⚛️ nuclear theory

Using $\gamma$ +jets to quantify medium-induced jet broadening in heavy-ion collisions

Using the JEWEL event generator, this study proposes a strategy to mitigate selection bias in heavy-ion collisions by re-clustering small-radius jets into "trimmed" jets, demonstrating that while this approach recovers some medium-induced broadening signals, the full effect is best observed by analyzing the radial profiles of sub-leading subjets beyond their cone radii.

Original authors: Ankita Budhraja, Marco van Leeuwen, Wouter Waalewijn

Published 2026-07-15

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Ankita Budhraja, Marco van Leeuwen, Wouter Waalewijn

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 authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a high-energy particle collision as a chaotic mosh pit at a rock concert. When two heavy nuclei smash together, they create a super-hot, super-dense soup of particles called a "quark-gluon plasma" (QGP). If a high-speed particle (a "jet") tries to zoom through this soup, it doesn't just glide; it gets battered, loses energy, and gets pushed around by the crowd.

For a long time, scientists trying to study this soup faced a tricky problem: a "selection bias." It's like trying to study how a crowd pushes people around by only looking at the people who managed to run out of the mosh pit without getting tired. If you only pick the runners who still have high energy, you miss the ones who got slowed down the most. This made it look like the crowd was actually narrowing the paths of the runners, when in reality, the soup was supposed to be broadening them, scattering them wide.

The Photon Anchor
To fix this, the authors (Ankita Budhraj, Marco van Leeuwen, and Wouter J. Waalewijn) proposed a clever trick using a "photon-jet" event. Think of the photon as a super-secure anchor. Unlike the jet, the photon doesn't carry any "color charge" (it's invisible to the soup), so it flies straight out of the mosh pit untouched. By measuring the photon's speed, scientists know exactly how fast the jet should have been going. Any difference tells them how much the jet got slowed down.

The Big vs. Small Bucket Problem
The authors ran simulations using a tool called Jewel to see what happens. They first tried to catch the jet in a giant bucket (a "large-R" jet with a radius of R = 1.2). In their simulations, this giant bucket caught all the scattered particles, and the results were clear: the jet in the soup was indeed wider (broader) than in a vacuum.

However, there's a catch. In a real heavy-ion collision, the "underlying event" (the background noise of the mosh pit) is so loud and fluctuating that a giant bucket is impossible to use in a real experiment; it would just fill up with noise.

The "Trimmed" Solution
So, the team tried a new strategy: Jet Trimming. Instead of one giant bucket, imagine catching the jet with a bunch of tiny, high-quality buckets (small-radius jets with rsub = 0.2) and then gluing them together to pretend they are one big bucket. They only glue together the tiny buckets that have enough energy (a minimum transverse momentum of psub T > 20 GeV) to ignore the background noise.

What the Simulations Showed
When they tested this "trimmed" method in their simulations:

The Core Shrinks: If they looked only at the very center of the glued-together jet (the tiny rsub = 0.2 core), it looked narrower in the soup. This is the same confusing "selection bias" effect that plagued earlier studies. The trimming process accidentally threw away the wide, scattered particles that were the most interesting.
The Tail Reveals the Truth: But, when they increased the size of the tiny buckets to rsub = 0.5, the hidden signal appeared. The distribution of the jet's width showed a clear "tail" of broadening in the soup, just like the giant bucket did. This suggests that the medium-induced broadening happens at wider angles and lower energies, which only larger buckets can catch.

One vs. Two Pieces
The authors also looked at the internal structure of these jets, splitting them into cases where the jet was just one piece (1-subjet) or two pieces (2-subjet).

One Piece: When the jet was just a single piece, it looked significantly narrower in the soup.
Two Pieces: When the jet had a "sub-leading" piece (a second, smaller chunk), that second piece showed a clear sign of broadening.

The Radial Profile Clue
To understand why the single piece looked narrow, they looked at the "radial profile" (how energy is spread out from the center). They found that for the single-piece jets, the soup actually pushed energy outside the tiny bucket radius. Because the trimming procedure only looks inside that tiny radius, it misses the broadening. However, if you look beyond the radius of the tiny bucket, the soup's effect becomes obvious again.

The Bottom Line
This paper, based entirely on simulations using the Jewel event generator, suggests that "jet trimming" is a viable way to study how the quark-gluon plasma broadens jets, but only if you look at the right size of sub-buckets (around rsub = 0.5) or look at specific parts of the jet (like the sub-leading piece in a two-piece jet). If you stick to the smallest buckets, you'll still see the misleading narrowing effect. The authors hope this strategy will guide future real-world experiments to finally see the true, broadening nature of the quark-gluon plasma.

Technical Summary: Using $\gamma$ +jets to Quantify Medium-Induced Jet Broadening in Heavy-Ion Collisions

Problem Statement
Jets produced in ultra-relativistic nucleus-nucleus collisions serve as calibrated probes of the quark-gluon plasma (QGP). However, characterizing the microscopic properties of the QGP through inclusive jet substructure measurements is complicated by a "jet fragmentation bias." In lead-lead (PbPb) collisions, jets selected at a fixed reconstructed transverse momentum ( $p_T^{jet}$ ) are enriched in those that lost comparatively little energy to the medium. Highly quenched jets migrate to lower $p_T$ and fall outside the selection window. Consequently, comparisons to proton-proton (pp) baselines at the same $p_T$ conflate genuine medium modifications with a selection bias toward less quenched jets. This bias produces an apparent narrowing of the jet radiation pattern across various substructure observables, obscuring the expected medium-induced broadening of the perturbative jet shower.

Methodology
The authors utilize the Jewel Monte Carlo event generator to simulate $\gamma$ +jet events in pp and PbPb collisions at $\sqrt{s} = 5.02$ TeV. The study employs a specific strategy to mitigate selection bias and access medium-induced broadening:

Photon Tagging: The photon transverse momentum ( $p_T^\gamma$ ) is used as a baseline to determine the recoil jet's initial momentum. Since photons do not interact with the QGP, they provide an anchor for the initial parton momentum, partially circumventing the fragmentation bias inherent in inclusive jet measurements.
Large- $R$ Baseline: The authors first establish a baseline using direct large-radius ( $R=1.2$ ) jets. These jets recover a greater fraction of energy transported to large angles, rendering the momentum imbalance ( $x_{j\gamma} = p_T^{jet}/p_T^\gamma$ ) nearly insensitive to overall energy loss. Simulations show that for $R=1.2$ , the $x_{j\gamma}$ distributions in pp and PbPb are nearly identical, and the jet girth (a measure of transverse spread) exhibits clear broadening in PbPb.
Jet Trimming (Experimental Proxy): Recognizing that large- $R$ jets are experimentally inaccessible due to underlying event fluctuations in heavy-ion collisions, the authors propose a "jet trimming" procedure. This involves re-clustering a collection of small-radius ( $r_{sub} = 0.2$ ) jets into a proxy for the large- $R$ jet. A minimum transverse momentum cut ( $p_T^{sub} > 20$ GeV) on the subjets acts as a natural grooming criterion to suppress background.
Differential Analysis: The study analyzes the internal structure of these trimmed jets by separating events into 1-subjet and 2-subjet configurations. It further investigates the jet girth distribution and the jet radial profile ( $\rho(r)$ ) to distinguish between vacuum radiation and medium-induced effects.

Key Contributions and Results

Validation of Trimming: The trimming procedure ( $r_{sub}=0.2, p_T^{sub}>20$ GeV) successfully recovers the physics of direct large- $R$ jets regarding the momentum imbalance $x_{j\gamma}$ , bringing the distribution closer to unity compared to standard small- $R$ jets. However, the trimming procedure does not fully recover the medium-induced broadening signal observed in direct large- $R$ jets when using the standard $r_{sub}=0.2$ radius.
Dependence on Subjet Parameters: The recovery of the broadening signal is highly dependent on the subjet clustering radius. With $r_{sub}=0.2$ , the jet girth distribution in PbPb shows narrowing (similar to inclusive measurements). However, increasing the subjet radius to $r_{sub}=0.5$ reveals a clear signal of medium-induced broadening. This indicates that the broadening originates from radiation at relatively low transverse momentum and large angles, which is lost when the subjet radius is too small.
Subjet Configuration Dependence:
- 1-Subjet Configurations: In events where the trimmed jet contains only a single subjet, the jet girth shows strong narrowing in PbPb compared to pp. The authors attribute this to the fact that medium-induced radiation is not sufficiently concentrated to be clustered into an energetic subjet; thus, selecting for a single subjet effectively removes the broadening component.
- 2-Subjet Configurations: For the sub-leading subjet in 2-subjet configurations, a visible signature of medium-induced broadening is observed (mean girth increases from $0.0588$ in pp to $0.0617$ in PbPb).
Radial Profile Analysis: The jet radial profile reveals that contributions from medium-induced broadening are distinct from radiation in pp collisions. While the subjet core ( $r < r_{sub}$ ) shows narrowing in PbPb, the region beyond the subjet radius ( $r > r_{sub}$ ) shows that the PbPb distribution is less suppressed than the pp distribution. This suggests that the broadening is located at angles larger than the subjet radius.

Significance and Claims
The paper claims to provide an experimentally viable strategy to expose medium-induced jet broadening using $\gamma$ +jet events, thereby reducing the selection bias effects that plague inclusive jet measurements. The primary significance lies in demonstrating that:

Bias Reduction: Photon-tagged events combined with large- $R$ reconstruction (or its trimmed proxy) offer a significantly less biased probe of medium-induced angular modifications than inclusive jets.
Recovery of Physics: While standard trimming ( $r_{sub}=0.2$ ) recovers the momentum balance, recovering the broadening signal requires capturing radiation over sufficiently large angular scales (larger $r_{sub}$ ) or analyzing the radial profile beyond the subjet radius.
Substructure Sensitivity: The study highlights that the dependence of jet quenching on subjet configurations (1-subjet vs. 2-subjet) offers a new window into the microscopic properties of the QGP, specifically regarding color decoherence and independent energy loss.

The authors conclude that this approach provides guidance for future experimental studies of jet-medium interactions, suggesting that analyzing the radial profile beyond the subjet radius or utilizing larger subjet radii in trimmed jets can reveal the distinct signatures of medium-induced broadening that are otherwise hidden by selection biases.

More like this