The discovery of new molecular and material composites can dramatically advance technology for how energy is used and stored. The possible design space for such new materials is large, but discovery is limited by the time and difficulty in carrying out each experimental study by trial and error. Transition metals are elements in the periodic table that impart unique, useful properties to materials, molecules and their chemistries through unusual behavior related to their open-shell, localized electrons. First-principles, computational screening has emerged as a powerful tool for the design of function based upon such ions. Approximate density functional theory (DFT) is a computationally efficient method suitable for high-throughput simulation for discovery. However, current approximate implementations of DFT produce critical errors and reduce DFT’s accuracy particularly for the study of the most promising new materials (i.e., with transition metals). The objective of this project is to separate and assess distinct sources of error in approximate DFT for transition metal chemistry. This project will develop methods to enable this error assessment and then identify strategies to mitigate these errors without sacrificing present computational efficiency in approximate DFT for high-throughput screening. The project will also apply these new methods to enable high-throughput screening of new materials for use in redox flow batteries, which are a key strategy for large-scale energy storage.