⚛️ phenomenology

All three-loop four-point massless integrals with one off-shell leg

This paper presents the complete set of massless three-loop four-point master integrals with one off-shell leg, which are essential for next-to-next-to-next-to-leading-order QCD predictions for Higgs or vector-boson production with a jet at the LHC, by constructing a canonical basis through integrand analysis and providing their solutions as iterated integrals and high-precision series expansions.

Original authors: Thomas Gehrmann, Petr Jakubčík, Jungwon Lim, Andreas von Manteuffel, Cesare Carlo Mella, Lorenzo Tancredi

Published 2026-10-16

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Thomas Gehrmann, Petr Jakubčík, Jungwon Lim, Andreas von Manteuffel, Cesare Carlo Mella, Lorenzo Tancredi

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

In the high-energy collisions that take place inside the Large Hadron Collider, protons smash together at nearly the speed of light, creating a shower of new particles. Among the most important of these are the Higgs boson and various vector bosons, particles that carry the fundamental forces of nature. To understand exactly how these particles are created, and to test whether our current theories of physics hold up under extreme conditions, scientists must predict the behavior of these collisions with extraordinary precision. This requires calculating the probability of specific outcomes, such as a Higgs boson appearing alongside a jet of other particles. These calculations are not simple arithmetic; they involve summing up the contributions of countless possible ways the collision could happen, including scenarios where particles briefly pop in and out of existence as "virtual" loops. The more precise the prediction, the more loops must be included, and the more complex the mathematics becomes.

For decades, physicists have been able to make predictions that include two such loops of virtual activity. However, to match the incredible precision of modern experiments, they must now push these calculations to include three loops. This leap introduces a massive increase in difficulty, turning manageable problems into mountains of algebra that were previously impossible to climb. The core of this challenge lies in a set of mathematical objects known as integrals, which represent the sum of all possible paths a particle could take. To solve the three-loop problem, researchers needed to find a complete set of these integrals for a specific, difficult scenario: a massless particle decaying into three others, with one of the incoming particles carrying extra energy. Until now, the full set of these building blocks was missing, leaving a gap in the theoretical predictions for the Large Hadron Collider.

A team of physicists has now filled this gap by constructing the complete set of these three-loop integrals. They did not simply calculate the numbers; they first had to find the right way to organize the problem. Imagine trying to solve a massive jigsaw puzzle where the pieces are not just shapes, but complex, shifting mathematical expressions. The researchers discovered that by looking at the structure of these expressions in a specific way, they could identify a special set of "master" pieces that serve as the foundation for all others. They found a way to rewrite these integrals so that their behavior became much simpler and more predictable, revealing a hidden order in the chaos. This new organization allowed them to solve the equations that describe how these integrals change as the energy of the collision changes.

The result is a complete library of 371 distinct master integrals for the most complicated family of these problems, along with solutions for many other related families. The researchers showed that despite the immense complexity of the individual pieces, the entire collection can be described using a surprisingly small and elegant set of mathematical functions. They identified a specific list of twenty "letters" that act as the alphabet for these functions. By combining these letters in specific sequences, they could write down the exact solution for every integral needed. This alphabet includes simple linear terms as well as more complex algebraic terms that had never been seen in this context before. The team proved that all the solutions fit together perfectly within this framework, confirming that no hidden complexities were lurking in the non-planar, or twisted, geometries of the particle interactions.

To make these abstract solutions useful for real-world physics, the team also developed two powerful methods to calculate the numerical values of these integrals for any specific collision scenario. The first method uses accelerated series expansions, which are like highly refined approximations that become more accurate the more terms are added. This approach allows for precise calculations even in difficult regions where the particles are moving in specific directions. The second method involves numerically transporting the solution along a path through the mathematical space of the collision, moving from a known starting point to the desired destination. Both methods were tested and shown to produce results with high precision, providing the necessary tools for experimentalists to compare their data with the most advanced theoretical predictions available.

The significance of this work extends beyond just filling a gap in a calculation. By providing the complete set of integrals, the researchers have enabled the first fully precise predictions for the production of vector bosons and Higgs bosons in association with a jet, including all the subtle effects of the strong nuclear force. This is a crucial step toward understanding the fundamental nature of matter and forces at the highest energies. The work also revealed that the mathematical structure governing these three-loop interactions is more constrained and orderly than one might expect, with the solutions confined to a compact space of functions. This suggests that the universe's laws, even at the most intricate levels of quantum mechanics, follow a deep and elegant logic that can be uncovered with the right mathematical tools. The findings are now available for other scientists to use, paving the way for the next generation of precision tests at the Large Hadron Collider.

Technical Summary: All three-loop four-point massless integrals with one off-shell leg

Problem Statement
This work addresses the computation of the complete set of massless three-loop four-point master integrals with one off-shell external leg. These integrals constitute the scalar building blocks for the purely virtual interference of the three-loop $2 \to 2$ amplitude with the tree amplitude. They are essential for achieving next-to-next-to-next-to-leading order (N $^3$ LO) QCD predictions for Higgs-boson plus jet ( $pp \to H+j$ ) and vector-boson plus jet ( $pp \to V+j$ ) production at the Large Hadron Collider (LHC). While planar families and specific non-planar subsets had been previously evaluated, the full set of non-planar families required for full-colour amplitudes remained incomplete.

Methodology
The authors employ a strategy based on the construction of integral bases directly from their integrands, rather than transforming existing bases via kinematic-dependent changes of variables. The core methodology involves:

Integrand Analysis for Canonical Bases: The authors seek a basis of integrals where the differential equations take the canonical form $dJ = \epsilon \, d\tilde{A} \, J$ . To identify such a basis, they construct integrands with only logarithmic singularities (simple poles) and unit leading singularities (non-zero maximal iterated residues equal to $\pm 1$ ). This is achieved through an improved iterated-residue method.
Handling Algebraic Obstructions: A significant technical challenge arises from intermediate residues becoming algebraic (involving square roots) during the residue calculation. The authors implement an extended search for variable transformations (including linear shifts, ratio substitutions, and rescalings) to manage these roots. Crucially, they avoid brute-force rationalization, which often introduces high-degree denominators, by computing residues directly on the algebraic cover when necessary.
Use of Generalized Cuts: For the most complex family (Family H), the authors utilize a hybrid approach. They determine numerators on specific generalized cuts (e.g., a double cut) where the residue calculation is feasible, and then lift these results to the full uncut family by adding "contact terms" (numerator terms proportional to the cut propagators) to cancel higher-order poles and non-constant singularities away from the cut.
Differential Equations and Solutions: Once the canonical bases are established, the authors solve the resulting differential equations. The solutions are expressed in terms of Chen iterated integrals over a specific alphabet of 20 letters. Boundary conditions are fixed using regularity at pseudo-thresholds and asymptotic behavior, transported to a common base point.
Numerical Evaluation: Two complementary strategies are presented for numerical evaluation:
- Accelerated Series: Local generalized series expansions around the corners of the kinematic triangle, improved by Bernoulli-like variable changes to accelerate convergence.
- Numerical Transport: Direct numerical integration of the canonical differential equations along paths in complexified kinematic space, using series expansions to provide boundary values at regular points.

Key Contributions and Results

Complete Set of Master Integrals: The paper provides the canonical bases and analytic solutions for all 46 integral families relevant to this kinematics. This includes 18 irreducible families and 4 reducible families (plus their crossings).
Family H: The most complex family, H, contains 371 master integrals with a 19-dimensional top sector. This family was previously a bottleneck for full-colour calculations. The authors successfully constructed its canonical basis, overcoming algebraic obstructions that had prevented previous attempts.
Function Space and Alphabet: All systems close on a common 20-letter alphabet, comprising 6 linear letters, 2 algebraic letters (involving a square root $\rho = \sqrt{-xyz}$ ), and 12 quadratic letters.
Transcendental Basis: The solutions for all 22 families (and their crossings) are expressed in terms of a minimal set of 1,318 "trident functions." This is only 36 functions more than the basis required for the leading-colour approximation. The authors demonstrate that despite the combinatorial complexity of the graphs, the function space is highly constrained, with new letters (algebraic and quadratic) appearing only at specific transcendental weights (starting at weight 3 for algebraic letters in families H and D).
Numerical Performance: The authors provide proof-of-concept numerical evaluations. The series expansion method achieves high precision (6–8 digits) across the decay region with truncation order 20. The numerical transport method achieves median accuracy of ~11 digits for weight-six functions in both decay and scattering regions.

Significance
The paper claims to supply the missing integral input required for three-loop vector-boson-plus-jet production beyond the planar approximation. By providing the complete set of master integrals and their solutions, the work enables the calculation of full-colour N $^3$ LO QCD predictions for $H+j$ and $V+j$ processes. The construction of the canonical bases directly from integrands with unit leading singularities serves as a robust method for handling the algebraic complexities inherent in non-planar three-loop topologies. The results confirm that the function space for these amplitudes is manageable and structured, adhering to conjectured adjacency conditions derived from cluster algebras.

More like this