Machine-learning-revealed reaction statistics via 3D spectroimaging for copper sulfidation of adhesive layers in rubber/brass composite

Abstract The sulfidation of Cu derived from Cu-Zn alloy (brass) in S-containing rubber, which is used for plating steel-cord-reinforced rubber tires, is suggested to be the key reaction for adhesive behavior between brass and rubber in tires. However, the heterogeneous structures of rubber/brass interfaces have prevented us from understanding the sulfidation of metallic Cu in brass and the formation of Cu sulfides at the brass surface and buried rubber interface. We report visualizing the three-dimensional spatial location and chemical states of Cu species in a rubber/brass composite by three-dimensional (3D) X-ray spectroimaging with X-ray absorption fine structure-computed tomography during the aging process of the rubber/brass composite for the first time. Machine-learning derived reaction statistics for the 3D spectroimaging data revealed the reaction mechanism of the Cu sulfidation in the heterogeneous rubber/brass composite.