Precision QCD corrections to gluon-initiated diphoton-plus-jet production at the LHC

TL;DR

This paper advances the precision of QCD calculations for gluon-initiated diphoton-plus-jet production at the LHC by computing two-loop amplitudes, developing validation tools, and applying machine learning for amplitude evaluation.

Contribution

It provides the first full-colour virtual QCD corrections to diphoton-plus-jet production via gluon fusion, along with a library of infrared functions and machine learning techniques for amplitude evaluation.

Findings

Computed next-to-leading QCD corrections for diphoton-plus-jet production.

Developed a library of infrared functions up to NNLO in QCD.

Made amplitudes publicly available for phenomenological studies.

Abstract

In this thesis, we present recent advances at the precision frontier of higher-order quantum chromodynamics (QCD) calculations. We consider massless two-loop five-point amplitudes, with a particular focus on diphoton-plus-jet production through gluon fusion. We build a library of infrared functions up to at most next-to-next-to-leading order (NNLO) in QCD, which can be used to validate amplitudes and construct counterterms in subtraction schemes at NNLO. We review progress in the novel use of machine learning technology to optimise the evaluation of amplitudes in hadron collider simulations. We present the full-colour virtual QCD corrections to diphoton-plus-jet production through gluon fusion, discussing the new techniques developed to calculate these non-planar two-loop amplitudes. We use these amplitudes to compute the next-to-leading QCD corrections to the differential cross sections of diphoton-plus-jet production through gluon fusion at the Large Hadron Collider. We also present the leading-colour double-virtual corrections to hadronic trijet production. All derived amplitudes are made available in a public implementation that is ready for further phenomenological application.