This project will establish Energy Materials Chemistry Integrating Theory, Experiment and Data Science (EM-CITED) as a multidisciplinary research effort focused on accelerating discovery of scientific knowledge via incorporation of data science and artificial intelligence in materials chemistry. It will use data from theory and high throughput experimentation to build models that address fundamental gaps in understanding of the synthesis and stability of functional energy materials in extreme environments. EM-CITED will have 3 focus areas of research: (i) Open Representations of Energy Materials, (ii) Energy Materials Synthesis Prediction, and (iii) Catalyst Evolution Prediction and Classification. The Open Representations focus area addresses a critical limitation in the adoption of machine learning algorithms in materials and solid-state chemistry research, the present lack of appropriate machine representations of energy materials. Data representations are key to predicting beyond the confines of existing data, motivating the plan to learn representations using data science, resulting in a suite of data representation schemes that will be open to the community to enable their specific implementation of machine learning algorithms. Learning appropriate machine representations for materials properties will be critical to the second focus area, which will establish a framework for using first principles calculations to create a descriptor database from which interpretable models will be trained to predict the phases that are synthesizable via a specific method. The establishment of such models will enable searching for synthesis parameters for a given candidate material, which will help bridge the present gap between high throughput computational materials design and experimental realization of the resulting materials. While this focus area will include analysis of surface energies and diffusive transport in a breadth of thermal and reactive-atmosphere environments, the final focus area explores materials dynamics in even farther-from-equilibrium conditions. Many energy technologies incorporate (photo)electrocatalysis where materials are operated in extreme chemical, electrochemical, and illumination environments. While discovery of operationally stable electrocatalysts is a primary goal in several basic energy science areas, the dynamic evolution of materials in far-from-equilibrium electrochemical environments is a cross-cutting research topic for which a predictive model will require integration of materials and catalysis sciences. The interdisciplinary EM-CITED team will establish principled methods for leveraging advancements in computer science to accelerate scientific discovery, which will require not only applying machine learning methods but also advancing scientific reasoning capabilities with Artificial Intelligence. By making datasets, physical representations, and algorithms open to the community, EM-CITED will transform basic energy research by providing broadly-applicable tools for building predictive models and demonstrating the principled integration of computer science in the physical sciences.