Achieving nondestructive, high-precision quantitative elemental analysis at the nanoscale has long been a challenge in materials characterization. The proposed Grayscale Bayesian Elemental Quantification Analysis (GBEQA) method effectively circumvents issues such as insufficient detection depth, destructiveness, and matrix effects encountered by conventional techniques by integrating grayscale information from synchrotron radiation nano-CT with a Bayesian inversion framework. By constructing a grayscale probability model and generating probability mass functions through convolution with standard materials, GBEQA achieves unsupervised element quantification without requiring prior knowledge for the first time. The validation on FeNi alloys and Ni0.9Co0.05Mn0.05(OH)2 (NCM90 precursor) demonstrates high agreement with inductively coupled plasma atomic emission spectrometry (ICP-AES) while successfully characterizing nanoscale element distribution heterogeneities inaccessible to conventional techniques. This study establishes a novel paradigm for elemental analysis that combines nondestructive, 3D resolution, and universality, holding significant implications for advancing research on complex systems such as alloys and energy materials.
Zhang et al. (Thu,) studied this question.