“…We demonstrate how the developed methodology can be used to model excited-state dynamics in unprecedentedly large Si nanocrystals (NCs), as well as in periodic solids, such as a graphitic carbon nitride monolayer and a titanium-based metal–organic framework (MOF). In the last decade, colloidal NCs have demonstrated their potential to replace the forerunner silicon solar cells that work similarly to dye-sensitized solar cells while having flexibility in production with size-tunable optoelectronic properties due to their quantum confinement properties. − Among them, silicon NCs (Si NCs) are of particular interest due to lack of their toxicity, the element’s pre-existing prevalence in the electronic industry, and its natural abundance. − Our second system of interest, the graphitic carbon nitride monolayer, has recently attracted a lot of interest as a material for photocatalytic applications such as solar-driven water splitting. − The third system is the MIL-125-NH 2 MOF, which is based on titanium and has different photovoltaic and photocatalytic applications such as in CO 2 reduction. − In this work, we assess the quality of electronic structure and NACs as computed using the xTB, the GFN2 parametrization, and DFT methods. We assess the new scheme for computing NACs that utilizes the Libint2 library for analytical computation of time–overlap integrals.…”