Rare top quark production and decays at ATLAS and CMS
aa r X i v : . [ h e p - e x ] M a y Rare top quark production and decays at ATLAS and CMS
K. Skovpen (on behalf of the ATLAS and CMS Collaborations)
IIHE - VUB, Pleinlaan 2, 1050 Brussels, Belgium
The most recent studies in the top quark sector are reviewed with the focus on the rareproduction mechanisms and suppressed decays. The experimental results obtained with theATLAS and CMS detectors in proton-proton collisions at the center-of-mass energy of 13TeV include the measurements of the associated production of top quark pairs with vectorbosons (t¯tW, t¯tZ, t¯t γ ), the first evidence for the t(¯t) γ q process, the first observation of thet(¯t)Zq production, the study of the t¯t + b¯b and t¯t + t¯t processes, as well as searches forlepton flavour violation in top quark decays and effective field theory interpretations. Theexperimental results show good agreement with the theoretical predictions. The top quark takes an important place in the standard model (SM). It’s mass and distinctiveexperimental decay signature in the experiment makes it possible to study a number of very rareproduction mechanisms, as well as extremely rare decay modes. Top quarks produced in pairs,as well as singly produced particles, were already observed at the LHC. The rare productionof top quarks with vector bosons and additional quarks is associated with small cross sectionsand is also challenging due to complex final states. Study of these rare processes allows us toprobe interactions of the top quark with other SM particles and to search for possible anomalousphenomena.
The study of the top quark pair production (t¯t) in association with a W or Z boson is importantbecause these topologies can receive sizeable contributions from new physics, and, in addition,the t¯tZ production represents the main channel to directly measure the top quark couplingsto the Z boson. Moreover, the precise measurements of the production cross sections of theseprocesses are essential for the study of the t¯tH production where the t¯tW and t¯tZ events representone of the dominant backgrounds in the multilepton analysis channels.The study of the t¯tW production at ATLAS 1 is done in the dilepton same-sign and trileptonchannels, while the t¯tZ process is looked for in the dilepton opposite-sign, trilepton and four-lepton final states3. In the analysis of the t¯tZ process using dilepton opposite-sign final states theprompt lepton background originates from Z+jets and t¯t events, while these processes representa non-prompt lepton background in the other channels, also for the case of t¯tW. The promptlepton background is additionally associated with the diboson production and is one of thedominant backgrounds along with the non-prompt leptons. The analysis proceeds with definingcontrol and signal regions with multiple exclusive event categories based on the number of leptonssplit into different flavour and sign with an additional selection based on the number of jets andb-tagged jets. c (cid:13) he analysis of the t¯tZ production allows to perform an effective field theory (EFT) study ofanomalous contributions to the t¯tZ vertex with obtaining constraints on the Wilson coefficientsof the respective dimension-six operators. In such interpretation the t¯tZ event rate can beexpressed as a quadratic function of the Wilson coefficients where the linear terms results fromthe interference between beyond the SM (BSM) and SM operators. The fits to the measureddistributions are done to obtain the EFT constrains in the case when both the quadratic andlinear terms are kept, as well as when the quadratic terms are omitted. The obtained constraintsrepresent competitive results to the existing direct and indirect limits. The inclusive t¯tZ and t¯tWcross sections are measured with ≃
10% and ≃
20% precision, respectively, and are comparableto next-to-leading order (NLO) theoretical uncertainties, as shown in Fig. 1.
W cross section [pb]tt Z c r o ss s e c t i on [ pb ]tt Best fit68% CL95% CLNLO prediction
ATLAS -1 = 13 TeV, 36.1 fbs Figure 1 – The results of the simultaneous fit to the t¯tZ and t¯tW cross sections with the 68% and 95% confidencelevel contours compared to the NLO theoretical predictions 3.
The analysis at CMS 2 studies final states with two same-sign leptons for the t¯tW process,while the t¯tZ production is looked for in the trilepton and four lepton final states 4. The promptand non-prompt lepton backgrounds are validated in control regions in data. The analysis uses animproved multivariate-analysis-based (MVA) lepton identification with respect to the previousiterations of these studies. The t¯tW and t¯tZ cross sections are extracted from combined fit overseveral exclusive event categories defined by the final MVA-based discriminant and the totalnumber of jets and b-tagged jets. The measured inclusive cross sections show good agreementwith NLO predictions.By including more data it becomes possible to probe differential cross sections of the t¯tZproduction. The study done at CMS measures differential distributions of the t¯tZ cross sec-tion using kinematic variables sensitive to t − Z anomalous interactions 5. The measured crosssections are interpreted in two frameworks. The first approach uses an anomalous-coupling La-grangian based on the neutral vector and axial vector current couplings, as well as the weakmagnetic and electric dipole interaction couplings. The second interpretation is EFT-basedwhich considers four dimension-six operators which induce electroweak dipole moments andanomalous neutral-current interactions. The t¯tZ inclusive cross section in this recent analysis isnow measured with an improved precision of ≃ γ production represents an important study of the top-photon elec-troweak couplings where the kinematic distributions of the radiated photon, such as transversemomentum, are especially sensitive to new physics contributions. The measurement of the t¯t γ differential cross section also provides an important information on the t¯t spin correlations and igure 2 – Comparison between data and MC prediction for differential t¯tZ cross sections as a function of thetransverse momentum of the Z boson (left) and the cosine of the angle between the Z boson and the negativelycharged lepton from the Z boson decay in the Z boson rest frame (right) 5. The hatched band includes the theoryuncertainties in the prediction. charge asymmetry, and is complementary to the other t¯t measurements. The t¯t γ process isstudied at ATLAS in the channels with one or two leptons and the results of the differentialmeasurements are compared to leading-order and NLO predictions 6. The previously observeddisagreements in the large values in the distribution of the azimuthal angular difference betweenthe two leptons are now significantly mitigated with moving to the NLO event generation. The associated production of a top quark with a photon (t(¯t) γ q) is an important process whichis sensitive to the charge, as well as the electric and magnetic moments of the top quark. Thesearch for this process is done at CMS in the t-channel considering the final state with one muon,one photon, one b-tagged jet and one forward jet 7. Figure 3 – The boosted decision tree output distribution for data and SM predictions after the fit in the analysisof the t(¯t) γ q process 7. The presence of the forward light flavour energetic jet is a very characteristic signature of thesingle top quark associated production with vector bosons. The dominant background includeshe t¯t γ production, among other contributions. The analysis uses a boosted decision tree-baseddiscriminator to suppress various backgrounds, as shown in Fig. 3. This study resulted in thefirst evidence for this process at 4.4 (3.0) σ observed (expected).The production of a top quark in association with a Z boson (t(¯t)Zq) is sensitive to anomalousWWZ triple-gauge and tZ couplings. The analysis of this production at CMS was done in thefinal state with three leptons 8. This study uses an improved lepton identification which allowedto boost the final sensitivity in this search. A simultaneous fit is performed over several eventcategories to extract the signal with the sensitivity of 8.2 (7.7) σ observed (expected) leading tothe first observation of this process. Comparisons after the final event selection criteria betweendata and predictions are shown in Fig. 4. E v en t s / G e V Data tZq m Nonprompt e / ZttWZ ZZX / tXtt (*) g XMultiboson Total unc.
CMS (13 TeV) -1 (Z) (GeV) T p D a t a / P r ed . Stat. unc. Total unc. E v en t s / . Data tZqWZ Ztt m Nonprompt e / ZZX / tXtt (*) g XMultiboson Total unc.
CMS (13 TeV) -1 | h Recoiling jet | D a t a / P r ed . Stat. unc. Total unc.
Figure 4 – Comparison between the number of expected and observed events for the distributions of the recoilingjet | η | (left) and the transverse momentum of the Z boson after the additional selection on the final discriminant(right) 8. The production of t¯t with additional jets is associated with large theoretical uncertainties dueto the presence of two different scales of the top quark mass and the jet transverse momen-tum. The measurements of the associated production of top quark pairs with b quarks (t¯t + b¯b)are done at ATLAS in single lepton and dilepton final states 9. This study is an importanttest of QCD predictions with providing a better estimation of one of the main backgrounds inthe t¯tH(H → b¯b) analysis. The t¯t + b¯b component is extracted from data using MC templatesdefined by the flavour of additional quark-jets. The measured inclusive fiducial cross sectionsgenerally exceed the t¯t + b¯b NLO predictions but are still compatible within the total uncer-tainties. The inclusive cross section is measured with precision of ≃
20% and is better than inthe theoretical calculations. The measured cross sections in fidual region is presented in Fig. 5.Another rare process that allows to study the QCD predictions is the four top quark pro-duction that is also sensitive to the top quark Yukawa coupling. The search for this process atATLAS is done in single lepton and dilepton opposite-sign channels 10. Events are categorisedbased on the number of jets and b-tagged jets. There are several validation regions defined to igure 5 – The measured fiducial cross sections, with t¯tH and t¯tV(V = W , Z , γ ) contributions subtracted fromdata, compared with t¯t + b¯b predictions 9. perform the measurement of the b tagging efficiencies adapted to the topology of these eventsand to extrapolate it to the signal regions. The results are combined with the previously pub-lished dilepton same-sign and multilepton results. This combination has resulted in the exclusionof the cross section of t¯t + t¯t production down to ≃ × the predicted value at 2.8 (1.0) σ ob-served (expected), with the final limits presented in Fig. 6. The analysis also includes an EFTinterpretation optimised for four-top contact interactions. ✵ ✷ ✹ ✻ ✽ ✶✵ ❙(cid:0)tttt s ✴tttt s ✥ m ✾✁✂ ✄☎ ✆✝✞✝✟ ✠✡✄✠✞❈✝✡☛☞✌✌ ☞✝✆☛✍✎ ✴ ✟✏✝✆☛✍✎✌✝✡✑✆☛ ✆☛✍✎ ✴ ✒✌ ☞✝✆☛✍✎ ❆✓✔❆✕ ✲✖✥ ✶✗ ✘☛✙✚ ✗✻✎✶ ✛❈s ✜✌✢✣✤✤✤✤ s ✶ – ❊✦✍☛✧✟☛☞ s ✷ – ❊✦✍☛✧✟☛☞✒❈s☛✏❖☛☞ ✥✶✣ m ❊✦✍☛✧✟☛☞ ✜ Figure 6 – Summary of the 95% confidence level limits on the t¯t + t¯t production relative to the SM prediction inthe individual channels and for the combination 10.
The search for the t¯t + t¯t production in similar final states is also done at CMS 11. Severalevent categories are defined based on the number of reconstructed jets and b-tagged jets usedin a simultaneous fit to extract the signal. A combination with dilepton same-sign and trileptonresults is also performed. The EFT interpretation was done for four-fermion operators whichcontribute to the t¯t + t¯t production. The combined sensitivity to the t¯t + t¯t production reaches1.4 (1.1) σ observed (expected). The EFT interpretations are becoming an essential part of many analysis studying top quarks.The full EFT at NLO interpretation is done for the top quark production in the dilepton finalstate at CMS, which mainly includes top quarks produced in t¯t and t(¯t)W processes 12. Therere several types of operators which contribute to the production of these events, includingoperators associated with the Wtb couplings, chromomagnetic dipole moment, triple gluon fieldand flavour-changing neutral currents. The interference between the t¯t and t(¯t)W processes isremoved. The EFT constrains are set through the fit of a Neural Network discriminants trainedto distinguish between the EFT contributions and the SM prediction.
In addition to the rare top quark production, one can also search for rare decays of these particles.One such analysis is done at ATLAS to search for the charged lepton-flavour violation (LFV)with the model-independent approach in three-particle decays of top quarks 13. The LFV decaysof top quarks are extremely suppressed in the SM and any deviations from these zero rates wouldindicate the presence of new physics. The analysis is based on the study of the three-leptonfinal state to set limits several EFT operators, including the axial-vector, scalar, pseudo-scalarand lepton-quark interactions. The probability of observing LFV top quark decays is excludeddown to the ≃ − level and this constraint is more stringent than the current indirect limitsset at ≃ − . The experiments done with the ATLAS and CMS detectors at the LHC provide us with a greatopportunity to study very rare processes with top quarks. Recently, we have observed for thefirst time the process with the production of single top quarks in association with a Z boson,as well as have obtained the first evidence for the single top quark production with a photon.The study of the underlying physics in these processes and searches for BSM phenomena willproceed with even more data in the coming years.
References
1. ATLAS Collaboration,
JINST , S08003 (2008).2. CMS Collaboration, JINST , S08004 (2008).3. ATLAS Collaboration, arXiv:1901.03584 (Submitted to Phys. Rev.
D), 2019.4. CMS Collaboration,
JHEP , 011 (2018).5. CMS Collaboration, CMS-PAS-TOP-18-009, https://cds.cern.ch/record/2666205, 2019.6. ATLAS Collaboration, arXiv:1812.01697 (Submitted to Eur. Phys. J.
C), 2019.7. CMS Collaboration,
Phys. Rev. Lett. , 221802 (2018).8. CMS Collaboration,
Phys. Rev. Lett. , 132003 (2019).9. ATLAS Collaboration, arXiv:1811.12113 (Submitted to
JHEP ), 2019.10. ATLAS Collaboration,