⚛️ high-energy experiments

Correlations between $\ensuremath{\mit{\Upsilon}(n\text{S})}\xspace$ production and the underlying event in $\ensuremath{p\!p}\xspace$ collisions with the ATLAS detector

Using 139 fb $^{-1}$ of 13 TeV proton-proton collision data, the ATLAS experiment investigates correlations between hard $\Upsilon(n\text{S})$ production and the underlying event, revealing that associated charged-particle multiplicities are systematically lower for excited $\Upsilon$ (2S) and $\Upsilon$ (3S) states compared to the ground state at low transverse momentum, with this difference diminishing as momentum increases.

Original author: Baidyanath Sahoo (On behalf of the ATLAS Collaboration)

Published 2026-11-11

📖 4 min read🧠 Deep dive

CC BY 4.0

Original author: Baidyanath Sahoo (On behalf of the ATLAS Collaboration)

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the author. For technical accuracy, refer to the original paper. Read full disclaimer

In the heart of the universe, matter is built from tiny, fundamental particles that zip around at incredible speeds. When scientists smash protons together at nearly the speed of light, they create a chaotic burst of energy where these particles scatter and interact. Sometimes, this collision is so violent that it briefly creates a hot, dense soup of free-moving quarks and gluons, the building blocks of protons and neutrons. This state of matter, known as a quark-gluon plasma, was once thought to exist only in the massive collisions of heavy atomic nuclei, like gold or lead. However, recent observations suggest that even in smaller collisions between single protons, hints of this exotic soup might appear. To understand how this happens, physicists study the "underlying event," which is the background chatter of particles produced alongside the main collision. By looking at how these background particles behave when specific, heavy particles are created, researchers can learn how the violent core of a collision influences the calmer surroundings.

A researcher using the ATLAS detector at the Large Hadron Collider has taken a fresh look at this problem by focusing on a specific family of heavy particles called bottomonium. These particles are made of a bottom quark and its antimatter partner, and they come in three distinct versions that differ in their mass and how tightly they are bound together. The scientist examined nearly 140 trillion collisions of protons at an energy level of 13 trillion electron volts. They identified events where a bottomonium particle was created and then carefully counted the number of other charged particles that appeared in the same collision. They did this separately for the three different versions of the bottomonium particle, treating each as a unique tag to see if the surrounding environment changed depending on which version was formed.

The results revealed a clear and surprising pattern. When the lightest and most stable version of the bottomonium particle was created, the collision produced a significantly higher number of accompanying charged particles. In contrast, when the heavier, excited versions were formed, the number of surrounding particles dropped noticeably. In the collisions with the lowest momentum, the events producing the heaviest version had about 17 percent fewer charged particles than those producing the lightest version. This difference was not random; it was a consistent trend that held true across the data. As the momentum of the bottomonium particle increased, the gap between the different versions narrowed, but the heavier versions still consistently produced fewer background particles than the lightest one.

This finding challenges the current models that physicists use to simulate these collisions. The researcher compared their real-world data with predictions from a widely used computer simulation program. The simulation suggested that all three versions of the bottomonium particle should be accompanied by roughly the same number of background particles, regardless of their mass or energy. The simulation failed to reproduce the hierarchy the scientist observed in the data. Even when the researcher adjusted the simulation to account for different ways particles might reconnect or decay, the discrepancy remained. This indicates that our current understanding of how heavy particles form and interact with the surrounding environment is incomplete.

The study also looked at how these particles were distributed in space relative to the main collision. The difference in particle counts was not just a result of the main particles balancing each other out in a jet-like spray. Instead, the missing particles seemed to come from the softer, more diffuse background of the event. This suggests that the formation of the heavier, excited bottomonium states somehow suppresses the production of other particles in the surrounding area. The researcher concludes that the way these heavy particles are born is intimately linked to the activity of the entire collision, and that the specific state of the heavy particle acts as a sensitive probe of the complex forces at play. These measurements provide a new and strict test for theories describing how matter behaves in these extreme conditions, pointing toward a deeper connection between the creation of heavy particles and the soft, underlying dynamics of the collision.

Technical Summary: Correlations between $\Upsilon(nS)$ Production and the Underlying Event in $pp$ Collisions

Problem Statement
Recent observations at the LHC and RHIC suggest that small collision systems, such as proton-proton ($pp$) collisions, may exhibit signatures similar to the quark-gluon plasma (QGP) typically associated with large systems. While QGP-like signatures have been noted in the soft sector, constraints on partonic energy loss in the hard sector remain. A critical open question involves the interplay between hard scattering processes and soft QCD dynamics, specifically the correlations between hard and soft particle production. Previous measurements by the CMS Collaboration indicated a decrease in the relative production of excited $\Upsilon(nS)$ states with increasing charged-particle multiplicity. This paper addresses the inverse perspective: using the $\Upsilon(nS)$ state as a hard probe to characterize the accompanying charged-particle activity (the underlying event, or UE) and determine if the formation of different bottomonium states correlates with distinct soft hadronic environments.

Methodology
The analysis utilizes the complete Run 2 $pp$ dataset collected by the ATLAS detector at $\sqrt{s} = 13$ TeV, corresponding to an integrated luminosity of 139 fb $^{-1}$ . Events with an average of 50 or more simultaneous interactions were excluded, resulting in an average pile-up of approximately 33.7.

Event Selection: $\Upsilon(nS)$ candidates were reconstructed via the dimuon decay channel ( $\Upsilon \to \mu^+\mu^-$ ). Muons were required to have $p_T > 4$ GeV and $|\eta| < 2.4$ . Oppositely charged muon pairs were selected within an invariant mass range of $8.2 < m_{\mu\mu} < 11.8$ GeV and $|y_{\mu\mu}| < 1.6$ .
Charged Particle Selection: Primary charged particles were selected with $0.5 < p_T < 10$ GeV and $|\eta| < 2.5$ . Tracks were required to be within 0.5 mm (transverse) and 0.75 mm (longitudinal) of the primary vertex. The muon tracks forming the $\Upsilon$ candidate were excluded from the charged-particle observables.
Signal Extraction: Due to the overlap of the three $\Upsilon(nS)$ peaks and the smooth dimuon background, the analysis measured charged-particle distributions in five mass intervals (three signal regions and two sidebands). A fit was performed in each $p_T^{\mu\mu}$ interval using three signal components (modeled by Crystal Ball and Gaussian functions) and a third-degree polynomial background.
Corrections: Detector acceptance, reconstruction, and trigger efficiencies were corrected using simulation (Pythia 8 with A14 tune and CTEQ6L1 PDFs) and data-driven scale factors. Pile-up contributions were estimated and subtracted using a procedure developed for ATLAS $Z$ -tagged UE measurements.

Key Contributions and Results
The study measured the multiplicity, transverse momentum ( $p_T$ ), and azimuthal separation ( $\Delta\phi$ ) of charged particles associated with $\Upsilon(1S)$ , $\Upsilon(2S)$ , and $\Upsilon(3S)$ mesons as a function of the $\Upsilon$ transverse momentum ( $p_T^{\mu\mu}$ ).

Multiplicity Hierarchy: A distinct hierarchy in charged-particle activity was observed. In the lowest measured $p_T^{\mu\mu}$ interval, events containing $\Upsilon(1S)$ mesons exhibited a mean charged-particle multiplicity of $29.5 \pm 0.7$ . In contrast, $\Upsilon(2S)$ and $\Upsilon(3S)$ events showed systematically lower multiplicities, reduced by $3.6 \pm 0.4$ and $4.9 \pm 1.1$ particles, respectively. This corresponds to relative reductions of $12 \pm 1\%$ and $17 \pm 4\%$ compared to the $\Upsilon(1S)$ sample.
$p_T$ Dependence: The difference in multiplicity between the ground state ( $\Upsilon(1S)$ ) and excited states ( $\Upsilon(2S)$ , $\Upsilon(3S)$ ) decreases as $p_T^{\mu\mu}$ increases but remains positive across the entire measured range.
Kinematic Distributions: The subtracted transverse-momentum and azimuthal distributions indicate that the observed differences are not solely driven by jet activity balancing the $\Upsilon$ momentum. The subtracted $\Delta\phi$ distributions remain positive around $\Delta\phi = \pm\pi/2$ , suggesting a substantial soft component in the differences between states.
Model Comparison: Pythia 8 predictions, both with and without feed-down contributions from higher bottomonium states (e.g., $\chi_b$ decays), failed to reproduce the observed hierarchy. The model predicted very similar multiplicities for all three states. Varying the colour-reconnection scheme also failed to resolve the discrepancy.

Significance
The paper claims that the observed multiplicity hierarchy provides a strong constraint on the modeling of bottomonium production and its correlation with the surrounding soft event activity. Specifically, the data indicate that events producing excited bottomonium states are accompanied by systematically lower charged-particle activity than those producing ground states, with the effect most pronounced at low $p_T^{\mu\mu}$ .

The author concludes that the observed state dependence of charged-particle activity, particularly at low transverse momentum, motivates further studies of the relationship between bottomonium production and the hadronic environment in high-multiplicity $pp$ collisions. The results challenge existing Monte Carlo models (specifically Pythia 8) to better describe the interplay between quarkonium formation and soft QCD dynamics.

Technical Summary: Correlations between Υ(nS)\Upsilon(nS)Υ(nS) Production and the Underlying Event in $pp$ Collisions

More like this

Technical Summary: Correlations between $\Upsilon(nS)$ Production and the Underlying Event in $pp$ Collisions