Niobium-Based High Rate Electrodes for Lithium-Ion Batteries: From Materials Design to Electrochemical Origins
Rui Xia is a PhD student in the research group Inorganic Materials Science. Supervisor is prof.dr.ir. J.E. ten Elshof and co-supervisor is prof.dr.ir. M. Huijben from the Faculty of Science & Technology.
The Ph.D. project described in this thesis aimed to enhance the charging speed of lithium-ion batteries (LIBs) by structural design. Niobium-based materials are one of the new material systems for fast-charging LIBs. Thus, niobium-based materials were chosen as the objects of investigation. Structural design can go in different directions: tuning the crystallite size of the materials, changing the size of the lithiation channels, aligning the lithiation channels with defined direction and length, etc. Thus, in the project, the structural design by a different method was applied to manipulate the structure on different dimensions (micrometer range, nanometer range, and atomic level). By the end of this project, the fast-charging performance was linked to the electrochemical reaction rate (and to the ohmic resistance and diffusion process) by extending the Butler-Volmer equation to determine the origin of the fast-charging behavior of niobium-based materials. This thesis contains two parts, the structural design of niobium-based materials and the decoupling of different subprocesses that influence the intrinsic material properties of the electrode materials of LIBs and can explain their fast-charging behavior.
The first part contains three different sections. In the first section, addressed in Chapter 2, the grain size and particle size of the Nb18W16O93 electrodes were tuned by variation of calcination time. Tungsten bronze Nb18W16O93 powders with different sizes were successfully synthesized. The grain size of the Nb18W16O93 powders varied from 60 nm to 130 nm, and the particle size of the Nb18W16O93 powders varied from 0.5 μm to 1.5 μm. The electrochemical performance of the smaller Nb18W16O93 crystals was better than larger Nb18W16O93 crystals regarding both cycling stability and rate performance. STEM results showed ~2 nm wide grain boundaries formed by local reconstruction, which explains the different particle and grain sizes. These grain boundaries could be fast lithiation channels in the Nb18W16O93 crystal. Furthermore, by detailed investigation of the STEM images of the a-b plane, the Nb18W16O93 crystal structure was reconstructed from the structure in the database. The reconstructed Nb18W16O93 crystal has connected lithiation channels by a connection path in the a-b plane, further enhancing the lithiation diffusion process.
In the second section, addressed in Chapter 3, a new niobium-based material, nickel niobate (NiNb2O6), was explored for lithium-ion batteries since it has a one-direction lithiation channel that was thought to promote fast charging. Nickel niobate has a single type of one-dimensional channel for lithium-ion transport and storage, leading to a single voltage plateau at 1.65 V during charge-discharge cycling, which is different than the structure of Nb18W16O93 with three different types of channels. Thus, it is, in principle, easier for practical use. The oxidation of all three transition metal ions (Ni, Nb) enables the structure containing 3 lithium ions up to Li3NiNb2O6 with a capacity of about 244 mAh g−1 at 0.5 C. Nickel niobate shows a high diffusion coefficient of 10−12 cm2 s−1 (confirmed by sweep rate cyclic voltammetry, GITT, and EIS), which enables fast cycling at high current densities resulting in high capacities of 220, 165, 140, and 50 mAh g−1 for 1, 5, 10, and 100 C, respectively. Furthermore, the origin of the stable reversible lithiation process in NiNb2O6 is a minimal volume change upon lithiation/delithiation, which leads to a capacity retention of 81% after 20 000 cycles at 100 C. Finally, full cell systems against LiFePO4 and NCM811 cathodes show stable charge-discharge kinetics under 5 C, demonstrating the promising energy storage performance of nickel niobate anodes in practical battery devices.
In the third section, addressed in Chapter 4, niobium tungsten oxide (NbWO) nanorods were formed on (001), (110), and (111) oriented niobium doped strontium titanate (NbSTO) substrates with aligned lithiation channels. Epitaxial NbWO nanorods were successfully grown on top of (001), (110), and (111) oriented NbSTO substrates. The growth direction of the nanorods followed the symmetry of the underlying substrates. The NbWO nanorods on (001) and (110) oriented NbSTO substrates were in-plane oriented. However, the NbWO nanorods on (111)-oriented NbSTO substrate was 35.2 degrees out-of-plane growth-oriented, which makes them interesting for fast lithium-ion intercalation. The NbWO nanorods on (111)-oriented NbSTO substrate had the most practically well-aligned lithiation channels. Furthermore, the NbWO nanorods were revealed to have an unexpected atomic structure. A layer-by-layer crystal stack of 2D glass structures was found in niobium tungsten oxide nanorods. The proofs for an amorphous a-b plane and a crystalline c direction in the NbWO nanorods were found using RSM, STEM, etc., techniques. The atomic structures of differently sized NbWO nanorods were confirmed to be the same. Thus, to the best of our knowledge, it can be concluded that a new type of atomic arrangement besides crystalline, quasi-crystalline, and amorphous structure was found in the niobium tungsten oxide system. In this atomic arrangement, only one mirror symmetry plane exists, the plane perpendicular to the crystal direction. Also, NbWO nanorods show exceptionally fast lithium-ion intercalation ability. 14400 pulses grown NbWO nanorods on NbSTO111 substrate showed 38% capacity retention from 5 C to 1000 C. The capacity retention after 10000 cycles at 250 C is over 80%. Thus, this NbWO nanorod film has a huge potential to be used as fast charging electrode in lithium-ion micro-batteries.
The second part of the thesis contains two sections. In the first section, discussed in Chapter 5, an extended form of the Butler-Volmer equation was used to investigate the separate influences of reaction rate, lithium-ion diffusion, and ohmic resistance in the battery on its fast-charging property. The extended Butler-Volmer equation was rearranged into a version with an ohmic resistance factor and shape and size factor, which can simulate the relation between the current peak voltages and the peak currents from a series of sweep rate CV measurements. The model was applied to a series of well-known electrode materials as case studies to prove its validity of the model. The model was further extended to calculate the overpotential at non-current peak positions. The overpotential at non-current peak positions shows a broadening of the CV peak that can be related to slow lithium-ion diffusion. The new model described in chapter 5 can serve as a guide for electrode material modification as well as for identifying new suitable materials for high-rate batteries.
In the second section, Chapter 6, the model from Chapter 5 is used to decouple the influences of reaction rate and diffusion on the fast charging ability of niobium-based electrode materials. The reaction rate of in niobium-based electrodes was found to be larger than the reaction in TiO2, which indicates that the superior rate performance of niobium-based electrode materials is particularly related to the faster reaction of Nb5+/Nb4+. A conclusion that can be drawn from the work presented in this thesis is that further investigations of niobium-based materials may be more meaningful for the practical use of fast-charging electrodes than titanium-based materials. The gravimetric and volumetric energy density of different LIBs (full cell systems) indicates that high volumetric energy and power density LIBs require niobium-based electrodes.