In the field of energy storage and Li-ion batteries, searching for new (positive) electrode materials with better electrochemical performances than those of transition-metal oxides is of permanent concern. To that aim, very simple concepts of chemical bonding can be used to find out the origin of the electrode limitations and to guide experimentalists for the design of new promising materials. This local approach was recently applied to hybrid architectures, such as metal−organic frameworks (MOFs), and allowed some of us to demonstrate the first reversible lithium insertion into the MIL53(Fe) positive electrode. In this paper, we combine first-principles density functional calculations and local chemical bond analyses to fully interpret the redox mechanism of this material. Its reactivity versus elemental lithium is investigated as a function of (i) the lithium composition from xLi/Fe = 0−1, (ii) the lithium distribution over the most probable Li sites, and (iii) the OH/F substitution ratio along the redox chains. The results show that the MIL53(Fe) is a weak antiferromagnet at T = 0 K with iron ions in the high-spin state (Fe3+, S = 5/2). It reacts with lithium through a two-step insertion/conversion mechanism. The insertion reaction is perfectly reversible and proceeds in two steps: first, a single-phase reaction whose capacity increases linearly with the fluorine content in the starting material, then a two-phase reaction that ends around xLi/Fe = 0.5 due to the stabilization of a localized Fe2+/Fe3+ mixed-valence state along the inorganic chains. Further lithium insertion into Li0.5MIL53(Fe) is shown to provoke an irreversible conversion reaction due to a complete loss of the local interactions between the inorganic and organic networks of the MOF architecture. On the basis of this interpretation, several alternatives to improve the capacity of these materials can be proposed by means of appropriate ligand functionalization and/or use of electrochemically active molecules within the large open space occurring in such porous materials.

The paper presents a fast and efficient method to compute phase diagrams using first-principles methods and statistical physics coupled with experimental techniques. The method is applied to lithium intercalated graphite for which evidence of experimental anomalies have recently questioned the currently accepted stage II/stage I two-phase transformation in the 0.5 ≤ x ≤ 1.0 composition range. Our coupled experimental and theoretical analysis has led to a full determination of the Li x C6 phase diagram in this range of composition, revealing previously unknown phase transitions and the characterization of at least two new phases.