Method for extracting the intrinsic diffusion coefficient from grain boundary diffusion depth profile

Numerical research is done on grain boundary diffusion of Cu-Ni bilayer thin films. Together, the composition-dependent interdiffusion coefficient and the position of the grain boundary are considered. Fick's second law provides a quantitative assessment of the effects of grain size, volume, grain boundary diffusion coefficient, diffusion temperature, and time on the depth profile. Regarding the situation where the plateau values of the Ni concentration in the Cu layer and the Cu concentration in the Ni layer are almost the same for the case where the Ni intrinsic diffusion coefficient is significantly bigger than that for Cu, a detailed explanation is provided. We may produce a profile that is like the direct volume diffusion profile with a relative inaccuracy of 7.53% by using a method that is provided to eliminate the impacts of grain boundaries and grain size on the grain boundary diffusion profile. For most of the circumstances covered in this study, the technique is effective. The transformation of the grain boundary diffusion profile yields a composition-dependent interdiffusion coefficient, and the calculation's outcome is consistent with the theoretical conclusion drawn from the Darken equation. Thin films of Cu (150 nm)/Ni (150 nm)/SiOx/Si (100 nm) were produced as examples, and Auger electron spectroscopy was used to determine the respective depth profiles. The calculated diffusion parameters, namely, the pre-exponential factor and the activation energy, are well within the bounds of the available data.

Automated summary

Numerical research is conducted on grain boundary diffusion of Cu-Ni bilayer thin films, considering the composition-dependent interdiffusion coefficient and the position of the grain boundary. Fick's second law provides a quantitative assessment of the effects of various factors on the depth profile. A method is proposed to eliminate the impacts of grain boundaries and grain size on the grain boundary diffusion profile, yielding a profile similar to the direct volume diffusion profile. The calculated diffusion parameters are within the bounds of available data.