A new method to measure pore radius distribution of powders

The pore radius distribution plays a significant role in characterizing the porous structure of powder particles, yet quantifying radii spanning 100 nm-10 µm-including both open and closed pores-remains challenging. Here, we propose a new method to measure the pore radius distribution. The method comprises two parts. First, the cross-section radius distribution is measured by an automated routine combining scanning-electron-microscopy (SEM) and deep-learning models. Second, a novel algorithm was developed to convert the cross-section radius distribution into a pore radius distribution. This requires a numerical solution to Wicksell's corpuscle problem, where the new algorithm outperforms the commonly used Saltikov-GCO method. We apply the proposed method to powder samples and compare the result with data from synchrotron X-ray tomography. Our approach provides more information on the distribution and agrees with the result of synchrotron X-ray tomography at a larger scale. As a secondary outcome, the algorithm can also be applied to geology and metallurgy when 3D grain size distribution is calculated from 2D grain size distributions.