Centrality dependence of charged-hadron pseudorapidity distributions in oxygen-oxygen collisions at $\sqrt{s_\mathrm{NN}}$ = 5.36 TeV

The CMS experiment reports the first measurement of charged-hadron pseudorapidity distributions in oxygen-oxygen collisions at $\sqrt{s_\mathrm{NN}}$ = 5.36 TeV, revealing that while the particle density per participating nucleon in central collisions matches that of lead-lead collisions, the data exhibit deviations from simple scaling laws that highlight the significant role of collision geometry and finite-size effects in light ion systems.

The Gist

The Big Picture: Smashing Tiny Oranges

Imagine the Large Hadron Collider (LHC) as the world's most powerful particle accelerator. Usually, scientists smash together giant, heavy nuclei like lead (PbPb) or xenon (XeXe). Think of these as smashing two heavy watermelons together.

In this new study, the CMS collaboration decided to smash something much smaller: Oxygen nuclei. If lead is a watermelon, oxygen is like a small orange. They smashed these "oxygen oranges" together at incredibly high speeds (5.36 TeV) to see what happens when you create a tiny, super-hot fireball of matter.

Why Do This?

Scientists want to understand the Quark-Gluon Plasma (QGP). This is a state of matter that existed just fractions of a second after the Big Bang, where particles melt into a soupy, fluid-like state.

• The Mystery: We know big collisions (like watermelons) create this soup. But can tiny collisions (like oranges) do it too?
• The Advantage: Oxygen is a "doubly magic" nucleus, meaning its internal structure is very neat and predictable (like a perfectly stacked pyramid of oranges). This makes it easier for scientists to calculate what should happen theoretically, allowing them to test their models more strictly than with messy, deformed heavy nuclei.

What Did They Measure?

The team looked at the charged particles (like tiny, electrically charged marbles) that flew out of the collision. They measured two main things:

1. How many particles came out? (Multiplicity)
2. Where did they fly? (Pseudorapidity, or $\eta$)

Think of pseudorapidity as a measure of the angle. If you throw a handful of confetti, some flies straight forward, some backward, and some to the sides. The scientists mapped out this "confetti pattern" to see how the collision debris was distributed.

Key Findings

1. The "Sweet Spot" of the Collision
When the two oxygen nuclei hit head-on (the most "central" collision), they produced a massive burst of particles.

• The Result: In the center of the explosion, they found about 135 charged particles per unit of angle.
• The Comparison: This is about 15 times fewer particles than you get from smashing lead nuclei, which makes sense since oxygen is much smaller. However, when they adjusted for the size of the nuclei, the "particle density per participant" was surprisingly similar to the big lead collisions. This suggests that even a tiny "orange" collision creates a fluid-like soup similar to a "watermelon" collision.

2. Testing the Theories (The Crystal Ball)
Scientists have computer programs (called Monte Carlo generators) that try to predict what happens in these crashes. The researchers compared their real data against these digital simulations:

• HIJING: This model predicted too many particles in the center.
• EPOS LHC: This model predicted too few particles everywhere.
• AMPT: This model got the total number of particles right, but the shape of the distribution wasn't perfect.
• TRAJECTUM: This is a hydrodynamic model (treating the collision like a fluid). This was the winner. It matched the real data best, especially for the head-on collisions. This confirms that oxygen collisions really do behave like a fluid.

3. The Shape of the Explosion
The paper found that while the total number of particles scales with the energy of the collision (just like in bigger systems), the way the particles spread out depends heavily on the geometry (the shape and size) of the collision.

• The Analogy: If you drop a large rock in a pond, the ripples are big and smooth. If you drop a small pebble, the ripples are smaller and behave differently near the edges. The oxygen collisions showed that "finite-size effects" (being small) matter a lot. The rules that work for big watermelons don't apply perfectly to small oranges.

The Conclusion

This paper is the first time anyone has measured the detailed particle spray from oxygen-oxygen collisions at this energy level.

• What it proves: Even in these tiny collisions, the matter behaves like a near-perfect fluid (QGP).
• What it teaches us: The hydrodynamic model TRAJECTUM is currently the best tool we have for describing these events.
• The Takeaway: While the general rules of particle production hold true, the specific "shape" of the collision depends on the size of the nuclei. Smashing small, neat oxygen nuclei gives us a cleaner, more precise way to test our understanding of the universe's earliest moments than smashing messy, heavy nuclei.

In short: We smashed tiny oranges at light speed, found they turned into a fluid soup just like big watermelons do, and confirmed that our best fluid-dynamics computer models are on the right track.

Technical Summary

Technical Summary: Centrality Dependence of Charged-Hadron Pseudorapidity Distributions in Oxygen-Oxygen Collisions at $\sqrt{s_{NN}} = 5.36$ TeV

Problem and Motivation
The formation of the quark-gluon plasma (QGP) and the study of strongly interacting matter under extreme conditions are central goals of ultrarelativistic heavy-ion physics. While collective behavior and QGP-like signatures have been established in large nucleus-nucleus (AA) systems (e.g., PbPb, XeXe), the transition to smaller systems remains a critical area of investigation. Oxygen-oxygen (OO) collisions occupy an intermediate regime between proton-nucleus (pA) and heavy-ion (AA) systems. As a "doubly magic" nucleus with a well-constrained density profile, $^{16}$O offers a unique opportunity to reduce theoretical uncertainties associated with initial-state modeling and nuclear deformation compared to heavier ions.

Despite recent observations of collective flow in OO collisions, precise measurements of charged-particle multiplicity and its pseudorapidity ($\eta$) dependence are required to quantitatively constrain initial-state entropy production and test scaling properties across different collision systems. Specifically, understanding how particle production scales with system size and collision geometry in light ions is essential for validating theoretical models of initial conditions and hydrodynamic response.

Methodology
This analysis utilizes data recorded by the CMS experiment at the LHC in 2025, corresponding to an integrated luminosity of approximately $2 \, \mu\text{b}^{-1}$ of minimum-bias OO collisions at a nucleon-nucleon center-of-mass energy of $\sqrt{s_{NN}} = 5.36$ TeV.

• Event Selection: Hadronic events were selected requiring energy deposits in the forward hadronic (HF) calorimeters and coincident beam signals in the Beam Pickup Timing for Experiments (BPTX) detectors. Events with multiple interaction vertices (pileup) were rejected. The sample was enriched for hadronic interactions by demanding at least two energy deposits $>4$ GeV in each HF calorimeter and a primary vertex within $|v_z| < 15$ cm.
• Centrality Determination: Centrality was defined by the scalar sum of transverse energy ($E_T$) in the HF calorimeters, calibrated against the Glauber model. The $E_T$ spectrum was fitted using a two-component ansatz (scaling with participating nucleons $N_{part}$ and binary collisions $N_{coll}$) to determine centrality percentiles and the corresponding average number of participating nucleons ($\langle N_{part} \rangle$).
• Particle Reconstruction: Primary charged hadrons were reconstructed using "tracklets"—pairs of pixel clusters from distinct layers of the silicon pixel detector. This method allows for efficient reconstruction of low-$p_T$ particles. Tracklets were selected based on angular distance ($\Delta r < 0.5$) to suppress combinatorial background. Corrections were applied for geometric acceptance, reconstruction efficiency, event selection efficiency, and background fractions, utilizing simulations from HIJING, EPOS LHC, and AMPT generators interfaced with GEANT4.
• Systematic Uncertainties: Major sources of systematic uncertainty include tracklet combination consistency (1–9%), primary vertex position consistency (1–4%), model dependence of corrections (1–7%), and centrality calibration (0.5–8% in peripheral collisions). Statistical uncertainties were negligible (<0.1%).

Key Results
The paper presents the first measurement of the charged-hadron pseudorapidity density ($dN_{ch}/d\eta$) in OO collisions for $|\eta| < 2.4$ across various centrality intervals.

• Midrapidity Densities:
  • Integrated over all centralities (0–100%), the average charged-hadron pseudorapidity density in the midrapidity region ($|\eta| < 0.5$) is $\langle dN_{ch}/d\eta \rangle = 41.8 \pm 1.1 \, (\text{syst})$.
  • For the most central events (0–5%), the density is $135.0 \pm 4.0 \, (\text{syst})$.
• Comparison with Models:
  • HIJING: Overpredicts the yield in central collisions.
  • EPOS LHC: Systematically underpredicts the absolute yield across all $\eta$, though it provides the best description of the $\eta$ distribution shape among the event generators.
  • AMPT: Best reproduces the absolute yields, particularly at midrapidity.
  • TRAJECTUM (Hydrodynamic Model): Provides the best overall description of the data, predicting results within $\sim 1\%$ for central collisions and $\sim 10\%$ for peripheral collisions. This holds for both NLEFT and PGCM nuclear density descriptions.
• Scaling and System Size Dependence:
  • The $dN_{ch}/d\eta$ values follow the expected system-size ordering (OO < XeXe < PbPb) at fixed centrality.
  • When normalized by the number of participating nucleons ($\langle N_{part} \rangle$), the per-participant yield in central OO collisions is consistent with that observed in central PbPb collisions at similar energies.
  • However, the centrality dependence of the normalized yield in OO collisions deviates from the trends seen in larger systems (XeXe, PbPb). Specifically, at a given fraction of participating nucleons ($\langle N_{part} \rangle / 2A$), the per-participant yields in OO are lower than in XeXe and PbPb.
  • The data exhibit deviations from simple participant and system-size scaling, highlighting the role of collision geometry and finite-size effects in light ion collisions.

Significance and Claims
The authors claim that these measurements provide the first constraints on final-state particle production and entropy production in OO collisions. The results demonstrate that while the overall energy-scaling behavior observed in nucleus-nucleus collisions is preserved, the centrality dependence of normalized particle production in the smallest ion-ion system studied at the LHC deviates from the trends of larger systems.

The paper concludes that the hydrodynamic model TRAJECTUM offers the most robust description of the data, particularly in central collisions. The consistency of per-participant yields between central OO and central PbPb collisions suggests that particle production mechanisms in central collisions remain comparable even in substantially smaller systems, supporting the presence of collective dynamics beyond independent nucleon-nucleon interactions. Conversely, the observed deviations in the centrality dependence underscore the importance of collision geometry and finite-size effects in light ion collisions, providing new constraints for models of initial-state entropy production and particle production dynamics.