⚛️ phenomenology

A POWHEG generator for di-jet production in polarized proton-proton collisions

This paper presents a new POWHEG-based Monte Carlo generator for simulating di-jet production in polarized proton-proton collisions at next-to-leading order accuracy, providing critical insights into parton-shower effects and selection criteria for the Relativistic Heavy Ion Collider's spin program.

Original authors: Ignacio Borsa, David Betz, Barbara Jäger

Published 2026-02-06

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Ignacio Borsa, David Betz, Barbara Jäger

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 the proton not as a solid marble, but as a bustling, chaotic city made of tiny, speeding particles called quarks and gluons. Scientists have long known that these particles have a property called "spin," which is like a tiny internal compass needle pointing either up or down. The big mystery in physics is: How do all these spinning needles add up to make the proton's total spin?

To solve this, scientists smash protons together at incredibly high speeds in a giant machine called the Relativistic Heavy Ion Collider (RHIC). They want to see what happens when they align the "compass needles" of the colliding protons.

This paper introduces a new, super-smart simulation tool (a Monte Carlo generator) designed to predict exactly what happens during these crashes, specifically when two "jets" (sprays of particles) are produced.

Here is a breakdown of what the authors did, using simple analogies:

1. The Problem: The "Blurry" Prediction

Think of the old way of predicting these crashes like trying to predict the path of a single billiard ball hitting another. You can calculate the math perfectly for that one hit (this is called "fixed-order" calculation).

However, in the real world, when particles collide, they don't just bounce; they often spit out extra tiny particles (radiation) in a chaotic spray. The old math gets "unstable" or confused when trying to account for these extra sprays, especially when the particles are moving in very specific, tricky directions. It's like trying to predict the path of a billiard ball while ignoring the fact that the table is vibrating and the balls are sometimes shooting out confetti.

2. The Solution: The "Smart Simulator"

The authors built a new program using a framework called POWHEG. Think of this as upgrading from a simple billiard calculator to a full-motion video game engine.

The Upgrade: This new engine doesn't just calculate the main crash; it also simulates the "confetti" (the extra particles) that gets sprayed out. It combines the precise math of the main crash with a realistic simulation of the messy aftermath.
The Spin: Crucially, this engine is designed specifically for polarized collisions (where the compass needles are aligned). Before this, scientists had to use a generic engine and then try to "re-weight" the results manually, which was like trying to fix a blurry photo by squinting at it. This new tool takes the spin into account from the very first line of code.

3. Testing the Engine (Validation)

Before trusting the new simulator, the authors ran it against known data and other computer codes.

The Check: They compared their results with older, simpler calculations. They found that for simple, broad questions, the new tool agreed perfectly with the old math.
The "Pathological" Fix: They discovered that in certain tricky situations (where two jets are almost perfectly back-to-back), the old math would sometimes spit out impossible negative numbers or wild swings. The new simulator, however, smoothed these out perfectly, just like a video game engine handles physics better than a spreadsheet. It realized that the "confetti" (radiation) naturally prevents these impossible scenarios.

4. Comparing to Reality (RHIC Phenomenology)

Finally, they used their new tool to predict what the STAR collaboration (a team of scientists at RHIC) actually sees in their detectors.

The Match: They compared their predictions to real data from collisions at two different energy levels (200 GeV and 510 GeV).
The Result: The predictions were already very close to the real data using just the basic math. However, when they turned on the "full simulation" (including the parton shower/confetti), the predictions got even closer to the real-world measurements in some specific areas.
The Takeaway: While the "confetti" didn't change the big picture too much, it helped fine-tune the details, making the theory match the experiment better.

Summary

In short, the authors built a high-definition, spin-aware simulator for particle collisions. It fixes the mathematical glitches of older methods and provides a more accurate way to understand how the spin of the proton is built from its tiny parts. This tool is now available for other scientists to use to analyze data from the RHIC collider, helping them solve the mystery of the proton's spin.

Technical Summary: A POWHEG Generator for Di-jet Production in Polarized Proton-Proton Collisions

Problem Statement
Understanding the spin structure of the proton, particularly the contribution of gluon helicity, remains a central challenge in hadronic physics. While polarized parton distribution functions (PDFs) are extracted from global analyses, significant uncertainties persist in the gluon distribution. High- $p_T$ di-jet production in longitudinally polarized proton-proton ($pp$) collisions offers enhanced sensitivity to the gluon distribution compared to inclusive observables. However, theoretical tools for simulating these processes are limited. Existing fixed-order Next-to-Leading Order (NLO) calculations, while reliable for inclusive observables, can become unstable in phase space regions sensitive to soft or collinear radiation. Furthermore, experimental analyses at the Relativistic Heavy Ion Collider (RHIC) rely on fully exclusive event simulations that include parton showers, hadronization, and detector corrections—features not consistently addressable within a purely fixed-order framework. Currently, simulations for polarized processes often rely on unpolarized parton-shower generators supplemented by reweighting procedures based on Leading Order (LO) spin asymmetries, which lacks full NLO accuracy.

Methodology
The authors present a new Monte Carlo event generator for di-jet production in longitudinally polarized $pp$ collisions, implemented within the POWHEG BOX V2 framework. This implementation achieves NLO accuracy matched to parton showers (NLO+PS).

Theoretical Framework: The generator extends the POWHEG formalism to polarized initial states. The differential polarized cross-section is factorized into polarized PDFs and partonic cross-sections. The NLO calculation includes polarized Born contributions ( $\Delta B$ ), virtual corrections ( $\Delta V$ ), and real-emission corrections ( $\Delta R$ ), utilizing helicity amplitudes for $2 \to 2$ and $2 \to 3$ processes available in the literature.
Implementation Details:
- The POWHEG Sudakov factor is generalized to the polarized case ( $\Delta_{pol}$ ), replacing unpolarized Born and real-emission terms with their polarized analogues.
- Infrared divergences are handled via a polarized extension of the FKS subtraction scheme.
- Collinear factorization terms utilize polarized Altarelli-Parisi splitting functions.
- To handle potentially negative weights inherent in polarized cross-sections (even at LO), the generator employs a specific event weighting scheme ( $n_+$ and $n_-$ ) to preserve NLO accuracy for the inclusive cross-section.
- The implementation was validated using two methods: a direct implementation of helicity amplitudes and an alternative approach using MadGraph5_aMC@NLO with modified interfaces to handle polarization.
Simulation Setup: Phenomenological studies were conducted for RHIC center-of-mass energies of $\sqrt{s} = 200$ GeV and $\sqrt{s} = 510$ GeV. The generator uses the NLO polarized PDF set BDSSV24. Parton showering is performed using PYTHIA6 (Perugia 2012 tune). While the shower itself is based on helicity-averaged codes (as no polarized shower codes are publicly available), the authors argue that the leading logarithmic contributions are correctly captured.

Key Contributions

First NLO+PS Generator for Polarized Di-jets: This work provides the first consistent NLO+PS Monte Carlo generator for di-jet production in longitudinally polarized $pp$ collisions, filling a gap in tools available for the RHIC spin program.
Validation of Fixed-Order vs. Resummation: The paper rigorously compares fixed-order NLO results with the hardest-emission (LHE) and fully showered (NLO+PS) predictions. It demonstrates that while inclusive observables show good agreement, exclusive distributions (e.g., invariant mass $M_{jj}$ , transverse momentum of the third jet $p_{T}^{jet3}$ , and azimuthal separation $\Delta\phi_{jj}$ ) exhibit significant deviations. Specifically, fixed-order predictions can become unstable or negative in regions of soft/collinear radiation (e.g., back-to-back configurations), whereas the POWHEG approach regulates these via Sudakov suppression, yielding physically meaningful results.
Critical Assessment of Selection Criteria: The authors analyze the perturbative stability of experimental selection cuts, noting that low- $p_T$ cuts can enhance discrepancies between fixed-order and resummed predictions.

Results

Validation: The fixed-order implementation matches previous NLO codes (dFFSV) within numerical uncertainties. The comparison between NLO, LHE, and NLO+PS levels reveals that Sudakov effects significantly suppress cross-sections in regions sensitive to soft radiation (e.g., low $M_{jj}$ or small $\Delta p_T$ ).
RHIC Phenomenology: The generator was used to compare predictions with recent STAR collaboration measurements of the longitudinal double-spin asymmetry ( $A_{LL}$ $A_{LL}$ ) at $\sqrt{s} = 200$ $s = 200$ GeV and $\sqrt{s} = 510$ $s = 510$ GeV.
- At the NLO level, predictions generally agree with experimental data within uncertainties.
- The inclusion of parton-shower effects (NLO+PS) leads to a mild enhancement of the predicted $A_{LL}$ at larger di-jet invariant masses.
- For the $\sqrt{s} = 510$ GeV data, the NLO+PS predictions show improved agreement with experimental measurements compared to fixed-order NLO, particularly for specific jet topologies (Forward-Forward and Central-Central).
- The study confirms that hadronization and multi-parton interaction effects largely cancel out in the ratio defining the spin asymmetry.

Significance and Claims
The paper claims that this new Monte Carlo generator provides a useful tool for simulating di-jet production at longitudinally polarized hadron colliders. Its primary significance lies in offering a systematic alternative to LO-based reweighting approaches currently used in experimental analyses. By providing consistent NLO+PS simulations, the tool supports current and future analyses by the STAR and PHENIX collaborations at RHIC, potentially improving the extraction of polarized PDFs and the understanding of the proton's spin structure. The authors modestly note that while parton-shower effects are generally small for inclusive observables, their inclusion improves agreement with data in specific search regions and provides a more robust theoretical description of exclusive kinematic configurations.