Relationships Among Structure, Magnetism and State of Charge in Positive Electrode Materials for Metal-Ion Batteries

Abstract

Polyanionic framework materials containing 3d transition metals such as iron, cobalt and manganese are attractive candidates as electrodes in lithium and sodium ion batteries due to their thermal stability, long cycle life and environmental friendliness. LiFePO4 is already used in some commercial lithium ion batteries as a positive electrode material where these are key attributes, but it still has lower energy density and higher costs compared to the more commonly used LiCoO2. This thesis describes a combined physical properties and magnetic structures some of these materials, aimed at improving our understanding of their solid-state chemistry, and ultimately their performance as battery materials by relating those physical and magnetic properties to the state of charge of the battery. A variety of polyanionic materials including phosphates, pyrophosphates and silicates were prepared using solid-state synthesis. All compounds were intensively characterized using specific heat capacity and magnetic measurements, X-ray, neutron and synchrotron X-ray diffraction techniques. Low-temperature neutron diffraction was used to solved and refined the magnetic structures. During the lithium extraction process, the magnetic properties can vary significantly because it involve redox reaction of transition metals. Measuring the magnetic properties of working electrode materials can therefore potentially provide information about local structural changes including the introduction of defects, decomposition and phase segregation. The magnetic properties of chemically delithiated samples were also studied so that they could be used as reference materials. In the first part of the thesis, the phosphates family AM1-xM′xPO4 (A = Li, Na; M = Mn, Fe; M′ = Zn, Mn, Fe) are intensively studied. These phosphates are modified in two ways: by doping with magnetic and non-magnetic transition metals. It was found that all compounds exhibit antiferromagnetic ordering at low temperatures, but the nature and ordering temperatures depend on doping. In the course of this work, the magnetic structures of two types of sodium phosphate were determined for the first time,triphylite and maricite NaFePO4. The triphylite type showed similar crystal and magnetic properties as LiFePO4, while the maricite type demonstrated a transition from commensurate (T < 12 K) to incommensurate (12 < T < 13 K) magnetic phases. A spin-flop transition in the commensurate phase was also observed. These results are discussed in the context of spin frustration on the Fe2+ sites. The second type of cathode material studied was the pyrophosphate A2M1-xM′xP2O7 (A = Li, Na; M = Fe; M′ = Co, Mn). Varying the compositions of these materials led to significant changes in crystallographic and electronic structure with remarkable effects on the magnetic properties and structures.The magnetic structures of Li2(Fe1-xCox)P2O7 and Na2(Fe1-xMnx)P2O7 solid solutions were explored in the course of this work. The crystal and magnetic structures of the silicates γ-Li2MnSiO4 and β-Li2CoSiO4 were investigated and their magnetic structures solved for the first time, including for chemically delithiated versions of Li2CoSiO4. Magnetic property measurements confirmed that Co had oxidised from Co2+ to Co3+, confirming that delithiation was successful while also serving to demonstrate the sensitivity of magnetic measurements to lithium content. In addition, the structural evolution of Li2CoSiO4 was tracked by in situ S-XRD and revealed no phase transformation during cycling. In summary, the outcome of this study is an extension of the state of knowledge of the low-temperature magnetic properties and structures of polyanion-based transition metal oxides, and a demonstration of the sensitivity of those properties and structures to electrochemical state. Further refinement of this approach could lead to a new tool for developing improved positive electrode materials for rechargeable batteries. The work also yielded crucial missing information concerning the electronic ground state of these materials, required for future high-level computational studies aimed at predicting properties including ionic conductivity.