ATLAS finds further confirmation of evidence for four top quark process

24th March 2021 | By

In a new result released this week, the ATLAS Collaboration studied the production of four top quarks at once in LHC collisions. This is the heaviest particle final state ever seen at the LHC, and provides physicists with a unique opportunity to study the top quark’s relationship to the Higgs boson. Its production rate could be affected by various new theories that go beyond the Standard Model, providing a unique window to search for new physics.

When four top quarks decay in the ATLAS detector, they can lead to some very complicated collision events! The four top quarks produce four W bosons and four jets – collimated sprays of particles – originating from bottom quarks. The W bosons then, in turn, each decay into two jets or one charged lepton (electron, muon or tau leptons) and an invisible neutrino. This leads physicists to search for 12 distinct particle signatures in the detector, with up to twelve jets produced in a single event.

For this result, ATLAS physicists studied the full LHC Run-2 dataset and focused on four-top-quark decays producing one charged lepton or two leptons with opposite electric charge. Despite accounting for the lion’s share of four-top-quark events (about 57%), these signatures are easily overshadowed by other, much more-common Standard Model processes with similar decay products.

Figure 1: The boosted decision tree score output for the signal region. The data are shown in black; the simulated signal in red. The y-axis shows the number of events. The band includes the total uncertainty on the post profile-likelihood fit (post-fit) computation. The dashed red line shows the signal distribution normalised to the background yield. The ratio of the data to the total post-fit computation is shown in the lower panel. (Image: ATLAS Collaboration/CERN)

Top quarks can be produced in association with additional jets, which is the main source of background for this analysis. This is made particularly challenging when the additional jets are initiated by b-quarks, a process which is still not well understood. To account for this difficult background, ATLAS physicists used simulations based on the best available theoretical predictions. These were then corrected using real data events with fewer jets from b-quarks, avoiding any contamination from four-top-quark events. They then extrapolated this correction to events with more jets from b-quarks. The result: a precise evaluation of the background and a reduced uncertainty on its contribution to the final result.

To best distinguish between four-top-quark and background events, ATLAS physicists used a multivariate discriminant called a Boosted Decision Tree (BDT). The BDT was trained on several distinct features of the signal, such as the large amount of transverse momentum of the many decay products, the kinematic properties of jets initiated by b-quarks, and the magnitude of the missing transverse momentum from the invisible neutrinos. The BDT ultimately gave each event a relative score between minus one and plus one, indicating the likelihood that an event is a four-top-quark decay. A four-top-quark signal would ideally peak at one, as shown in Figure 1.

ATLAS has found evidence of the four-top-quark process with an observed significance of 4.7 standard deviations – just shy of the conventional requirement of 5 standard deviations to claim an observation.

Figure 2: Observed and expected event yields displayed as a function of the ratio of the post-fit signal (S) and total background (B) yields, and shown logarithmically to easily display a large range of this variable. The bins in all fitted regions are ordered and grouped in bins of log10(S/B). The signal is shown for the best-fit signal strength, μ = 2.2, and the Standard Model prediction, μ = 1.0. The lower panel shows the ratio of the data to the post-fit background prediction, compared to the signal-plus-background prediction with the best-fit signal strength, and the Standard Model prediction. The shaded band represents the total uncertainty on the background prediction. (Image: ATLAS Collaboration/CERN)

ATLAS’ new measurement of four-top-quark production (shown in Figure 2) was combined with a previous measurement that looked at events with multiple leptons with the same dataset. The combined four-top-quark production cross section is measured to be 25 +7-6 fb, which is consistent with the Standard Model prediction of 12.0 ± 2.4 fb within 2.0 standard deviations. The existence of the four-top-quark process is therefore favoured with an observed significance of 4.7 standard deviations. This provides stronger evidence for this process than the 2.6 standard deviations that had been expected and is just shy of the conventional requirement of 5 standard deviations to claim an observation.

The study of four-top-quark production will certainly benefit from further data from the LHC’s upcoming operation. However, the biggest improvements to the measurement’s precision will come from new developments in the theoretical understanding of the background processes and more sophisticated methods for signal separation. An approach that incorporates all of these improvements will allow scientists to explore this exciting process in even greater detail in the future and verify its compatibility with the Standard Model.