sign in sign up
<<
Home , Top quark

ATL-PHYS-PROC-2018-023 . 07 May 2018 ig oo rdcinwt aro o urs( top of pair a with production Higgs ( fusion ( -gluon of by production boson Higgs in loops via (entering coupling Yukawa (left) top for The diagrams Feynman – 1 Figure the for diagrams Feynman the shows of 1 measurement Fig. the physics. in new channels for hint production a and be decay might field the coupling value Yukawa Brout-Englert-Higgs , SM the top the SM of from this heaviest (VEV) of quark the value value top with expectation Model The vacuum the boson Standard from Higgs interest. the the SM particular by of the of boson coupling is Higgs The quark, a the for top physics. of predicted 1 particle those Run of GeV the with 125 (SM) at consistent to experiments close are CMS mass interactions and a and ATLAS with the (LHC) by Collider discovered been Large has LHC boson the Higgs at The coupling Yukawa quark Top 1

c g g C-Y40license. CC-BY-4.0 the under licensed is work This Collaboration. ATLAS the of benefit the for CERN oo ihams ls o15GVadapi ftpqak.A xeso vnsoe the over events of excess Furthermore, other An deviations. with standard quarks. result 2.8 of this top of significance Higgs of of expectation observed combination an Model an pair the to with a Standard compared found and deviations, is the standard GeV processes of 4.1 Model 125 Standard presence to from the close background expected mass for a examined with are boson candidates, charged- of ig oo ea oe to modes decay boson Higgs ( pair quark lumi- top integrated an a W W to with of corresponding boson energy dataset fb center-of-mass Higgs a 36.1 the using of states of nosity final production multileptonic associated in the performed for search A H ∗ ig oo rdcini soito ihatpqakpair quark top a with association in production boson Higgs , ττ → and γγ . t , 6,aeu eLmn ae92-128Mrele France Marseille, 13288 - 902 Case - Luminy de avenue 163, b ,where ), ZZ • 1 PM i-asil nvri´ n CNRS/IN2P3, Universit´e and Aix-Marseille CPPM, oetWl nbhl fteALSCollaboration ATLAS the of behalf on Wolff Robert at fpoo-rtncliindt eoddb h TA xeieta a at experiment ATLAS the by recorded data collision -proton of ∗ r agtd ee nlsae,ctgrzdb h ubradflavour and number the by categorized states, final Seven targeted. are √ s λ √ H t 13 = s nesvatpqaklos h relvlpoeso associated of process tree-level The loops. quark top via enters 3TVa h ag arnClie.Hgsbsndcy to decays boson Higgs Collider. Hadron Large the at TeV 13 = b gg ¯ b , m H γγ → e ihteALSdetector ATLAS the with TeV t H 7 e i h formula the via GeV 173 = and rdcin (centre) production, ZZ t • tH ¯ t ∗ , . b gg → erhsfo h TA xeietuigthe using experiment ATLAS the from searches , W gg → 4 ` t H → sbifl reported. briefly is tH ¯ and H lostedrc measurement. direct the allows ) λ n h ig oo east pairs to decays boson Higgs the and ) t H h nietmaueetue the uses measurement indirect The . H 1 γ γ → , 2 → t esrdpoete iespin like properties measured Its . γγ γγ g g sidctdb lc circle. black a by indicated is ) easad(right) and decays v λ t 4 e.Aydeviation Any GeV. 246 = = λ √ t 2 • m λ t t /v speitdby predicted is t tH ¯ t ≈ tH ¯ is ) ihthe with 1 production. t H t ¯ In Run 1 the ATLAS and CMS experiments performed a combined measurement of properties3. The measured ratio of observed over expected ttH¯ production , 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.

√ 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 √ 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 (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 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 H → ττ 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 and bottom quarks, all channels require at least one b-tagged . 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 pair (ttW¯ , ttZ¯ ) and di-boson production (VV ) with similar final ATLAS q mis•id t Wt t Zt Diboson s = 13 TeV τ Fake had Non•prompt 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 conversions. They are dominant in the 2`SS, 2`SS+1τhad and 3` SRs. The 2`OS tt¯events with an 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 . 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 fb•1 s = 13 TeV, 36.1 fb•1 ttW ttZ 2ℓSS Diboson Non•prompt Events / bin Post•Fit q mis•id Other 3 Tot. 10 Uncertainty Pre•Fit Bkgd. Tot. Stat. ( Stat. , Syst. ) 2.1 2ℓOS + 1τ 1.7 + +1.6 +1.4 had − 1.9 ( − 1.5 , − 1.1 ) +1.6 +1.1 +1.1 1ℓ + 2τhad •0.6 2 − 1.5 ( − 0.8 , − 1.3 ) 10 1.3 ℓ •0.5 + +1.3 +0.2 4 − 0.9 ( − 0.8 , − 0.3 ) 1.8 3ℓ + 1τ 1.6 + +1.7 +0.6 had − 1.3 ( − 1.3 , − 0.2 ) 1.7 10 2ℓSS + 1τ 3.5 + +1.5 +0.9 had − 1.3 ( − 1.2 , − 0.5 ) +0.9 +0.6 +0.6 3ℓ 1.8 − 0.7 ( − 0.6 , − 0.5 ) +0.7 +0.4 +0.5 1.51 ℓ 1.5 2 SS − 0.6 ( − 0.4 , − 0.4 ) 1.25 +0.5 +0.3 +0.4 combined 1.6 1 − 0.4 ( − 0.3 , − 0.3 ) 0.75 −2 0 2 4 6 8 10 12 Data / Pred. 0.5 −1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1 Best•fit µ for mH=125 GeV BDT output t Ht

Figure 4 – (left) the distribution of the final BDT output in the 2`SS SR5 and (right) the best-fit signal strength 5 µttH¯ in single channels and their combination in the ttH¯ → multilepton analysis . with the uncertainty on ttW¯ subtraction in the CR of the fake lepton efficiency measurement for the non-prompt light lepton estimate in the 2`SS and 3` channels. The results for the best-fit signal strength µttH¯ are shown in Fig.4 (right) for the single channels and their combination. All results are compatible with each other and with the SM +0.5 expectation of µttH¯ = 1. Combining all channels, a value of µttH¯ = 1.6−0.4 with a significance of 4.1σ has been observed, while 2.8σ is expected. The systematic uncertainties with the largest impact on this measurement are ttH¯ signal cross section from QCD scale and PDF variations, jet energy scale and resolution and non-prompt light lepton estimate with a large impact of CR ¯ +230 statistics. The measured cross-section for the ttH production is σttH¯ = 790−210 fb compared to SM +35 the expectation of σttH¯ = 507−50 fb.

3 Combination with other searches for ttH¯ production at the ATLAS experiment

The results of this analysis have been combined with the measurements of ttH¯ production √ searches in other Higgs boson decay channels with the 36.1 fb−1 of pp collisions at s = 13 TeV. 6 The search with the Higgs boson decaying to b-quark pair observed a value of µttH¯ = 0.8 ± 0.6; 7 +0.7 the di-photon decay channel found µttH¯ = 0.6−0.6; and the search with H → ZZ → 4` in a 10 GeV mass window around the Higgs boson mass8 observed a 68% confidence level upper limit on µttH¯ of 1.9. The search for ttH¯ production in multileptonic final states has a major impact on the combination results. The combined best-fit value is µttH¯ = 1.2 ± 0.3. The ob- served (expected) significance of this excess over the background-only hypothesis is 4.2 (3.8)σ. In conclusion, the ATLAS experiment found evidence for the ttH¯ production at 13 TeV5.

References

1. ATLAS Collaboration, Phys. Lett. B 716, 1 (2012), arXiv:1207.7214 [hep-ex]. 2. CMS Collaboration, Phys. Lett. B 716, 30 (2012), arXiv:1207.7235 [hep-ex]. 3. ATLAS and CMS Collaborations, JHEP 08, 045 (2016), arXiv:1606.02266 [hep-ex]. 4. ATLAS Collaboration, JINST 3, S08003 (2008). 5. ATLAS Collaboration, Phys. Rev. D 97, 072003 (2018), arXiv:1712.08891 [hep-ex]. 6. ATLAS Collaboration, Phys. Rev. D 97, 072016 (2018), arXiv:1712.08895 [hep-ex]. 7. ATLAS Collaboration, CERN-EP-2017-288, arXiv:1802.04146 [hep-ex]. 8. ATLAS Collaboration, JHEP 03, 095 (2018), arXiv:1712.02304 [hep-ex].