UC Berkeley

The tremendous success and growth of Lithium (Li)-ion based energy storage in a broad range of applications is likely to strain our natural resources. Projected growth of Li-ion production towards 1 TWh/year will require more than a million tons of Co/Ni combined, which constitutes a very sizeable fraction of the annual production of these metals. The recent development of Li-excess cation-disordered rocksalt (DRX) cathodes, which require no separate Li layer and transition metal (TM) layer in the cation sublattice, is providing an avenue for the Li-ion battery field to develop high energy density cathodes with more abundant and less expensive metals, such as Mn, Fe and Ti. Unlike the traditional layered NCM cathodes in which Li ions transport within the well-defined Li layers, the Li ions diffuse through a “3-D” Li rich pathway (‘0-TM’ channel) in DRX cathodes, i.e., the feasible Li hops require that the intermediate tetrahedral sites along the Li diffusion path are coordinated only by Li (no transition metal). The complexity of this class of DRX cathodes mainly lies in the fact that lacking in long-range order though, DRX cathodes, in most cases, present various types of cation short-range order (SRO), which controls the frequency and connectivity of Li migration channels, thus, macroscopic Li transport.In this thesis, two general strategies to engineer the cation SRO in DRX cathodes will be demonstrated through compositional design, combining electrochemical tests, advanced characterizations and computational investigations. The first part of the thesis will talk about the control of SRO in DRX cathodes through fluorine doping on anion sublattices. The strong Li-F interaction modifies the cation SRO in DRX materials by forming Li-rich domains around F ions. Four well-chosen compositions within the Li–Mn–O–F DRX chemical space with different Li content and F content were synthesized, characterized, electrochemically tested and modeled: Li1.333Mn(III)0.667O1.333F0.667 (HLF67), Li1.333Mn(III)0.5Mn(IV)0.167O1.5F0.5 (HLF50), Li1.333Mn(III)0.333Mn(IV)0.333O1.667F0.333 (HLF33), and Li1.25Mn(II)0.167Mn(III)0.583O1.333F0.667 (LLF67). While all compositions tested achieve higher than 200 mAh/g initial capacity the material with high Li-excess (1.333 per formula unit, LixMn2-xO2-yFy) and moderate fluorination (0.333 per formula unit) achieves 349 mAh/g initial capacity and 1068 Wh/kg specific energy. Higher fluorination (0.667 per formula unit) leads to a less efficient Li diffusion network, resulting in a decrease in capacity (specific energy) to be 256 mAh/g (822 Wh/kg). However, the activation of Mn2+/Mn4+ redox enabled by F substitution significantly improve the cycle life of DRX cathodes, with more than 85% retained after 30 cycles even upon charging to 5.0V. It can thus be concluded that the Li-site distribution, which can be significantly modified by fluorine substitution, plays a more important role than the metal-redox capacity in determining the initial capacity, whereas the metal-redox capacity is more closely related to the cyclability of the materials. A design map will also be presented, which shows that the macroscopic Li transport is strongly controlled by the Li-excess level and F content within a fixed DRX chemical space and can be correlated to the observed capacities of a DRX cathode, as corroborated by electrochemical tests. Similar computational evaluations will also be shown in various different DRX chemistries to demonstrate that fluorination is a generalized strategy to engineering SRO in DRX cathodes. The second part of the thesis will talk about the control of SRO in DRX cathodes through TM doping on cation sublattices. With a fixed Li and F content, TM doping can frustrate the cation SRO in DRX cathodes by increasing configurational entropy to form so-called high-entropy (HE) DRX, thus improve the Li transport, i.e. capacity and rate capability in a significant way. As a demonstration, a group of DRX cathodes containing two, four or six TM species are designed. It is shown that SRO in these DRX cathodes systematically decreases, as a consequence, energy density and rate capability systematically increase as more metal cations are mixed together, even though the total metal content are fixed. Remarkably, a HE DRX cathode with six TM species achieves 307 mAh g−1 (955 Wh kg−1) at a low rate (20 mA g−1), and retains more than 170 mAh g−1 when cycling at a high rate of 2,000 mA g−1. The compatibility between various TM ions in DRX compounds is also investigated to facilitate future experimental design and a phase-pure HE DRX compound containing twelve TM species was successfully synthesized as a proof-of-concept.