UC Berkeley

Lithium-ion batteries will play a large role in transitioning from fossil fuels to renewable energy sources. Given the scarcity of cobalt as well as the serious human rights abuses involved in its sourcing, it is imperative that we move towards cobalt-free battery cathode materials. Lithium-excess disordered rock salt materials (DRX) are a new class of materials that offer a cobalt-free alternative to traditional state-of-the-art nickel manganese cobalt (NMC) batteries that are currently ubiquitous in consumer electronics like cell phones and laptops as well as electric vehicles. While promising, DRX materials still need optimization before they can be manufactured and used as replacements for current lithium-ion battery technologies. Oxygen loss at the surface of DRX cathodes at high states of charge leads to unacceptable capacity and voltage fade. Partial substitution of oxygen with fluorine has been shown to mitigate oxygen loss and even adds capacity to DRX cathodes by lowering the redox-active transition metal valence state. In order to further improve DRX cathodes, a better understanding of the particle surface and morphology is needed, as well as further insights into oxygen loss at the surface of the cathode.In this work, we develop a novel method of modeling disordered oxide surfaces and use it to study the highly fluorinated archetypal DRX material Li2MnO2F. Surface energies were calculated for the low miller index facets {100}, {110}, {111} and {112} using varying surface chemistries. Boltzmann averaging was used to obtain a representative surface energy for each facet weighted by both multiplicity and energy. It was found that the {100} type surface is lowest in energy and dominates the equilibrium particle shape resulting in a cubic shape. The low-energy particle surfaces were also enriched in lithium and fluorine which bode well for lithium mobility at the particle surface. The same low miller index facets, enriched in lithium and fluorine, were then studied for their propensity to lose oxygen at different states of delithiation: 0%, 25% and 50%. Boltzmann averaging was again used to calculate a representative oxygen evolution energy (E ̃O) (the resistance to losing oxygen) for 1 2 each facet and delithiation state. As expected, the E ̃O at the surface lowered for all facets as we progressively delithiated Li2MnO2F. However, in contrast to traditional layered oxide cathode materials, Li2MnO2F remained much more robust to oxygen loss. The {110} and {112} surface facets were most resistant to oxygen loss, with their E ̃O remaining positive even at the 50% delithiation state. Taking a closer look at the {100} type facet, which is predicted to be most exposed on the particle surface, we found that coordination with lithium is highly predictive of a low E ̃O and hence a propensity to lose oxygen at the surface. The relationship between fluorination and mechanical cracking in DRX cathodes was also explored. Well-formed single grains of a DRX oxide and oxyfluoride were prepared and used to explore their intrinsic chemomechanical behaviors. The unique role of fluorine in modulating the nanoscale evolution of oxyfluoride mechanics was shown, where directional cracking was observed upon delithiation. In contrast, delithiation leads to randomly oriented cracks in the DRX oxide baseline sample. X-ray nanotomography measurements and 2D XANES imaging revealed the correlation between the aligned cracks and the variation in Mn oxidation state. Crystallographic analysis through TEM/SAED further determined <001> directions as the preferential cracking direction in the oxyfluoride. DFT calculations of different surface decorations revealed that fluorination leads to an increased concentration of Li+ on the (001) planes, which is expected to lead to preferential Li movements along the <001>-family directions. The unique chemomechanical behavior of DRX oxyfluoride enables its enhanced electrochemical cycling stability. Battery recycling as a means to meet battery demand was also explored. A model was developed to predict the amount of recycled battery power over time that could be used to meet battery demand. Even with generous assumptions, our model found that it is not likely that recycled battery material would make up half of global demand until around 2050. If demand does not level off in the future, battery recycling will never be able to meet demand because of the time lag between battery manufacturing and recycling. The Reaction Calculator App was designed for the Materials Project Website. This app allows users to input the reactants and products for a reaction and get out a balanced reaction and reaction enthalpy which are useful for predicting and planning reactions and syntheses. The logic of the app as well as the user-friendliness and robustness to error was explored.