Title: Localization of peripheral reactions and sensitivity to the imaginary potential

Abstract: The aim of the present study is to make for the first time in the literature a systematic and quantitative assessment of the evaluation of the imaginary part of the optical potential calculated within the folding model and its consequences on the localization of surface reactions. Comparing theoretical and experimental reaction cross sections, for some light projectiles on a $^9$Be target, it has recently been shown that a single-folded s.f. (light-) nucleus-$^9$Be imaginary optical potential is more accurate than a double-folded d.f. optical potential. Within the eikonal formalism for the cross sections and phase shifts, the single-folded potential was obtained using a n-$^9$Be phenomenological optical potential and microscopic projectile densities. This paper is a follow-up in which we systematically study a series of different light and medium-mass projectile induced reactions on $^9$Be. Our results confirm that the s.f. cross sections are larger than the d.f. cross sections and the effect increases with the projectile mass. Furthermore the strong absorption radius parameter extracted from the $S$ matrices calculated with the s.f. has a stable value $r_s$ =1.3 - 1.4 fm for all projectile masses in the range of incident energies 40-100AMeV. This indicates that a clear geometrical separation can be made between the region of surface reactions, the region of strong absorption into other channels and the region of weak nuclear interaction. The d.f. results are instead much scattered and the separation between surface reactions and other channels does not seem to be consistent. Excellent agreement with recent experimental results confirms the validity of our approach.