QCD critical point

Collide two gold nuclei and count the protons that emerge. Even collisions run at the same energy give different counts. Researchers look for patterns in this variation that might signal a special point in hot nuclear matter. Think of a fluid’s liquid–gas boundary ending where the two phases become indistinguishable. QCD, the theory of the strong force binding quarks inside protons and neutrons, may have an analogous endpoint between an abrupt phase change and a smooth transition. The analogy illustrates the map’s shape; the particles and conditions are very different. Does this endpoint exist, and if so, under what conditions does it occur?[1] [3] [4]

QCD describes the strong interaction of quarks and gluons, the constituents of nuclear particles. In hot matter, a proposed boundary between two coexisting phases could end where their distinction vanishes. The open question is whether this endpoint exists at finite temperature in matter with an excess of baryons over their antiparticles, and where it lies. Baryons include protons and neutrons; experiments infer the possible phase structure from what emerges after nuclear collisions.

For a counting illustration, an event with 10 accepted protons and 2 accepted antiprotons has net-proton number 8. Another event gives another count; many events form a distribution. Its variance measures spread, while higher-order cumulants capture features beyond the spread. The fourth cumulant, for example, subtracts the Gaussian contribution from the fourth central moment. STAR’s 2025 study compared the fourth-to-second cumulant ratio, C4/C2, across central gold collisions. It found a downward departure from selected noncritical references around 19.6 GeV of collision energy per nucleon pair in the centre-of-mass frame. That is an energy-dependent clue, not the temperature or density of a located endpoint.

Net protons sample part of net baryon number: neutrons and particles outside the selected window are not counted. Models must connect that subset to the equilibrium fluctuations of QCD matter and to the matter’s expansion before particles cease interacting.[1] [4]

The endpoint’s existence and location remain unestablished. The STAR experiment found notable fluctuation patterns in 2025 and added higher-order measurements in 2026, but other processes also change proton counts. A collision makes a rapidly expanding, cooling sample, so the measured pattern must be traced back through that evolution before it can identify a phase boundary.[1] [2] [3]

An equilibrium critical point predicts long-range correlated fluctuations, but a finite-lived collision may not give them time to reach equilibrium. Conserved charges, variations in participant number and later interactions change the same observables. The 2026 fifth- and sixth-order results remained compatible, within uncertainties, with more than one set of expectations. Identifying an endpoint therefore requires a dynamical explanation of multiple measurements, including how they change when the accepted particle window or collision conditions change.[2] [3] [4]

The same strong force binds quarks inside nuclear particles and controls how those constituents behave in hot, dense matter. A phase map connects those two scales, testing how a known microscopic interaction produces very different forms of matter.[4]

An equation of state turns particle interactions into predictions for pressure and compressibility. Critical fluctuations probe how strongly baryon density responds when thermodynamic conditions change, complementing average particle yields. Connecting these responses to the observed count distributions would test QCD’s account of collective matter and improve the physical interpretation of collision evolution. Hot transient laboratory matter and cold neutron-star matter still require their own conditions and models.[3] [4]