Next-to-leading power endpoint factorization and resummation for off-diagonal “gluon” thrust

The lack of convergence the convolution integrals appearing in next-to-leading-power (NLP) factorization theorems prevents applications existing methods to resum power-suppressed large logarithmic corrections collider physics. We consider thrust distribution two-jet region for flavour-nonsinglet off-diagonal contribution, where a gluon-initiated jet recoils against quark-antiquark pair, which is power-suppressed. With help operatorial endpoint conditions, we obtain formula, individual terms are free from divergences convolutions and can be expressed renormalized hard, soft collinear functions four dimensions. This allows us perform first resummation endpoint-divergent SCET$_{\rm I}$ observables at leading accuracy using exclusively renormalization-group methods. presented approach relies on universal properties limits may serve as paradigm systematic NLP other $1\to 2$ $2\to 1$ physics processes.