A Bottom-up Computational Study of Doped Amorphous Carbon Clusters as Precursors of Carbon Nanodots

In this work, we present a bottom-up computational study of unsaturated amorphous carbon clusters partially doped with nitrogen and other typical chemical groups. We focus on their structure and electronic properties with the aim of providing possible models of the amorphous portions of the core of carbon nanodots, a class of fluorescent carbon-based nanoparticles whose complex atomistic structure is not yet fully understood. Using the GOAT-EXPLORE algorithm in combination with semiempirical and DFT methods, we generated and optimized a broad ensemble of structural isomers for C_<i>n</i>X clusters of up to 35 atoms of carbon, with X including nitrogen and typical functional groups found in the carbon nanodots (X = OH, NH₂, COOH, and CONH₂). Structural and electronic parameters were evaluated to understand how size, doping, and functionalization impact electronic delocalization, stability, and potential reactivity. Our findings highlight a structural evolution from linear to polycyclic and cage-like motifs with increasing size and demonstrate that functionalization mainly influences local electronic environments without drastically affecting the overall architecture.