In July 2025, ATLAS recorded the first-ever collisions of light ions at the LHC — oxygen–oxygen (O–O) and neon–neon (Ne–Ne) — marking a new phase in the study of strongly interacting matter. These short, intense runs provided a unique opportunity to explore the transition between conventional hadronic interactions and the formation of the quark–gluon plasma (QGP), a state of matter where quarks and gluons move freely rather than being confined inside protons and neutrons.

Light-ion collisions occupy an intermediate space between proton–proton and heavy-ion systems, offering a finely tunable environment to study how QGP-like behaviour emerges as a function of system size. The oxygen and neon data were collected between 1 and 9 July 2025 under tightly optimised conditions, with ATLAS employing advanced trigger strategies, high-throughput buffering of collision data, and the Zero Degree Calorimeter to monitor beam conditions and background suppression.

Figure 1. Collision evolution sketch from MADAI collaboration.

Figure 2: Comparison between an oxygen ion and a neon ion, showcasing the distinct “bowling pin” shape of the neon ion. (Image: Giacalone et al., arXiv:2402.05995)

New analyses presented at the 2025 CERN Jamboree reveal that both O–O and Ne–Ne collisions exhibit strong collective flow, evidenced by long-range azimuthal correlations with a clear harmonic hierarchy (v₂>v₃>v₄​). These results demonstrate that collective dynamics persist even in small nuclear systems. Comparing the two species provides an additional insight: differences in the flow coefficients reflect their distinct nuclear shapes, with the characteristic deformation of the neon nucleus leaving a visible imprint on the final-state particle distributions.

Figure 3: The ATLAS experiment observes a clear hierarchy between flow harmonics in oxygen–oxygen and neon–neon collisions, with v₂>v₃>v₄​. The rise-and-fall behaviour of the harmonics mirrors that seen in lead–lead collisions, indicating the presence of collective dynamics even in light-ion systems. Image: ATLAS Collaboration.

Figure 4. Comparison of elliptic (v₂) and triangular (v₃) flow coefficients measured by ATLAS in oxygen–oxygen, neon–neon, xenon–xenon and lead–lead collisions. The results reveal a different centrality dependence between light and heavy ions, showing how initial geometry drives the anisotropic flow in heavier systems, while geometry fluctuations dominate in lighter ones. Image: ATLAS Collaboration.

In parallel, ATLAS reported the first measurement of the dijet momentum imbalance (x_J) in oxygen–oxygen collisions. The results show medium-induced modifications reminiscent of jet quenching — a signature of parton energy loss — suggesting that even light-ion collisions can produce a medium dense enough to alter jet propagation. These findings are key to understanding where and how quark–gluon plasma–like (QGP-like) effects emerge.

Figure 5. Comparison of flow coefficients measured in oxygen–oxygen and neon–neon collisions at 5.36 TeV. The higher elliptic flow (v₂) observed in neon reflects its intrinsic nuclear deformation — the so-called “bowling-pin” geometry — which influences the final-state particle distributions. Image: ATLAS Collaboration.

Beyond QGP studies, the 2025 light-ion campaign also included proton–oxygen collisions, designed to refine nuclear parton distribution functions (nPDFs) for oxygen and to support cosmic-ray research by improving interaction models relevant for atmospheric cascades. This cross-disciplinary connection highlights how collider experiments can inform both high-energy and astroparticle physics.

Together, the O–O and Ne–Ne measurements offer valuable input for theoretical models describing the initial state and the hydrodynamic evolution of nuclear matter. Comparisons with frameworks such as TRENTo+PGCM and IP-Glasma indicate that both nuclear geometry and quantum fluctuations are essential for reproducing the observed flow patterns.

The light-ion results have opened a new experimental frontier for studying collective phenomena in the smallest QGP droplets ever produced in the laboratory. Ongoing ATLAS analyses — including identified-particle flow, photon production, and jet–medium correlations — promise to deepen our understanding of how the quark–gluon plasma emerges, one nucleus at a time.

Read more in the presentation by Riccardo Longo (University of Illinois at Urbana-Champaign, US) delivered during the CERN EP Jamboree on 16 September 2025: Longo_CERN_Jamboree_16Sept2025_v1.4.pdf

Cover image: Event display of a proton–oxygen collision recorded by the ATLAS Experiment on 1 July 2025. (Image: ATLAS Collaboration/CERN)