Science1 publisher3 min readPublished
STAR finds a dip in momentum fluctuations at RHIC's lowest gold-gold energies
STAR's proxy for temperature swings between gold-gold collisions stops falling and turns back up at the bottom of RHIC's energy range, a result now fixed in a dataset the collider finished taking in 2026.
The Scientist · Science desk

What happened
- STAR has found a dip in collision-by-collision variations of particle momenta from gold-gold collisions near RHIC's lowest energies, in a thinly explored region of the nuclear phase diagram.
- The fluctuations fall steeply as collision energy rises from 3 GeV, reach a minimum around 5.2 to 7.7 GeV, then climb again at higher energies.
- The departure from a smooth trend carries a significance of about 5 sigma, the threshold physicists use for calling a deviation statistically significant, in a paper published in Physical Review Letters.
- RHIC ran as a Department of Energy Office of Science user facility at Brookhaven National Laboratory from 2000 to 2026.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- constraint The quantity STAR plotted is a proxy for temperature fluctuations, so the result bounds how the transition can behave with energy without placing a critical point anywhere on the diagram.
- decision Groups planning low-energy heavy-ion running now have a specific energy band to target, and they have to match STAR's particle selection if their points are to sit on the same plot.
- precedent Because Brookhaven cannot retake the 3 GeV points, this becomes the reference dataset for the dense corner of the diagram, and any check on it depends on a different accelerator.
There is no thermometer you can lower into a gold-gold collision, so the collaboration used momentum instead. The analysis measured how far each particle's momentum sat from the average, then compared those deviations from one collision to the next. Hotter matter sends out particles with more momentum on average, so that spread stands in for how much the fireball's temperature moves collision to collision [9]. "This momentum correlation measurement gives us an experimental window into temperature fluctuations," said Chunjian Zhang, a junior faculty member at Fudan University and a co-leader of the STAR analysis [10]. "We can use it as a proxy to look for temperature fluctuations," he said [11].
RHIC's most energetic collisions melt the boundaries of protons and neutrons and free the quarks and gluons inside into a quark-gluon plasma. Lower-energy collisions leave the protons and neutrons intact and squeeze them to densities approaching the inside of a neutron star [14]. Stepping through collision energies charts different regions of the phase diagram, much as an earlier generation mapped the solid, liquid and gaseous phases of water [17].
Five sigma says the points fall off a smooth curve [6]. What bent the curve is a separate question, and the quantity being plotted is a proxy for temperature fluctuations, not a direct measure of them [11]. A critical point would bend it, and locating one has been a long-sought goal of the RHIC program [16]. STAR's own framing is that the dip may be a sign that the way nuclei transform into these exotic forms changes character at lower energies [15]. I think this is the strongest structural signal anyone has from the dense corner of the diagram; it is not yet a critical point.
The minimum occupies 2.5 GeV, from 5.2 to 7.7, inside a compared range that runs 197 GeV from the 3 GeV points up to previously published 200 GeV data. In absolute terms that is a narrow window, about 1.3 percent of the mapped span [18]. In multiples it looks less lopsided: the new low-energy points cover a factor of 2.6 in collision energy against a factor of about 67 for the full comparison [19].
An energy scan fails quietly when the detector's view changes with energy. Restricting the sample to particles emerging at right angles to the beams and near the middle of the debris keeps the selection the same at 3 GeV as it was at 200 GeV, across more than two decades of running [7]. It discards most of the collision debris to do that.
The collider's run is finished [8], so the follow-up will have to come from another machine. The phys.org account does not identify a facility that will extend the gold-gold scan below 7.7 GeV [20]. "The existence of a critical point would tell us about all the phases of matter that could have existed when the universe began," said Rutik Manikandhan of the University of Houston, a STAR member and a leader on the new analysis. "It would sharpen our understanding of the quark-gluon plasma, how it condensed into the protons and neutrons that make up visible matter, and how matter behaves inside neutron stars." [13]
What to watch
- Whether another low-energy heavy-ion experiment reproduces the dip using the same right-angle, mid-debris particle selection.
- Whether reanalysis of archived RHIC runs moves the minimum within the 5.2 to 7.7 GeV window.
- Whether theory calculations place a hypothesized critical point at the same energies as the measured dip.