ), where λt enters via top quark loops. The tree-level process of associated Higgs boson production with a pair of top quarks (ttH¯ ) allows the direct measurement. c CERN for the benefit of the ATLAS Collaboration. This work is licensed under the CC-BY-4.0 license. In Run 1 the ATLAS and CMS experiments performed a combined measurement of Higgs boson properties3. The measured ratio of observed over expected ttH¯ production cross section, SM +0:7 called signal strength, is µttH¯ = σttH¯ /σttH¯ = 2:3−0:6. This excess of events over SM background corresponds to an observed significance of 4.4 standard deviations (σ), compared to an expec- tation of 2.0σ. Assuming no contributions from new physics the combination of all measured Higgs boson decay and production channels observed a top Yukawa coupling of 0:87±0:15 times the SM prediction, consistent with the SM. p 2 ttH¯ in multileptonic final states at the ATLAS experiment with s = 13 TeV 2.1 Introduction In the Run 2 of the LHC searches for ttH¯ production profit from the increase of center-of-mass p energy from s = 8 TeV to 13 TeV because the ttH¯ production cross-section is enhanced by almost a factor of 4. The presented analysis considers proton-proton collisions data with an integrated luminosity of 36.1 fb−1 recorded by the ATLAS detector4 in 2015 and 2016. The search for ttH¯ production in multileptonic final states5 considers only final states with at least two same-sign leptons (SS) to suppress the dominant background from tt¯ events with oppositely charged leptons. Due to this selection in the ttH¯ signal event topology both the Higgs boson and one of the top quarks need to have at least one lepton in the decay chain. The final states are categorised in seven orthogonal channels by multiplicities of light leptons (`) and hadronically decaying tau leptons (τhad). Each channel has one signal region (SR), apart from 4` where two SRs are separated by the presence or absence of same-flavour, oppositely charged lepton pairs. The categorisation and the Higgs boson decay modes are demonstrated in Fig.2. The light 100 had ATLAS Simulation τ 90 s = 13 TeV 2 1ℓ+2τhad 80 H → other 70 → ττ H Number of 60 H → ZZ Signal Fraction [%] 1 2ℓSS+1τhad 2ℓOS+1τhad 3ℓ+1τhad 4ℓ 50 H → WW 40 30 20 2ℓSS 3ℓ 0 10 0 2lSS+1 2lOS+1 3l+1 1l+2 2lSS 3l SR 4l Z-enriched4l Z-depleted τ τ 2 3 4 τ τ had had 1 had had Number of light leptons Figure 2 { (left) the categorisation of the seven analysis channels by multiplicities of light leptons and hadronically decaying tau leptons5 and (right) the contribution of the Higgs boson decay modes in the eight signal regions5. ∗ lepton channels with no τhad (2`SS, 3` and 4`) target mainly H ! WW and H ! ττ decays with leptonically decaying tau leptons, while the other channels are more sensitive to the H ! ττ decays, where at least one tau lepton decays hadronically. Because the top quarks from the ttH¯ production decay into W bosons and bottom quarks, all channels require at least one b-tagged jet. To suppress backgrounds with low jet multiplicities the basic cut on number of jets is Njet ≥ 2. On top of that, the 2`SS and 2`SS+1τhad (2`OS+1τhad and 1`+2τhad) channels require at least four (three) jets. 2.2 Backgrounds The composition of the backgrounds in the eight SRs is shown in Fig.3. There are two kinds of dominant backgrounds in the analysis. The irreducible backgrounds are SM backgrounds coming from prompt leptons. The main irreducible backgrounds are associated W or Z boson production with a top quark pair (ttW¯ , ttZ¯ ) and di-boson production (VV ) with similar final ATLAS q misid t Wt t Zt Diboson s = 13 TeV τ Fake had Nonprompt Other 2ℓSS 3ℓ SR 4ℓ Z−enr. 4ℓ Z−dep. 2ℓSS +1τhad 2ℓOS+1τhad 3ℓ+1τhad 1ℓ+2τhad Figure 3 { Background composition in the eight signal regions5. 3ℓ ̅ W CR 3ℓ ̅ Z CR 3ℓ VV CR 3ℓ ̅ CR states. Their estimates rely on Monte Carlo (MC) simulation and are validated in 3` control regions (CRs). Further rare backgrounds of ttW¯ W , tH, tZ, ttt¯ t¯, VVV and W tZ production are estimated from MC simulation, too. The reducible backgrounds have at least one fake, non-prompt or charge mis-reconstructed lepton. Non-prompt light leptons come mainly from b-hadron decays in tt¯, single-top and tW production or photon conversions. They are dominant in the 2`SS, 2`SS+1τhad and 3` SRs. The 2`OS tt¯events with an electron of mis-identified charge enter mainly in the 2`SS SR. The channels with τhad have big contributions of fake τhad from light flavour jets and mis-identified electrons. The estimate of these backgrounds is using different data-driven techniques. Dedicated boosted decision trees (BDTs) using lepton properties are designed to reduce these backgrounds. 2.3 Multivariate analysis in the 2`SS channel As shown in Fig.3 top left, the dominant backgrounds in the 2 `SS channel are ttV¯ production and non-prompt light leptons. Two independent event BDTs are trained to discriminate the ttH¯ signal against these backgrounds. The input variables to the BDTs are lepton properties like transverse momenta of the leptons, jet and b-tagged jet multiplicities, angular distances between the leptons and closest jets and the missing transverse momentum. The final BDT output is the combination of the two BDTs with a maximised signal significance. Its distribution of data agrees well with the post-fit prediction as shown in Fig.4 (left). 2.4 Statistical model and results A maximum-likelihood fit with the ttH¯ signal strength µttH¯ as parameter of interest is performed in 8 SRs and 4 CRs simultaneously. The BDT shape is used in five of the SRs, e.g. in the 2`SS SR. The 4 CRs and the 3`+1τhad and 4` SRs with low statistics enter the fit as single event counts. The total number of bins is 32. The systematic uncertainties are described by nuisance parameters (NPs). The total number of NPs is 315 with 191 experimental NPs, 83 NPs from data-driven reducible background estimates and 41 NPs related to signal and background modelling. To decrease the processing time of the fit, NPs are dropped if the size of the corresponding systematic uncertainty is less than 1 %. To reduce local statistical fluctuations of the estimate templates they are smoothed by redistributing bin contents. It has been checked, that the impact of these procedures is negligible on the expected ttH¯ signal significance. Some NPs need some special attention, e.g. the ttW¯ cross section uncertainty is fully anti-correlated 4 10 ATLAS Data ttH ATLAS s=13 TeV, 36.1 fb1 s = 13 TeV, 36.1 fb1 ttW ttZ 2ℓSS Diboson Nonprompt Events / bin PostFit q misid Other 3 Tot.