The bending and orientation of layers, along with any potential pore structures in disordered carbons, are crucial for their performance in many applications. Therefore, it is essential to characterize them. Traditionally, detection and analysis of fringes in high-resolution transmission electron microscopy (HR-TEM) images are used for characterization. However, fringes are affected by instrumental artifacts, such as spectral modulation by the microscope’s contrast transfer function and multiple overlapping structures in projection, which can lead to misinterpretation. Here, we propose to visualize and further analyze the nature of curved, graphite-like stacked layers in disordered carbons by examining changes in Bragg disc orientation in scanning electron nanobeam diffraction (SEND) patterns. The new approach uses non-negative matrix factorization (NMF) to reduce the complexity of the SEND patterns. It enables the detection and characterization of curved layers over a larger field of view than is readily achievable with individual HR-TEM images, thereby enhancing the statistical relevance of the methodology. Moreover, the approach yields structural information about curved layers of multiple intertwined and overlapping regions. It allows separation of contributions to overlapping layers of different pores and enables independent analysis of their curvatures. The potential of the new approach is demonstrated by analyzing simulated and experimentally measured SEND data from disordered carbons.
Leave a Reply