Scientists Just Recreated the Plasma That Existed Before Anything Else in the Universe

Scientists at CERN have captured the clearest evidence yet that the plasma filling the universe milliseconds after the Big Bang behaved as a dense, responsive fluid, not a loose particle cloud. The breakthrough came from tracking how a single quark tears through the plasma and leaves a wake behind it, much like a boat cutting through water.

The finding closes a long-standing debate in high-energy physics and opens a new window into the behavior of matter during the universe’s very first microseconds. It also validates a theoretical model years in the making, one that predicted exactly this kind of fluid-like rippling. The study, carried out by the CMS Collaboration at CERN’s Large Hadron Collider, was published in Physics Letters B in December 2025.

What Is Quark-Gluon Plasma, and Why Does It Matter?

In the instants following the Big Bang, the universe was too hot and energetic for ordinary matter to exist. Instead, space was filled with a searing trillion-degree mixture of quarks and gluons, the fundamental constituents of protons and neutrons, moving at nearly the speed of light. This state, known as quark-gluon plasma (QGP), lasted only a few millionths of a second before cooling enough for particles to bind together into the matter that now shapes the cosmos.

According to the CMS Collaboration’s published research, QGP is not only the first liquid to have ever existed in the universe, it is also the hottest, estimated to have reached temperatures of several trillion degrees Celsius. Physicists have long theorized it was a near-“perfect” liquid: smooth, frictionless, and flowing with almost no internal resistance.

Recreating this primordial state is exactly what scientists do at CERN’s Large Hadron Collider, where heavy lead ions are accelerated to near light speed and smashed together. The collisions produce tiny, fleeting droplets of QGP lasting less than a quadrillionth of a second, enough time, with the right technique, to take a snapshot of the early universe.

The “Wake Tag” Technique That Changed Everything

Previous attempts to detect wake effects in the plasma ran into a fundamental problem. As MIT physics professor Yen-Jie Lee explains, when two quarks are produced and travel in opposite directions, “one quark overshadows the wake of the second quark,” making it nearly impossible to isolate what the plasma is actually doing.

The team’s solution was elegant. Rather than hunting for quark-antiquark pairs, they searched for rare collision events that produce a single high-momentum quark traveling back-to-back with a Z boson, a neutral elementary particle that passes through QGP without interacting with it at all. According to Lee, this made the Z boson a perfect “tag”: whatever ripples appeared in the plasma opposite the boson could be attributed entirely to the lone quark.

Sifting through data from 13 billion heavy-ion collisions, the researchers identified roughly 2,000 events containing a Z boson. In each case, they mapped the energy patterns throughout the plasma and consistently found fluid-like splashes and swirls in the direction opposite the boson, a wake, created by a single quark. “We’ve gained the first direct evidence that the quark indeed drags more plasma with it as it travels,” Lee said.

Theory Confirmed, New Questions Opened

The observed wake patterns aligned closely with the predictions of the “hybrid model,” developed by MIT’s Krishna Rajagopal and collaborators. That model, grounded in anti-de Sitter/conformal field theory, predicts that a jet of quarks moving through QGP should generate a positive wake on the jet side and a negative wake, essentially a depleted region, near the Z boson. According to Rajagopal, “this is something that many of us have argued must be there for a good many years, and that many experiments have looked for.”

The CMS paper notes that among the four theoretical models tested against the data, those incorporating medium-response effects, meaning the plasma’s active reaction to the passing quark, provided a significantly better description than those that did not. Models lacking any recoil mechanism consistently failed to reproduce the characteristic dip structures observed in the data.

What remains open is the precise mechanism behind the response. As the paper acknowledges, greater statistical precision will be needed to determine whether the data favors a full hydrodynamic wake model or a medium-recoil-with-holes approach. Lee’s team plans to apply their Z-boson tagging technique to additional collision datasets, with the goal of measuring the size, speed, and decay rate of quark wakes, ultimately building a more complete picture of the exotic fluid that once filled the entire universe.