Tau & co.: the search for new physics with the heaviest leptons

21 March 2023 | By

Could matter be broken down endlessly? Are the known elementary particles the true fundamental building blocks of matter or are they composite? Are there new elementary particles with similar properties as the existing known ones? Two recent searches by the ATLAS experiment shed light on these profound questions!

The LHC is the perfect instrument to study the puzzle of matter. It acts like a giant microscope with the most fantastic spatial differentiation in the world. Not only can it produce new types of particles, it can reveal the structure – if any – of the known particles. One particle under investigation by the ATLAS experiment is the tau lepton, a heavier analogue of the electron. If it is composite, then the LHC could produce its excited state: the excited tau lepton. The LHC might also produce vector-like taus, particles with similar properties as the Standard Model tau lepton, except that it would have a spin of 1 rather than a spin of 1/2.

Tau leptons are the clue

Both types of hypothetical particles – the excited tau and the vector-like tau – would decay to a Standard Model tau lepton and some other particles. The ATLAS detector, acting like an oversized camera, has great potential to capture the handful of collisions – out of trillions – in which these rare particles could be produced. However, there are many collisions where only Standard Model particles are created that look very similar. These are considered as background, and precisely estimating their expected number is essential. With this at hand, physicists can compare it with the number of observed collisions. This leads to a judgement on the existence – or not – of the rare hypothetical particles.

Figure 1: Data are compared with the expected event yields given just background processes (coloured area) or background plus signal processes (dashed red line). The considered signal is the excited tau lepton with mass 1.5 TeV. The S_T variable is, loosely speaking, the amount of energy carried by the most energetic particles in each proton-proton collision. (Image: ATLAS Collaboration/CERN)

The Standard Model tau lepton is a tough particle to detect. It has a short lifetime and decays before reaching the ATLAS detector. Two-thirds of taus decay to hadrons (particles made from quarks) and one-third of taus decay to electrons or muons. Hadronically-decaying taus are difficult to measure because their detector signature is very similar to that of ‘jets’ (collimated sprays of hadrons). Here, artificial intelligence helps tremendously to distinguish hadronic taus from jets.

Pushing knowledge of the tau lepton size

In the excited-tau hunt, the signal collisions would produce two Standard Model tau leptons and at least two jets. The main background processes are the production of a Z boson in association with jets, the production of top-quark pairs, and the production of events with many jets, with two misidentified as tau leptons. Figure 1 shows the background prediction (the coloured area) with its incredibly small uncertainty displayed by the hatched band. The observed data (black points) show no excess over the background prediction. The excellent prediction allowed physicists to exclude the existence of excited tau leptons if their production rate is larger than 1 in 100 trillion proton-proton collisions. If the tau lepton is divisible, then its size is less than one-tenth of one quintillionth of a metre!

The ATLAS Collaboration has released two new studies of the tau lepton, investigating whether this elementary particle may actually be composite in nature.

On the hunt for vector-like tau leptons

The vector-like tau model introduces an additional pair of leptons: the vector-like tau lepton and the vector-like tau neutrino. If these particles exist, they would decay into Standard Model tau leptons and produce signals with multiple leptons of various combinations. To leave no stone unturned, researchers looked into seven different final states. Figure 2 shows the comparison between what was expected from background collision processes (filled histograms) and the ATLAS data (black dots).

Figure 2: The agreement between the ATLAS data and the Standard Model prediction in the various experimental regions explored. (Image: ATLAS Collaboration/CERN)
Figure 3: Production cross-section of the vector-like tau and vector-like tau neutrino as a function of the mass of the vector-like tau. (Image: ATLAS Collaboration/CERN)

The agreement is not perfect, but it is not supposed to be! There are only a limited number of events that are collected by the ATLAS detector, so statistical fluctuations are expected. As the vector-like tau model does not predict the mass of the particles, researchers explored various hypotheses. In Figure 3, the red line shows the theoretical prediction of the production rate (cross section) for various possible vector-like-tau masses. The dotted black line shows how sensitive the ATLAS search for vector-like taus is, depending on the mass, while the solid black line shows what was observed in the data. The difference between dotted and solid lines shows there are more events than what was expected (as is also seen in Figure 2), but this is within statistical fluctuations of data.

The adventure continues

While new physics phenomena were not observed, these new results rule out significant regions being explored by theoretical models and will help guide future searches. The goal of the ATLAS physics programme is to perform as detailed a search as possible, looking for deviations from the Standard Model where new physics may be hidden. With more data being collected and analysis techniques being improved, a better understanding of the laws of nature can be achieved.