In this work, we determined the tilt angles of molecular units in hierarchical self-assembled materials on a single-sheet level, which were not available previously. This was achieved by developing a fast linescanning vibrational sum frequency generation (VSFG) hyperspectral imaging technique in combination with neural network analysis. Rapid VSFG imaging enables polarization resolved images on a single sheet level to be measured within a short time period, circumventing technical challenges due to long term optical setup instability. The polarization resolved hyperspectral images were then used to extract the supramolecular tilt angle of a self-assembly through a set of spectra-tilt angle relationships which were solved through neural network techniques. This unique combination of both novel techniques offers a new pathway to resolve molecular level structural knowledge of self-assembled materials. Understanding these properties can further drive self-assembly design from a bottom-up approach for applications in biomimetic and drug delivery researches.
Supporting Information for Imaging Orientation of A Single Molecular Hierarchical Self-Assembled Sheet: The Combined Power of A Vibrational Sum Frequency Generation Microscopy and Neural Network