Nanoscale Spectroscopic Imaging of Carriers and Heat in Photocatalysts

Photoexcited carriers (electrons and holes) perform photocatalytic solar energy conversion reactions when they transfer from a solid-state material and into a nearby or surface-adsorbed molecule. For example, light-driven water splitting occurs when molecular hydrogen and oxygen are produced from water because electron transfer drives hydrogen evolution, whereas hole transfer drives oxygen evolution. These photoexcited carriers often get trapped at nanoscale features such as dopants, defects, and grain boundaries, limiting their mobility and potentially reducing photocatalytic efficiency. However, identifying which nanoscale properties hinder solar energy conversion remains a key challenge as this requires a technique that can separately image carriers and photothermal heating on the timescale relevant to photocatalysis. Here, we develop photomodulated electron energy-loss spectroscopy (EELS) in the scanning transmission electron microscope (STEM) to map photocarrier and heat distributions at sub-nanometer length scales. This dissertation presents the experimental and computational developments critical for technique development. We first introduce EELS and the fundamental working principles of ultrafast and optically coupled TEM. We demonstrate the theoretical framework needed to differentiate carriers from photothermal heating in two systems: plasmonic core-shell nanoparticles and crystalline silicon. Then, we apply this method to image photoexcited carriers localized at surface oxygen vacancies and along select nanoparticle facets in water-splitting photocatalysts. We lastly perform optically coupled liquid phase TEM (LPTEM) of similar photocatalytic nanoparticles in carbon film liquid cells, aiming to expand this technique to operando measurements of photocatalysts in their working conditions. Overall, this thesis demonstrates photomodulated STEM-EELS as a tool for imaging photocatalytic structure–function relationships down to the atomic scale.