The influence of crystal structure and relative stability on the electrochemical properties of Li2FeSiO4 polymorphs as cathode materials for Li-ion batteries is investigated. Six Li2FeSiO4 forms related to the crystal structure of Li3PO4 have been considered: three known polymorphs crystallizing in layered structures (with space groups P21, Pmnb, and Pmn21) and three forms reported for other LiMSiO4 materials crystallizing in three-dimensional (3D) frameworks (with space groups Pmn21, Pbn21, and P21/n). While Li2FeSiO4 polymorphs are very close in energy, their energies begin to differ substantially upon removal of Li, rendering those with lower electrostatic energy more stable. The change in relative stability of polymorphs upon delithiation results in a driving force for a phase transformation of the exiting two-dimensional (2D)-Li2FeSiO4 polymorphs to a more stable structure having a 3D network of [SiO4] and [FeO4] tetrahedra. The resulting phase exhibits a voltage plateau that is predicted to be 0.3 V below that of the original phase for the first electron process (Li2FeSiO4/LiFeSiO4 couple). The calculated voltage–capacity curves for the first and second cycles of Li2FeSiO4 at room temperature, assuming transformation to the new polymorph occurs after the first cycle, are in excellent agreement with experiments. Independently of the polymorphs, removal of the second lithium ion occurs at too high of a voltage (above 4.7 V) and is accompanied by major structural rearrangements, precluding the utilization of any of the unmodified Li2FeSiO4 polymorph derived from Li3PO4 as high specific capacity material.

The microstructural complexity of Li-rich cathode materials has so far hampered understanding the critical link between size, morphology and structural defects with both capacity and voltage fadings that this family of materials exhibits. Li2MnO3 is used here as a model material to extract reliable structure-property relationships that can be further exploited for the development of high-performing and long-lasting Li-rich oxides. A series of samples with microstructural variability have been prepared and thoroughly characterized using the FAULTS software, which allows quantification of planar defects and extraction of average crystallite sizes. Together with transmission electron microscopy (TEM) and density functional theory (DFT) results, the successful application of FAULTS analysis to Li2MnO3 has allowed rationalizing the synthesis conditions and identifying the individual impact of concurrent microstructural features on both voltage and capacity fadings, a necessary step for the development of high-capacity Li-ion cathode materials with enhanced cycle life.

Olivine NaFePO4 has recently attracted the attention of the scientific community as a promising cathode material for Na-ion batteries. In this work we combine density functional theory (DFT) calculations and high resolution synchrotron X-ray diffraction (HRXRD) experiments to study the phase stability of NaxFePO4 along the whole range of sodium compositions (0 ≤x≤ 1). DFT calculations reveal the existence of two intermediate structures governing the phase stability at x = 2/3 and x = 5/6. This is in contrast to isostructural LiFePO4, which is a broadly used cathode in Li-ion batteries. Na2/3FePO4 and Na5/6FePO4 ground states both align vacancies diagonally within the ab plane, coupled to a Fe(2+)/Fe(3+) alignment. HRXRD data for NaxFePO4 (2/3 < x < 1) materials show common superstructure reflections up to x = 5/6 within the studied compositions. The computed intercalation voltage profile shows a voltage difference of 0.16 V between NaFePO4 and Na2/3FePO4 in agreement with the voltage discontinuity observed experimentally during electrochemical insertion.

State-to-state exact quantum probabilities of the N + N2 exchange reaction have been calculated on the recently proposed L4 potential energy surface fitted to high level ab initio points using full-dimensional time-independent quantum techniques. Thermal rate coefficient values calculated on L4 were found not to differ from those calculated on a previously proposed potential energy surface. On the contrary, state-specific reaction probabilities calculated on the two surfaces are shown to differ significantly. Arguments for attributing the difference to specific features of the considered potential energy surfaces are provided.