LAUSR

Abstract In this work, a large data set of SiCO amorphous clusters was created from predefined primary building units (PBUs), connected statistically using custom made Python interface to GULP program and employing ReaxFF force field. A series of amorphous SiCO clusters with various content of carbon incorporated into silica matrix, characterized by well-defined geometry were obtained, confirming that the proposed in this work bottom-up approach for amorphous materials modelling can be a good alternative to other methods, typically found in a literature (amorphous structure based on strongly distorted crystal structures, from a defined set of atoms via MD melt quenching or just by hand). The analysis of the structural properties of obtained clusters showed that the increasing content of carbon incorporated into cluster results in small elongation of Si–O bonds as well as shortening of Si–C bonds. Carbon content in a structure affects also the number of spontaneously formed Si3(C,O)3 cycles – the more carbon the greater number of cycles can be found.