In the hot, dense aftermath of heavy-ion collisions at the Large Hadron Collider, protons and neutrons melt into a foggy soup called the quark-gluon plasma. Inside this furnace, fast-moving quarks and gluons sprint through a medium that behaves like a near-perfect liquid, reshaping their journeys in ways that scientists are only beginning to decode. One of the most revealing ways to study this plasma is by watching jets—sprays of particles that emerge from a single energetic quark or gluon after a high-energy collision. But peering into the jet’s inside is tricky. The medium can scramble the jet’s energy, steering which jets we observe and which we don’t, a kind of experimental selection bias that clouds our view of the jet’s true inner structure.
A new study from the JETSCAPE Collaboration turns up the lights on this problem by focusing on photon-tagged jets. In these events, a hard photon is produced alongside a jet. Because photons barely interact with the quark-gluon plasma, they carry a pristine message from the moment of the collision: they tell us the jet’s original energy, making it easier to separate genuine medium effects from the quirks of how we select which jets to study. The result is not just a cleaner signal—it’s a sharper lens on how the jet’s hardest branchings—the crucial forkings that set the jet’s skeleton—get modified as they plow through the plasma.
Behind the study is a sprawling collaboration of theorists and simulators, centered on the JETSCAPE framework. The lead author is Yuta Tachibana, based at Akita International University in Japan, and the work gathers expertise from Duke University, Wayne State University, MIT, McGill, UC Berkeley, Lawrence Livermore National Laboratory, and many others. The paper explicitly frames γ-tagged jets as a powerful diagnostic tool, offering clearer signals of medium-induced structural changes than inclusive jets ever could. In short: photons and the jets that ride with them form a golden channel for listening to the quark-gluon plasma’s whispers rather than its roars.
Why γ-tagged jets change the game
Traditionally, researchers hunted for jet quenching by looking at jets in general—without tagging them to any particular partner. But in such inclusive samples, a complicated selection bias creeps in. Jets with certain shapes or splittings are more likely to lose energy outside the observed jet cone, making them underrepresented in the data. That bias can masquerade as a fundamental change in the jet’s internal branching, muddying the interpretation of what the medium is actually doing to the jet’s core. The new work argues that γ-tagging—requiring a hard photon to accompany a jet—pulls back the curtain on this bias.
Photons sail through the quark-gluon plasma with minimal interference, so their momenta stay close to the jet’s original energy at production. When experiments trigger on a photon’s transverse momentum, the jet they pair it with tends to have a similar initial energy, regardless of how much energy the jet has later lost to the medium. That means γ-tagged jets deliver a cleaner map of the jet’s initial conditions, letting physicists isolate genuine medium-induced changes in the jet’s internal structure from the distortions introduced by how experiments select jets to study.
The study’s simulations—vividly detailed and anchored in realistic heavy-ion collision conditions—show that γ-tagged jets become especially sensitive to the flavor of the initiating parton. Quark-initiated jets, which are more common in photon-tagged samples than in inclusive jet samples, respond differently to the plasma than gluon-initiated jets. In other words, by tagging on photons we’re not just reducing bias; we’re also enriching the sample with a parton type that reveals distinct medium-driven fingerprints in the jet’s hardest splittings.
One of the paper’s core messages is that γ-tagged jets afford a clearer signal of how the jet’s hard splitting patterns—those moments when a jet branches into two prongs with a certain angular separation—are reshaped by the medium. The Soft Drop grooming technique isolates those hard branchings from softer, later emissions, letting researchers quantify quantities like the splitting fraction zg, the splitting angle rg, and the momentum balance between the two branches. The γ-tagged analysis shows clearer, less bias-laden deviations in these observables than what you’d see in inclusive jets, especially for quark-rich samples. That’s a big deal for the field’s ability to pin down the jet–medium interaction mechanism.
What the multistage model reveals about hard splittings
The team uses a multistage Monte Carlo approach to simulate jet evolution through the quark-gluon plasma. In the first stage, jets are virtuality-ordered: they begin as highly off-shell partons that split in vacuum-like fashion, but with medium effects inserted as they propagate. As the partons shed virtuality and slow down, the simulation transitions to a Low-Virtuality Boltzmann transport (LBT) stage, which captures the elastic and inelastic interactions with the medium’s constituents. A switching criterion governs when a parton hops from the high-virtuality, “matter” phase to the LBT phase, and back again as needed. The result is a controlled, physically motivated picture of how a jet’s inner branching evolves under the influence of the QGP.
What this reveals, startlingly, is that the medium’s influence is not uniform across the jet’s history. The observable shifts in zg, rg, kT,g (the transverse momentum imbalance of the two prongs), and mg (the groomed mass) are dominated by the low-virtuality, later-stage evolution. In other words, the jet’s most critical “skeleton”—the hard branchings that define its core structure—gets reshaped most strongly when the partons’ virtuality has already fallen to relatively small values. This has a clean physical interpretation: early high-virtuality splittings happen so quickly and in such a way that the medium’s resolving power is not yet strong enough to leave a large imprint; the late, softer evolution, by contrast, is more susceptible to broadening and energy redistribution as partons interact more, scatter, and exchange momentum with the surrounding medium.
For γ-tagged jets, the simulations reveal a pronounced broadening of hard splittings in the quark-rich sample. Quark jets tend to start with narrower intrinsic splittings, so the medium can nudge them into larger-angle, higher-mg splittings—producing a noticeable bump in certain rg and mg distributions. Gluon jets, with their inherently broader vacuum splittings, show a different pattern: the medium still broadens, but the signature looks less dramatic in the same observables. This flavor dependence is precisely what γ-tagging helps expose, because it tilts the sample toward quark jets relative to inclusive jet measurements that mix more gluon jets in as the jet energy rises.
Crucially, the researchers show that some of the apparent suppression seen in inclusive jet measurements—where certain broad splittings appear to vanish or shrink—is not necessarily due to a fundamental hard-splitting modification. It can arise from selection bias: jets with broad, energetic splittings lose more energy outside the jet cone and are less likely to be counted in a given trigger. By comparing γ-tagged jets to inclusive jets, the study demonstrates that γ-tagged data can reveal genuine, intrinsic changes in the jet’s hard splittings that would be hidden in an inclusive sample.
Implications for experiments and the future of jet physics
What does this mean for how we study the quark-gluon plasma? First, γ-tagged (and Z-tagged) jets emerge as a powerful experimental tool to probe the jet’s hard substructure with far less bias. If experimentalists push this approach, they can extract cleaner information about how the plasma reshapes the jet’s most fundamental splittings, helping to constrain the microscopic mechanisms of jet quenching—whether it’s collinear radiation, scattering with medium constituents, or the fluid-like response of the medium itself.
Second, the work points toward a path for precision, model-driven analysis. The authors emphasize that these observables are ripe for Bayesian analyses that combine theory with data to infer the transport properties of the QGP. In such analyses, γ-tagged jets could provide a sharper set of constraints on quantities that describe how quarks and gluons exchange momentum with the medium, potentially narrowing down the zoo of competing theoretical models about jet quenching.
Third, the study dovetails with a broader experimental program. CMS and ALICE have already measured several groomed-jet observables in pp and Pb-Pb collisions, and the γ-tagged jet results offer a complementary lens. The simulations align with the trend that, at higher jet energies, the medium’s impact on hard splittings weakens—consistent with the idea that the jet’s core becomes more “resilient” as it borrows from higher-energy scales. Yet the γ-tagged picture also reveals richer flavor-dependent structure, suggesting that future measurements with higher precision could reveal bumps and fine features that cleanly map to the jet’s earliest hard branching in the medium’s presence.
As the field advances, the most exciting horizon is not a single measurement but a coordinated program: use γ- and Z-tagged jets to map the jet’s hard-substructure changes across a range of energies, centralities, and collision systems; compare those patterns to a suite of multistage jet models; and then weave the resulting constraints into a Bayesian framework that infers the QGP’s transport properties with unprecedented clarity. The JETSCAPE collaboration argues that this approach can sharpen our understanding of how jet quenching operates from the very first branching through to the jet’s soft periphery, tying together the microscopic quantum processes with the macroscopic flow of the medium.
One sentence that captures the study’s essence is this: photons offer a faithful invitation to watch how jets bend and broaden inside the quark-gluon plasma, and by listening carefully to γ-tagged jets we stand a better chance of turning the plasma’s whispers into precise, testable physics. The result is a clearer signal, a more nuanced map of jet-medium interactions, and a blueprint for experiments that want to peel back the layers of one of nature’s most extreme states of matter.
The paper’s authors are explicit about what remains to be done. They argue that as experimental data accumulate—especially from γ-tagged and Z-tagged jet measurements with higher precision—these observables should be folded into comprehensive Bayesian analyses that constrain the jet transport coefficient and related dynamics. In the meantime, the multistage approach and the emphasis on hard substructure offer a compelling narrative: the jet’s most consequential branching events, quiet in the early moments and loud in the late evolution, are where the QGP leaves its signature, and γ-tagging helps us hear it clearly.
In short, photon-tagged jets don’t just add a data point to the jet-quenching atlas. They reframes how we test the dialogue between a jet and the medium it traverses. If the field listens, the quark-gluon plasma may finally yield a more complete, physically transparent picture of its inner life—and our ability to read it will be stronger than ever.
Credit for the research goes to the JETSCAPE Collaboration, with the lead author Yuta Tachibana of Akita International University, and a broad cohort spanning Duke University, Wayne State University, MIT, McGill University, UC Berkeley, Brookhaven National Laboratory, and dozens of other institutions. The work is a bold step toward turning jet substructure from a topic of theoretical curiosity into a precise diagnostic of one of the universe’s hottest, most revealing laboratories for fundamental physics.
