A report from the ATLAS experiment
Hard-scattering processes in hadronic collisions generate parton showers – highly collimated collections of quarks and gluons that subsequently fragment into hadrons, producing jets. In ultra-relativistic nuclear collisions, the parton shower evolves in a hot and dense quark–gluon plasma (QGP) created by the collision. Interactions of the partons with the plasma lead to reduced parton and jet energies, and modified properties. This phenomenon, known as jet quenching, results in the suppression of jet yields – a suppression that is hypothesised to depend on the structure of the jet. High-momentum shower components with a large angular separation are resolved by the medium, however, it is thought that the plasma has a characteristic angular scale below which they are not resolved, but interact as a single partonic fragment.
Using 5.02 TeV lead–lead collision data taken at the LHC in 2018 and corresponding pp data collected in 2017, ATLAS has measured large-radius jets by clustering smaller-radius jets with transverse momenta pT > 35 GeV. (This procedure suppresses contributions from the underlying event and excludes soft radiation, so that the focus remains on hard partonic splittings.) The sub-jets are further re-clustered in order to obtain the splitting scale, √d12, which represents the transverse momentum scale for the hardest splitting in the jet – a measure of the angular separation between the high-momentum components.
ATLAS has investigated the effect of the splitting scale on jet quenching using the nuclear modification factor (RAA), which is the ratio between the jet yields measured in lead–lead and pp collisions, scaled by the estimated average number of binary nucleon–nucleon collisions. An RAA value of unity indicates no suppression in the QGP, whereas a value below one indicates a suppressed jet yield. The measurement is corrected for background fluctuations and instrumental resolution via an unfolding procedure.
The figure shows RAA for large-radius jets as a function of the average number of participating nucleons – a measure of the centrality of the collision, as glancing collisions involve only a handful of nucleons, whereas head-on collisions involve a large fraction of the 207 or so nucleons in each lead nucleus. RAA is presented separately for large-radius jets with a single isolated high-momentum sub-jet and for those with multiple sub-jets in three intervals of the splitting scale √d12. As expected, jets are increasingly suppressed for more head-on collisions (figure 1). More pertinently to this analysis, and for all centralities, yields of large-radius jets that consist of several sub-jets are found to be significantly more suppressed than those that consist of a single small-radius jet. This observation is qualitatively consistent with the hypothesis that jets with hard internal splittings lose more energy, and provides a new perspective on the role of jet structure in jet suppression. Further progress will require comparison with theoretical models.
ATLAS Collab. 2019 ATLAS-CONF-2019-056.