(Invited) Development of Layered NaMeO2 Materials for Na-Ion Batteries

Abstract

We first reported hard carbon//O3-NaNi1/2Mn1/2O2 full cell in 2011 [1] and have studied various layered oxides for Na-ion batteries for over a decade. [2, 3] Table 1 shows the variety of directly crystallizable oxides of A x MeO2 (A = Li, Na, and K) on 3d transition metal species. [3] Clearly, Na x MeO2 crystalizes with a wider variety of 3d transition metals without Na/Me cation-mixing unlike Li/Me due to the proper difference in ionic radii between Na+ and Me3+ ions. [3] The wide variety of 3d metals and crystal structures such as O3, P3, and P2 types including the distorted ones are found only in the Na system, unlike Li and K cases. Such structural variety gives us so wide selection of Na x MeO2 materials, which is highly attractive to develop high-performance positive electrodes. This talk will provide our strategies and insights into materials development based on our recent work on designing polymorphs and metal substitutions for Na x MeO2.The correlation between crystal polymorphism and electrochemical properties was investigated in detail by studying a series of Na-Mn-O layered oxides, including distorted and undistorted P2-type Na2/3MnO2, [4, 5] α-NaMnO2, [6] and β-NaMnO2. [7] These results show strong correlations between the electrochemical performance and crystal structure and provide new insight into utilizing the MnIII/IV redox couple.To maximize the energy density and cycle performance of the layered oxide materials, we studied O3-type Na2/3+x [Ni1/3Mn2/3 − x − y Fe x Ti y ]O2 (x = 1/6–1/3, y = 0–1/2), which can be regarded as solid solutions of Na2/3[Ni1/3Mn2/3]O2, Na2/3[Ni1/3Ti2/3]O2, and Na[Ni1/3Mn1/3Fe1/3]O2, as shown in the phase diagram of Figure 1. We synthesized 27 samples, which have O3-type structures in different compositions. Importantly, their electrode performance is significantly affected by the ratio of Na, Ni, Mn, Fe, and Ti in the materials, and we finally found the optimal material in the series of solid solutions is O3-type Na5/6[Ni1/3Mn1/6Fe1/6Ti1/3]O2. [2]Based on the experimental data of more than 100 samples synthesized over the past decade, we developed machine-learning models to predict reversible capacity, operating potential, and cyclability using composition descriptors. Moreover, we experimentally validated the optimal compositions predicted by machine learning and demonstrated new promising O3-type materials.[1] S. Komaba, W. Murata et al., Adv. Funct. Mater., 21, 3859 (2011).[2] S. Komaba, Chem. Lett., 49, 1507–1516 (2020).[3] K. Kubota, S. Komaba et al., Adv. Energy Mater., 8, 1703415 (2018).[4] S. Kumakura, S. Komaba, Angew. Chem. Int. Ed., 128, 12952–12955 (2016).[5] E. J. Kim, S. Komaba et al., ACS Appl. Energy Mater., 5, 12999–13010 (2022).[6] K. Kubota, S. Komaba et al., J. Mater. Chem. A, 9, 26810–26819 (2021).[7] M. Shishkin, S. Komaba, H. Sato et al., Chem. Mater., 30, 1257 (2018). Figure 1