Transverse Λ polarization in e+e- collisions

In this paper we study transverse polarization of Lambda hyperons in single-inclusive leptonic annihilation. We show that when the transverse momentum of the Lambda baryon is measured with respect to the thrust axis, a transverse momentum dependent (TMD) factorization formalism is required and the polarization is generated by the TMD polarizing fragmentation function (TMD PFF), D-1T(perpendicular to). However, when the transverse momentum of the Lambda baryon is measured with respect to the momentum of the initial leptons, a collinear twist-3 formalism is required and the polarization is generated by the intrinsic collinear twist-3 fragmentation function D-T. Thus while these measurements differ from one another only by a change in the measurement axis, they probe different distribution functions. Recently, Belle measured a significant polarization in single-inclusive Lambda baryon production as a function of the transverse momentum with respect to the thrust axis. However, this data can in principle be re-analyzed to measure the polarization as a function of the transverse momentum of the Lambda baryon with respect to the lepton pair. This observable could be the first significant probe of the function, D-T. In this paper, we first develop a TMD formalism for Lambda polarization; we then present a recent twist-3 formalism that was established to describe Lambda polarization. Using the TMD formalism, we demonstrate that the Lambda polarization at OPAL and Belle can be described using the twist-2 TMD factorization formalism. Finally, we make a theoretical prediction for this polarization in the collinear twist-3 formalism at Belle. (C) 2021 The Author(s). Published by Elsevier B.V.

Automated summary

In this paper, we investigate the transverse polarization of Lambda hyperons in single-inclusive leptonic annihilation, revealing that this polarization is influenced by transverse momentum dependent and collinear twist-3 formalisms. Different measurement methods probe different distribution functions, and data can be reanalyzed to extract more information.