Ordered and Disordered Polymorphs of Na(Ni2/3Sb1/3)O2: Honeycomb-Ordered Cathodes for Na-Ion Batteries

Jeffrey Ma,Shou-Hang Bo,Lijun Wu,Yimei Zhu,Clare P. Grey,Peter G. Khalifah
DOI: https://doi.org/10.1021/cm504339y
IF: 10.508
2015-01-01
Chemistry of Materials
Abstract:Na-ion batteries are appealing alternatives to Li-ion battery systems for large-scale energy storage applications in which elemental cost and abundance are important. Although it is difficult to find Na-ion batteries which achieve substantial specific capacities at voltages above 3 V (vs Na+/Na), the honeycomb-layered compound Na(Ni2/3Sb1/3)O-2 can deliver up to 130 mAh/g of capacity at voltages above 3 V with this capacity concentrated in plateaus at 3.27 and 3.64 V. Comprehensive crystallographic studies have been carried out in order to understand the role of disorder in this system which can be prepared in both disordered and ordered forms, depending on the synthesis conditions. The average structure of Na(Ni2/3Sb1/3)O-2 is always found to adopt an O3-type stacking sequence, though different structures for the disordered (R (3) over barm, #166, a = b = 3.06253(3) angstrom and c = 16.05192(7) angstrom) and ordered variants (C2/m, #12, a = 5.30458(1) angstrom, b = 9.18432(1) angstrom, c = 5.62742(1) angstrom and beta = 108.2797(2)degrees) are demonstrated through the combined Rietveld refinement of synchrotron X-ray and time-of-flight neutron powder diffraction data. However, pair distribution function studies find that the local structure of disordered Na(Ni2/3Sb1/3)O-2 is more correctly described using the honeycomb-ordered structural model, and solid-state NMR studies confirm that the well-developed honeycomb ordering of Ni and Sb cations within the transition-metal layers is indistinguishable from that of the ordered phase. The disorder is instead found to mainly occur perpendicular to the honeycomb layers with an observed coherence length of not much more than 1 nm seen in electron diffraction studies. When the Na environment is probed through Na-23 solid-state NMR, no evidence is found for prismatic Na environments, and a bulk diffraction analysis finds no evidence of conventional stacking faults. The lack of long-range coherence is instead attributed to disorder among the three possible choices for distributing Ni and Sb cations into a honeycomb lattice in each transition-metal layer, disrupting the Li2MnO3-type stacking of the honeycomb layers preferred in the ordered form of Na(Ni2/3Sb1/3)O-2. It is observed that the full theoretical discharge capacity expected for a Ni3+/(2+) redox couple (133 mAh/g) can be achieved for the ordered variant but not for the disordered variant (similar to 110 mAh/g). The first 3.27 V plateau during charging is found to be associated with a two-phase O3-P3 structural transition, with the P3 stacking sequence persisting throughout all further stages of desodiation.
What problem does this paper attempt to address?