Speaker
Description
In this talk, I present a new approach to studying next-to-leading power (NLP) contributions in inelastic processes, based on the background field method with the explicit inclusion of soft modes. I first detail the different structures contributing to the effective current operator at NLP in this framework and discuss how they differ from previous results. The method is then applied to semi-inclusive deep inelastic scattering (SIDIS) to derive the NLP factorization formula for the hadronic tensor at small transverse momentum. I explain how this formulation enables the definition of twist-3 distributions with properly subtracted rapidity and endpoint divergences. Finally, I present the most general form factors contributing to the $e + h \rightarrow e + \text{jet} + X$ cross section at this power accuracy and illustrate the formalism with phenomenological results for a specific azimuthal asymmetry. This work represents a step toward extending TMD factorization theorems beyond leading power, which is an essential ingredient for high-precision QCD predictions and future extractions of twist-3 distributions at facilities like the EIC.
