Developing novel non-fullerene acceptors (NFAs) by modifying A ₁-DA ₂D-A ₁ Y6 molecules is an effective strategy to improve the energy conversion efficiency of organic solar cells (OSCs). To understand the mechanism of regulating the photovoltaic performance by modifying the central fused-ring, end-group and inner-chain segments of Y6, D18:Y6, D18:AQx-2, D18:AQx-18, D18:BTIC-γ-2CN and D18:Z9 OSCs were systematically studied based on extensive quantum chemistry calculations, and the impacts of different chemical groups on molecular properties and photovoltaic performances were analyzed. The substitution of benzothiadiazole in the central fused ring of Y6 with quinoxaline, substituting quinoxaline with phenazine, introducing benzonitriles at the end group and phenyl groups into inner side chains enhance molecular skeleton planarity, slightly elevate the highest occupied molecular orbital energy and the lowest excitation energy, and enlarge light absorption efficiency. Introducing quinoxaline and phenyl groups cause the reduction of electrostatic potential difference between D18 and NFAs, on the contrary, introducing phenazine and benzonitriles induce the opposite effects. Introducing quinoxaline and phenazine generate negligible effects on charge transfer (CT) energies, whereas introducing benzonitrile and the phenyl group increase CT energies. Introducing phenazine and phenyl group generate sufficient energy difference between local excitation and CT to conquer exciton binding. The rate constants calculated using Marcus theory indicate that all molecular modifications of Y6 reduced exciton dissociation rates. However, the introducing benzonitrile increase CT and suppress charge recombination rates. The results reveal the inherent relationship between molecular structure and photovoltaic performance, providing the theoretical basis for design and development of NFAs.