Search for: All records

The surfaces of colloidal nanocrystals are frequently passivated with carboxylate ligands that exert significant effects on their optoelectronic properties and chemical stability. The binding geometries of these ligands are often experimentally investigated using vibrational spectroscopy, but the interpretation of the IR signal is usually not trivial. Here, using machine-learning (ML) algorithms trained on DFT data, we simulate an IR spectrum of a lead-rich PbS nanocrystal passivated with butyrate ligands. We obtain good agreement with the experimental signal and demonstrate that the observed line shape stems from a very wide range of “tilted-bridge”-type geometries and does not indicate the coexistence of “bridging” and “chelating” binding modes as has been previously assumed. This work illustrates the limitations of empirical spectrum assignment and demonstrates the effectiveness of ML-driven molecular dynamics simulations in reproducing the IR spectra of nanoscopic systems.