Non-factorizable contributions to deep inelastic scattering at large x
Ben D. Pecjak
TL;DR
The paper investigates DIS near the endpoint $1-x \sim \Lambda_{\rm QCD}/Q$ using soft-collinear effective theory (SCET) and a regions analysis in the Breit frame. It identifies hard, anti-hard-collinear, collinear, and soft-collinear modes as essential, and shows that soft-collinear interactions couple soft and collinear sectors in a way that prevents a clean perturbative factorization into hard, jet, and soft functions. One-loop matching reveals that finite pieces entail convolutions among jet, soft, and soft-collinear structures, while UV poles cancel only after accounting for soft-collinear contributions, signaling a breakdown of factorization driven by soft-collinear dynamics. Consequently, for $1-x$ in the region $\Lambda_{\rm QCD}/Q$, DIS cannot be factorized perturbatively into independent scale-separated functions; a multi-scale effective theory or non-perturbative input is required, though standard factorization may still hold away from the endpoint. The work also contrasts the SCET-based findings with diagrammatic endpoint analyses and with prior SCET treatments, highlighting the central role of the soft-collinear mode in redefining endpoint factorization in DIS.
Abstract
We use soft-collinear effective theory (SCET) to study the factorization properties of deep inelastic scattering in the region of phase space where 1-x = O(Lambda_{QCD/Q}). By applying a regions analysis to loop diagrams in the Breit frame, we show that the appropriate version of SCET includes anti-hard-collinear, collinear, and soft-collinear fields. We find that the effects of the soft-collinear fields spoil perturbative factorization even at leading order in the 1/Q expansion.
