Synopsis

Relativistic heavy-ion collisions allow us to study the thermodynamic properties of quantum chromodynamics (QCD) and the Quark-Gluon Plasma (QGP). In these collisions, high-momentum partons scatter and produce collimated sprays of particles known as jets. Because isolated "prompt" photons do not interact via the strong force, they traverse the nuclear medium unaffected and provide a clean, calibrated proxy for the initial kinematics of the recoiling parton.

By measuring the correlations between these prompt photons and the associated charged hadrons, we can map out how partons fragment and measure potential modifications or energy loss. This analysis focuses on smaller collision systems—proton-proton (pp) and proton-Lead (p-Pb) at $\sqrt{s_{NN}}=5.02$ TeV—to constrain initial-state Cold Nuclear Matter (CNM) effects independent of QGP formation.

Abstract

This work presents the measurement of isolated photon-hadron correlations and the first study of photon-tagged fragmentation in p-Pb at the Large Hadron Collider using pp and p-Pb data collected by the ALICE detector. Prompt photons produced at leading order in hard scatterings constrain the kinematics of the recoiling parton, enabling the study of parton energy loss and modification to the parton fragmentation function.

For photons with $|\eta|<0.67$ and $12

Introduction & Motivation

At high-energy colliders, analyzing the distribution of particles observed in detectors enables us to infer the dynamics of quarks and gluons. In larger heavy-ion systems like Pb-Pb, significant jet quenching is observed as partons lose energy traversing the QGP. However, interpreting these modifications requires a rigorous understanding of the initial state and potential cold nuclear matter (CNM) effects. Measurements in asymmetric p-Pb collisions act as a crucial control experiment.

The analysis utilizes data from the ALICE Electromagnetic Calorimeter (EMCal) to trigger on high-$p_T$ isolated clusters, extracting a highly pure sample of prompt photons through a data-driven template fit method. Because these photons balance the momentum of the opposing jet, the angular correlation and $p_T$ balance between the photon and hadrons provides an ideal handle on the underlying fragmentation function.

Event Mixing: Serial vs. Random Access Optimizations

A critical step in constructing two-particle correlation functions is isolating genuine physical correlations from detector acceptance effects. This is typically done through "event mixing," where trigger particles from one event are correlated with associated particles from a different, yet topologically similar, event.

Traditionally, events are categorized into multi-dimensional bins based on properties like multiplicity (V0 amplitude) and primary z-vertex. Triggered events are then mixed randomly with minimum bias events from the corresponding bin. This random access procedure is notoriously memory-intensive and computationally slow.

To optimize this, the analysis sidestepped binned random access by deploying the Gale-Shapley stable matching algorithm. By generating ordered preference lists for both $\gamma$-triggered events and minimum-bias events based on vertex and multiplicity similarity, events were matched and processed serially. This completely avoided the CPU constraints of random memory lookups and strict bin boundaries, serving as a highly elegant computational optimization for large dataset correlation analyses.

Results and Conclusions

The fully corrected correlation functions were used to extract the fragmentation function proxies for both collision systems. A direct comparison of the integrated yields revealed no statistically significant modification in p-Pb compared to the baseline pp collisions.

These results firmly constrain the magnitude of cold nuclear matter effects on parton fragmentation in this kinematic phase space. The data sets are well-described by baseline PYTHIA 8.2 Monte Carlo simulations as well as specific theoretical CNM models within systematic uncertainties, implying that the nuclear environment in p-Pb collisions at ALICE does not drastically alter the fragmentation of high-energy partons.