Algorithmic study of the structure of graphene-containing shungite based on scanning electron microscopy data

A brief review of the general characteristics of graphene-containing high-carbon shungite as a material for creating shields for electromagnetic radiation is presented. It is noted that the conductive properties, which determine the shielding ability, depend largely on the structure and distribution of graphene layers within the material. The scanning electron microscopy is noted as the most important tool to study the structure of shungite at the nanoscale. The carbon distribution maps obtained by this method were examined by means of structural discretization on three levels. Three gradations of homogeneity within one discretization cell were noted: high, when graphene layers are almost continuous, medium, when graphene layers have discontinuous character, low, when well-defined graphene layers are absent. The study was carried out by two methods, manual and machine. The manual processing technique is described using a movable mask sequentially passing through the cells of the map. The technique of machine processing consisting in Shannon entropy measurement for each cell separately is described. Comparison of the results of manual processing with machine processing is performed. It was found that, in general, the coincidence between the methods is 67%, and the lack of coincidence is 33%. At the same time, the best coincidence takes place in the medium degree of ordering (26%), which is noticeably higher than the coincidence in the high (17%) and low (23%) degrees of ordering. On the basis of the results obtained, it was concluded that there is a high degree of efficiency in replacing manual processing with machine processing. Some recommendations for further development of the work are given.