Quantum mechanics has always insisted that empty space is not truly empty. Now an experiment at Brookhaven National Laboratory has put part of that claim on a very small—and very measurable—footing.
In a paper published in Nature, the STAR Collaboration reports that certain particles emerging from high-energy proton collisions inherit a telltale spin alignment that points back to virtual quark–antiquark pairs in the quantum vacuum. In short: the lab caught a fingerprint of something that, by usual language, “came from nothing.”
The measurement
At the Relativistic Heavy Ion Collider (RHIC) on Long Island, protons were smashed together at √s = 200 GeV. Using roughly the 2012 STAR dataset—about 600 million minimum-bias p+p events—analysts searched for pairs of lambda hyperons (Λ) and their antimatter partners (Λ¯). Lambdas are convenient spies: they contain a strange quark whose spin largely determines the hyperon’s spin, and the direction of that spin can be read out from the angular distribution of the proton (or antiproton) in the Λ → pπ decay.
The key observation emerged when the team looked at Λ–Λ¯ pairs that were produced close together in angle and rapidity. Those short-range pairs showed a clear positive relative polarization: P_{ΛΛ¯} = 0.181 ± 0.035 (stat) ± 0.022 (sys). That signal corresponds to a 4.4σ significance relative to zero. Put another way, when these lambdas are born near each other, their spins are far from random—they tend to point the same way.
The result is not universal across all pairings. Λ–Λ and Λ¯–Λ¯ combinations (same-sign pairs) and widely separated pairs show no significant correlation. The correlation also weakens as the pair separation ΔR (a combination of rapidity and azimuth difference) grows, suggesting the effect is localized and can be disrupted by subsequent interactions.
STAR’s reconstruction relied on its time projection chamber and topology cuts to identify Λ decays at mid-rapidity (|y| < 1) in the pT window 0.5–5.0 GeV/c. A noteworthy experimental complication is feed-down: simulations indicate that only about 11% of reconstructed Λs are primary (directly produced); the rest come from decays of heavier strange baryons. The collaboration folded such contributions into model comparisons and systematic studies.
What the spin tells us
Why should we care about two tiny particles pointing spins the same way? The interpretation the authors favor connects to the QCD vacuum—the sea of fluctuating fields that, according to quantum chromodynamics, contains virtual quark–antiquark pairs and a nonzero quark condensate. The vacuum’s quantum numbers favor the creation of s–s¯ (strange quark) pairs in spin-triplet (parallel) configurations. If a collision liberates such a virtual pair and they become part of separate Λ and Λ¯, their original spin alignment can survive the messy hadronization process and be read out experimentally.
Model comparisons help anchor that picture. A simple SU(6) quark-model estimate, accounting for feed-down, predicts a Λ–Λ¯ correlation of about 9.6% under the assumption of fully spin-aligned initial s–s¯ pairs; the measured short-range signal is compatible with inheriting a large fraction of the initial quark-level alignment. Perturbative sources—gluon splitting g → s s¯—and hadronic final-state interactions were studied and found unlikely to produce the observed short-range magnitude in the measured kinematic region.
Beyond a check on specific models, the observation opens a new experimental window on two big, related QCD puzzles: how confinement turns partons into hadrons, and how emergent properties like mass and spin arise from the vacuum and strong-field dynamics.
Entanglement, decoherence and future probes
The language of entanglement naturally appears in this story: virtual pairs are produced as correlated quantum states, and the experimental signal is consistent with the preservation of that correlation through hadronization. The STAR team notes that a full quantum-information characterization (for instance, via partial-transpose / PPT tests) would require more complete correlation-tensor measurements, beyond the single relative-polarization parameter reported.
Crucially, the drop in correlation with pair separation provides an empirical handle on decoherence: interactions with the surrounding QCD medium or overlapping production of multiple pairs can degrade the original quantum link. That suggests a route to study the transition from quantum-coherent partonic states to the quasi-classical world of hadrons.
The authors also point to possible extensions: higher-momentum Λs (where gluon splitting matters more), other hyperon species, and heavy-ion collisions where a hot medium could restore chiral symmetry and modify vacuum condensates. The upcoming Electron‑Ion Collider and continued RHIC analyses will be fertile ground for those efforts.
Why this matters
This isn’t a spectacle of philosophical nothingness—it's a concrete, quantitative probe of how the vacuum participates in building the mass and structure of the particles around us. The three light quarks inside a proton contribute only a sliver of its mass; most of the rest comes from interactions with the gluon and quark fields of QCD. Tracing spin from virtual pair to real hadron offers a fresh diagnostic of that process.
There are practical reasons to care, too. Understanding how quantum correlations survive (or don't) through hadronization has implications for using hadrons as probes in precision QCD studies, and it forges a tangible link between high-energy nuclear physics and concepts from quantum information science.
The STAR result doesn’t close the book. It hands physicists a short, sharp experimental clue and asks them to follow the thread—map the momentum dependence, test other species, or design measurements that can distinguish entanglement from mere classical correlation. The vacuum may be busy, but physicists are finally beginning to read its fingerprints.