Photophysics of Molecular Probes for Amyloid-β Detection: Computational Insights into the Roles of Probe Linker and...

In this computational study, density functional theory (DFT) and time-dependent DFT methods (TD-DFT) were employed to study the optical properties of six families of molecules with donor (D), bridge (B), and acceptor (A) fragments that have potential for use as fluorescent molecular probes for the early detection of Alzheimer's disease. After validating our computational method against experimental data, using X-ray and absorption data, the equilibrium geometries and wave functions of the ground and first singlet excited states were systematically studied. Our simulations demonstrate that the S₁ states of these rod-like D-B-A fluorescent probes are twisted intramolecular charge transfer states with a predominant highest occupied molecular orbital-least unoccupied molecular orbital (HOMO-LUMO) character, the former localized primarily at the donor, whereas the latter at the acceptor site. Moreover, the influence of the bridge, donor, and acceptor fragments on molecules' absorption energies is explored, highlighting the influence of double and triple bonds and some specific modifications on the acceptor side, including the addition of electronegative atoms, pyranone derivatives, and their functionalization. By having the absorption energies of 324 probes in hand, machine learning models were trained to predict the absorption energies of molecules. The models were found to be predictive, which suggests a potential that predictive models for other crucial properties, such as emission and quantum yield, can also be trained if suitable training data sets are made available.