Book Description

Since its commercialization in 1991, the rechargeable Li-ion battery has revolutionized how we live, work, and communicate. Further, rapidly declining costs have enabled the ongoing electrification of heavy industry, from automotive to power utilities and even aviation. Despite these important advances, reaching the targets required to minimize the effects of climate change requires improvements to battery technology across all levels of research and development – from basic materials chemistry to optimizations in manufacturing and operations in the field. This work represents a cross-cutting effort that spans the breadth of this range. In the experimental portion, the fundamental electrochemistry of magnesium was studied in the context of developing new positive and negative electrode materials for magnesium batteries. In particular, we show how considering the system as two interfaces (cathode-electrolyte and anode-electrolyte) rather than three components (cathode, anode, and electrolyte) helps explain some of the nuances of magnesium compared to lithium electrochemistry. In the second part of this work, we address applied problems in the growing field of battery data science and analytics. We develop software tools and models for industry-relevant problems in basic battery data management, cell lifetime forecasting and validation, and specific applications in electric aviation. Magnesium batteries represent a compelling candidate for the next generation of battery chemistries. It offers the prospect of safer operation at lower costs to manufacture and ~60% greater energy density compared to Li-ion batteries. However, many problems at the anode-electrolyte and cathode-electrolyte interfaces need to be resolved. Many electrolytes passivate magnesium metal and are oxidized at the cathode, and the nature of the kinetics (both reaction and diffusion) is unclear. In the two main prongs of this section of this work we propose 1) a coating that provides kinetic stability to the anode, and 2) an electrochemical protocol for studying magnesium cathode materials with greater clarity and reliability. Together we believe these approaches may be married to each other to develop a robust, high-energy magnesium battery based on a magnesium metal anode and a metal oxide cathode. We also explore battery data science as it applies to the new electric aviation industry. There are three main contributions. First, we report a user-friendly software environment for battery data science. It is designed to streamline data management, data cleaning, and data analysis to help bridge the gap between the domain expertise of most battery scientists and the tools needed as the field becomes increasingly data intensive. Second, we use a neural network model to extend state-of-the-art cell cycle life predictions to include a wider range of datasets and testing conditions, such as C-rate, cell chemistry, and temperature. Third, we train an outlier detection algorithm to rapidly identify weak cell blocks in an electric aircraft battery pack. This is significant in that identifying weak cells helps determine the remaining useful life in a battery system, which is of utmost importance for flight operators. Each of these constitute pieces of an effort that is oriented toward developing standard operating procedures for battery safety, performance, and maintenance for electric aviation.