The production of leading neutrons in deep-inelastic scattering is usually considered as a tool to measure the pion structure function at small $x$ accessible at HERA. The main obstacle is the lack of reliable evaluations of the absorption corrections, which significantly suppress the cross section. We performed a parameter-free calculation within the dipole approach and found the absorption corrections to be nearly as strong, as for neutron production in $pp$ collisions. We also included the significant contribution of the isovector Reggeons with natural ($\rho$, $a_2$) and unnatural ($a_1$, $\rho \mathrm{\text{−}}\pi$ cut) parity with parameters constrained by phenomenology. With a certain modeling for the pion-to-proton ratio of the structure functions we reached good agreement with data from the ZEUS and H1 experiments, successfully reproducing the observed dependences on the fractional neutron momentum $z$, the photon virtuality $Q^2$, and the transverse momentum transfer $q_T$.

• Received 2 May 2012

DOI:https://doi.org/10.1103/PhysRevD.85.114025