(Invited) The P2 Sodium Layered Oxides in Na Ion-Batteries

The development of very large scale renewable energy systems requires optimizing the lifetime, the price and the material availability. From these points of view, sodium based batteries have to be investigated. Our research group studied layered oxides as positive electrode for 30 years. Now a special focus is devoted on P2-type layered phases. One of the main interests of this structure is the existence of an ion conduction plane made of face sharing trigonal prims which exhibits a high ionic diffusivity thanks to the existence of a large bottleneck (oxygen rectangle) for sodium diffusion. This structure is able to accommodate a lot of transition metal cations, allowing the optimization of the properties by cationic substitution. Studies was performed on the P2- Na x (Mn , Co)O 2 , Na x (Mn , Fe)O 2, Na x (Mn , Fe,Co)O 2 and Na x (Mn , Fe,Ni)O 2 systems with various amounts of transition metal ions. All materials crystallize in the hexagonal system (P6 3 /mmc space group). The structure is made of MO 2 slabs built of corner sharing MO 6 octahedra. The sodium ions are in trigonal prismatic sites. Half of them share edges (Na e ) with the MO 6 octahedra while the other one share faces (Na f ). As the two types of prisms share a common face, they cannot be occupied simultaneously at the atomic scale. The sodium distribution depends of the cationic charge distribution and of the sodium amount is order to minimize the Na + -Na + and Na + -M n+ electrostatic repulsions. This lead to a large number of ordered Na + distribution as it was shown in the case of the Na x CoO 2 systems. In materials with several transition metal cations, their statically distribution in the slabs prevent generally from Na + ordering. In all systems, the P2 type packing is preserved in the 0.3 < x < 1 range. The fully intercalated phase is difficult to be obtained due to the very low ionic conductivity when the amount of vacancies is very small. For these compositions, only the Na f prisms are occupied. The specific capacity is in the 130-170 mAh.g -1 depending on the voltage range. For large deintercalation amounts (x < 0.3) a large polarization occurs in discharge. The GITT experiment shows that this “polarization” is not kinetics. The in-situ XRD during battery cycling from the Na 0.7 (Mn 0.5 Fe 0.3 Ni 0.2 )O 2 phase shows the existence at high voltage of two phases with different interslab distances. The change in the shape of the diffraction line profiles is characteristic of the continuous formation of stacking faults. For manganese based materials, structural distortions (Jahn-Teller effect) occurs at the end of discharge. Even if these distortions are fully reversible, they have a very bad effect on the cycling behavior. In the intermediate composition range the reversibility is very good. Therefore, these materials can be interesting for practical applications in stationary batteries due to their low price and the availability of the raw materials.