Scientists working on CERN’s ALICE experiment have reported the first observation of a distinctive flow pattern among quarks in proton-proton collisions, a signal long associated with quark-gluon plasma (QGP) but previously seen only in collisions between heavy nuclei like lead. The peer-reviewed findings, published in Nature Communications on March 20, 2026, show that even in small collision systems, quarks appear to move collectively as a fluid, behavior that was thought to require the extreme energy densities produced when entire atomic nuclei smash together. The result challenges a central assumption in high-energy physics and reopens questions about how the earliest matter in the universe behaved.
What the ALICE Team Measured
The core finding centers on a quantity physicists call anisotropic flow, labeled v2, which describes how particles emerging from a collision spread unevenly in angle rather than spraying out uniformly. In heavy-ion collisions, a clear split appears at intermediate transverse momentum: baryons (three-quark particles like protons) group separately from mesons (two-quark particles like pions and kaons). That baryon-meson grouping is a fingerprint of collective motion at the quark level, not the hadron level, and has long served as one of the strongest pieces of evidence for QGP in lead-lead events.
For the first time, the ALICE collaboration has now detected this same baryon-meson pattern of v2 at intermediate transverse momentum in high-multiplicity proton-proton collisions at a center-of-mass energy of 13 TeV and in proton-lead collisions at 5.02 TeV. The pattern is consistent with a partonic flowing system, meaning the quarks themselves, not just the composite hadrons they form, are exhibiting collective behavior before they bind together.
This distinction matters because alternative explanations for flow-like signals in small systems, such as initial-state momentum correlations or color reconnection effects in the hadronization process, do not naturally produce the baryon-meson split observed here. The split specifically points to dynamics operating at the parton level, before quarks recombine into hadrons, which is the hallmark of a deconfined QGP phase. A complementary access portal provides authenticated readers with the detailed analysis and systematic checks behind the measurement.
Building on Strangeness Enhancement
The new result does not arrive in isolation. It follows an earlier ALICE discovery, published in Nature Physics, that reported the first signs of enhanced strangeness in high-multiplicity proton-proton collisions. In that study, the yields of strange and multi-strange hadrons relative to pions were found to rise with charged-particle multiplicity, eventually reaching levels comparable to those seen in lead-lead collisions at high multiplicity.
Strangeness enhancement was one of the original proposed signatures of QGP in the 1980s. Strange quarks are not present in the incoming protons, so producing them in abundance requires the kind of thermal energy bath that a deconfined medium provides. A separate journal version of that analysis laid out how the strange-hadron yields scale with event activity and how alternative hadronic scenarios struggle to reproduce the observed trend.
Finding that enhancement in proton-proton events was already surprising. Now, the addition of partonic flow evidence in the same class of collisions creates a second, independent line of evidence pointing toward QGP-like conditions in small systems. Taken together, these two signatures, strangeness enhancement and baryon-meson flow grouping, form a converging case. Each alone could be explained away by competing models. Two signatures in the same collision environment are harder to dismiss without invoking some form of collective partonic dynamics.
Why System Size Assumptions Are Under Pressure
For decades, the working model in heavy-ion physics assumed that QGP requires a minimum volume and energy density to form. Lead nuclei contain 208 nucleons each, so a head-on lead-lead collision creates a fireball with thousands of interacting partons. Protons, by contrast, contain just three valence quarks and a sea of gluons. The idea that two protons could generate a deconfined medium seemed implausible because the overlap region is tiny and short-lived.
The ALICE data now forces a reassessment. “Our results support the hypothesis that an expanding system of quarks is present even when the size of the collision system is small,” said Kai Schweda, the ALICE spokesperson. That statement frames the result not as a curiosity but as a direct test of whether system size or some other property, perhaps the local energy density or the initial collision geometry, determines whether quarks deconfine.
If geometry rather than bulk volume drives collective behavior, the implications extend well beyond CERN. It would mean that QGP-like conditions could arise in any sufficiently energetic interaction where the initial spatial configuration produces strong pressure gradients, regardless of the number of nucleons involved. That possibility has been discussed in theory papers for years, but experimental confirmation has remained elusive until now. The proton-proton and proton-lead measurements, taken together with established lead-lead results, now provide a continuous map of flow observables across system sizes.
Open Questions and Competing Models
Not everyone in the field is ready to declare that QGP forms in proton collisions. Several theoretical frameworks can reproduce some flow-like signals without requiring a thermalized medium. Color Glass Condensate models, for instance, attribute collective-looking patterns to correlations already present in the incoming proton wave functions. String-based models with color reconnection can also generate v2 signals, though they struggle with the specific baryon-meson separation the ALICE team reports.
The critical test going forward will be whether these competing models can simultaneously reproduce both the strangeness enhancement and the partonic flow grouping in the same multiplicity range. Matching one observable is necessary but not sufficient. The open HEPData tables for this analysis give theorists the raw measurements they need to run those comparisons, including differential v2 values by particle species and event activity. Phenomenology groups, such as those working at the Durham IPPP, are expected to confront a wide range of models with these data in the coming months.
One central question is whether a single theoretical description can span all system sizes. Hydrodynamic simulations, which treat the hot matter as a nearly perfect fluid, have been remarkably successful in describing lead-lead collisions. Applying the same hydrodynamic framework to proton-proton events is more controversial, because it requires assuming that local thermal equilibrium is reached on extremely short timescales in a minuscule volume. If hydrodynamics still works, it would reinforce the idea that QGP is a universal phase of QCD matter, insensitive to the overall system size as long as certain density thresholds are crossed.
Alternatively, if initial-state or hadronization-based models can account for the proton data without invoking a deconfined medium, while hydrodynamics remains essential for lead-lead collisions, the community may be forced to accept a more fragmented picture. In that scenario, similar-looking flow observables could emerge from qualitatively different physics depending on the system, complicating attempts to use small collisions as a clean proxy for the early universe.
What Comes Next for ALICE
The ALICE collaboration is already planning follow-up measurements to sharpen the interpretation. One priority is to extend the identified-particle v2 studies to even higher multiplicities and to rarer hadron species, including multi-strange baryons and heavy-flavor mesons. If the baryon-meson grouping persists across a wider momentum range and for particles containing charm quarks, it would strengthen the case that the underlying collectivity is genuinely partonic.
Another avenue is to examine correlations between flow and other observables, such as jet quenching or direct photon production, which have been key QGP signatures in heavy-ion collisions. While full-scale jet suppression may be difficult to observe in small systems, more subtle modifications to jet structure or hadron chemistry could reveal how any deconfined medium interacts with hard probes on short length scales.
On the theoretical side, the newly released data will serve as a benchmark for global analyses that try to extract transport properties of QCD matter, such as the shear viscosity to entropy density ratio, across different collision systems. If consistent values emerge from proton-proton, proton-lead, and lead-lead collisions, it would suggest that the same underlying fluid is being formed in all three, albeit at different scales. If not, it may indicate that the apparent collectivity in small systems is a different phenomenon altogether.
For now, the ALICE results have decisively moved the debate beyond a simple yes or no question about QGP in proton collisions. By demonstrating both strangeness enhancement and quark-level flow in small systems, the experiment has shown that the boundary between “elementary” and “collective” physics at the LHC is far blurrier than once thought. Future runs and refined analyses will determine whether that blur resolves into a universal picture of QCD matter, or into a richer, more complex landscape where multiple mechanisms conspire to shape the behavior of quarks and gluons at the highest energies.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
