A new breakthrough in the cathode material of cobalt-free lithium-ion batteries!
Brief introduction of the results
Lithium-ion batteries play an important role in decarbonizing transportation and power grids, but their reliance on cobalt, which is expensive and scarce on the planet, raises chain and sustainability issues. Although there have been multiple attempts to address this challenge, the removal of cobalt from Li(Nimnco)O2 remains difficult because it negatively affects its delamination and cycling stability.
Based on this,Brookhaven National Laboratory, Jianming Bai, and Argonne National Laboratory Feng Wang (corresponding author), et alA synthesis of a lithium-deficient composite structure containing layered and halite phases lini095mn0.The 05O2 method has better performance than the traditional layered structure lini095mn0.05o2。Through multi-scale correlation experimental characterization and computational modeling of the calcination process, the researchers revealed the role of lithium deficiency in inhibiting the halite-to-layered phase transition and crystal growth, resulting in small-sized composites with the desired low anisotropy lattice expansion and contraction. Therefore, the lack of lithium lini095mn0.05O2 exhibits a first-cycle coulombic efficiency of 90%, a capacity retention rate of 90%, and near-zero voltage decay over 100 deep cycles, demonstrating its potential as a cobalt-free cathode for sustainable lithium-ion batteries. This research provides a new idea for the development of high-performance battery materials that replace cobalt.
Background:
Due to its high storage capacity and low cost, high-nickel CAM is the material of choice for lithium-ion batteries (LIBs), especially electric vehicles. However, these high-nickel CAMs face key issues such as surface reconstruction, oxygen release, transition metal (TM) dissolution, bulk fatigue, and cracking, among others, which hinder their practical application. Over the past decade, significant efforts have been made to alleviate these issues, mainly focusing on lini1-x-ymnxcoyo2 with a hierarchical structure (where 1-x-y 0.).8）。During the calcination process, cobalt plays a key role by promoting the orderly arrangement of Li Ni, which is crucial for the structural order and cyclic stability of NMC. However, the increasing cost, environmental pollution from cobalt mining, and shortages require the removal of cobalt from LIBs, suggesting that the search for cobalt-free or cobalt-low alternative materials is important for the development of sustainable lithium-ion batteries.
**Reading guide
Figure 1In the synthesis of nickel manganese-based CAMS, the stoichiometry of lithium is used as a regulatory tool for structure and morphological control
As shown in Figure 1a, nickel-manganese-based CO-free cathodes (Lini1-XMNXO2) are favored for their high capacity, high thermal stability, and safety. However, due to the need to introduce Ni2+ to maintain charge neutrality, they suffer from lithium-nickel misalignment and cycle instability. Accumulating evidence suggests that lithium-nickel mixing is inevitable in nickel-manganese-based CAMS, and that the higher the manganese content, the more severe the lithium-nickel mixing is due to the inter-plane superexchange between magnetic Ni2+, Ni3+, and Mn4+ ions and the increased in-plane magnetic frustration.
Therefore, despite many efforts made in the past decade, capacity decay and voltage decay due to structural disorders have been the main challenges faced by cobalt-free layered cathodes in practical applications. High-nickel NMC CAMs are usually synthesized by calcination of hydroxides in an oxygen atmosphere with a Li source at high temperatures. The calcination process involves a multiphase transition and is strongly dependent on the composition of Li and TM. It is now known that lithiation, with the simultaneous binding of Li and O into the crystal lattice, is critical for phase transitions and crystallization processes. However, it is not clear how calcination conditions can result in large variations in particle size, ranging from tens of nanometers to hundreds of nanometers (as shown in Figure 1b). In the synthesis of high-nickel CAMs, an excess of a LI source (e.g., lithium hydroxide, LIOH) is added to the precursor to compensate for the loss of LI, ensuring the formation of a stoichiometric layered structure of LI in the calcination product of CAMS. However, the addition of additional lithium results in the formation of large-sized cuboid-shaped particles that can reach several hundred nanometers in diameter (Figure 1b; right), which contrasts sharply with rod-shaped particles of tens of nanometers in lithium-deficient CAMs (Figure 1b; Left). On the other hand, the extra lithium forms a lithium residue on the surface of the particles, which can cause surface instability during cycling.
Figure 2Lithium deficiency (095 li) and a lithium surplus (1Structure and electrochemical properties of nm9505 at 05 li).
As shown in the enlarged image of Figure 2a, a two-phase model was used for NM9505-0The XRD spectra of 95LIs were fitted. The calculated curves of 14° and 18° were directly compared with the experimental curves, 14The peaks at 5° belong to the layered phase (110), 152 ° and 17The peaks at 5° belong to the RS phase and the layered phase. In the CAMS synthesis process, Li-containing RS is formed as an intermediate phase and coexists with the layered phase for a long time during the slow two-phase transition. When LI is absent, a small amount of RS is retained. The local distribution of the two phases (layered phase and RS phase) and their symbiotic growth in single secondary and primary particles were further revealed by the combination of three-dimensional (3D) transmission X-ray microscopy (TXM) X-ray absorption near-edge structure (Xanes) spectroscopy and high-angle annular darkfield (HAADF) scanning transmission electron microscopy (STEM).
from NM9505-0A three-dimensional reconstruction of the TXM-Xanes of a secondary particle of 95Li (Fig. 2B) shows that Ni2+ is uniformly distributed throughout the particle without segregation, indicating a high degree of mixing of the layered and RS phases. Atomic Haadf-stem images further reveal the local distribution of lamellar and RS phases within a single primary particle, as shown in Figure 2. In the Z-contrast Haadf-Stem image (the intensity is about the same as Z1.)7 is proportional, where z is the atomic number), and the li and o atoms are not visible due to their low atomic number. However, the layered structure can be clearly seen, and the contrast between strength and weakness is formed under the action of TMS (Ni, Mn), indicating that RS and the layered structure are symbiotic (overlapping along the direction of the beam). As with the simulated atomic arrangement (below), the strongly contrasting layer represented by the green arrows is the overlap of the TM layer (red spheres) in the layered structure with the LITM (orange) in RS, while the weak contrast layer represented by the orange arrows is the overlap of the LI layer (green) in the layered structure with the LITM (orange) in RS. From this area, the spots represented by the black arrows represent the lamellar structure and the contribution of the RS, and are therefore stronger than those indicated by the red arrows, which belong only to the lamellar structure.
By combining volume XRD studies with local TXM and STEM analyses at different length scales, the authors were able to confirm that there was a significant increase in the number of x.r.sNanocomposites are formed in NM9505 of 95Li, consisting of the primary layered phase and the minor RS structure phase growing with each other within the same CCP oxygen framework and with an epitaxial orientation relationship (as shown in Figure 2D). For the sake of structure-performance relevance, the authors are described in 27-4.The NM9505 was measured at 0 in the 4 V voltage range95li and 1Electrochemical performance at 05LI (Figure 2E). Both show an overall similar voltage distribution. nm9505-0.The 95LI has a high first-cycle coulombic efficiency (CE 90%) and high discharge capacity (226 mAh g), which is much higher than the NM9505-105li (85% and 218 mAh g, respectively). DQ DV curve** is the charge discharge curve. Where, 4The transition from H2 to H3 at 2 V is thought to be one of the main causes of the instability of the high nickel cathode cycle, which can often be explained by the structural collapse and cracking of the cathode particles caused by large volume changes.
with NM9505-105Li, compared to NM9505-095Li exhibits weak and broad redox peaks, especially those associated with the H2-H3 transition. This is indicated with NM9505-105Li, compared to NM9505-0The phase change of 95Li is smooth, which is due to the reduced anisotropic structural change of the composite structure. As expected, at 05 down, NM9505-095LI maintains a high volume retention rate of 90% after 100 cycles, significantly better than NM9505-105li (75%;Figure 2f). After 100 cycles, when the magnification drops to 0At 1 C, most of the capacity is restored and the voltage is slightly attenuated (< 005 v;Figure 2G), again showing the composite structure at NM9505-0Structural robustness at 95li. In comparison, NM9505-105li observed rapid capacity decay (50 mAh g) and voltage decay (0.) over 100 cycles1 v)。
Figure 3The structure and morphology of NM9505 have a lot to do with the stoichiometry of lithium
By taking a lithium deficiency (090 and 095Li), close to stoichiometric (10 and 1025Li) and excess lithium (105 and 110Li) to systematically study the effects of lithium stoichiometry on structure, particle morphology and electrochemical properties. The resulting NM9505 powders have significantly different primary particle morphologies at different LI stoichiometry, although they have similar secondary particle morphologies. This difference is clearly shown in the SEM image (Figure 3A). The particle size distribution statistics in Figure 3b show a quantitative comparison, measuring the smallest size of more than 100 primary particles. The difference between them is evident because the lithium-deficient NM9505 particles are small, less than 100 nm (narrowly distributed, indicated by shaded areas), while the NM9505 particles that are close to stoichiometric and lithium excess are large (up to several hundred nm, widely distributed). By synchrotron XRD analysis, the LI content was investigated from 090li increased to 1The change of the shape of the diffraction peak at 10 Li, and the effect of Li stoichiometry on the structural order of the NM9505 series was studied (Fig. 3C). For example, as the LiTM ratio increases, all the diffraction peaks become clearer overall, while the distance between the (110) and (108) peaks associated with the layer increases as the LiTm ratio increases, reaching 100, which indicates the formation of an orderly lamellar knot. When li is greater than 1At 025, the peak shift of (104) was not obvious, indicating that Ni2+ generation was the least when Li was sufficient. The layered phase can be well refined105 and 110 L tissue of NM9505.
And for the LI content is less than 1In the sample of 025, the two phases coexisted, and the RS phase fraction decreased sharply as the Li content increased, and when the Li content reached 1At 025, the RS phase fraction decreased from 23% to nearly 3% (Figure 3D). The results showed that the RS content was strongly dependent on the LI stoichiometry. Due to the presence of 5% mn-induced Ni2+, the ratio of LiTm in the structure is less than 1, so that LiNi mixing occurs even in the case of an excess of Li-source. The two-dimensional TXM mapping of Xanes and Ni K-Edge further confirmed that when the LI content exceeded 1At 00, the ni valence state does not change much (Fig. 2). In addition, as shown in Figure 3d, the domain size of the layered phase increases with increasing lithium content, highlighting the critical role of lithiation in driving crystal growth. For comparison, the authors measured the electrochemical performance of NM9505 CAMS with different lithium content (Figure 3E). In the range of lithium deficiency and near stoichiometry, the first-cycle discharge capacity increases with increasing lithium content, and they all have a consistently high CE of nearly 90%. This may be due to the rapid intercalation dynamics of lithium in small primary particles44. In contrast, the first-cycle discharge capacity and CE are significantly reduced when the lithium content is too high due to the formation of a resistive surface layer of residual lithium. Overall, NM9505 exhibits high capacity retention in lithium deficiency and degraded cycling performance when lithium is in excess.
Figure 4NM9505 Effect of stoichiometry of lithium ions on phase transformation and crystallization during calcination
nm9505-0.95LI and NM9505-105LI underwent an overall similar physicochemical process, including (a) removal of water from TM hydroxides, (b) oxidation and lithiation of hydroxide precursors by reacting with O and Li salts, and (c) mass transport (including all substances such as TM, Li, and O) resulting in particle sintering. In both cases, lithium-bearing RS forms in the middle and then gradually transforms into a layered phase. When the temperature reaches 600°C, the RS content remains at a high level, and the layered domains are still small, indicating that the crystallization kinetics at low temperature is slow. As the temperature increases further, the RS ratio decreases rapidly, transforming into a layered phase due to heat-driven NI oxidation and LI incorporation (Figure 4E, insert). As a result, the lamellar region grows abruptly (Figure 4f, insert). At high temperatures, the kinetics of structural ordering and crystal growth are strongly dependent on the stoichiometry of lithium, as shown in Figures 4e and 4f. with NM9505-095Li, compared to NM9505-1The transition of Rs to the layered phase in 05Li is much faster, resulting in lower RS scores and larger layered domain sizes, which is consistent with the ex situ results (Figure 3D). 
Figure 5Mechanistic understanding of the stoichiometry of lithium regulates crystal growth and interparticle fusion during calcination
Combined with multi-scale experimental observations, the calcination process was computationally modeled to better understand the effects of lithium stoichiometry on phase propagation and crystallization, so as to understand the size and size distribution of primary particles in the final CAM. The primary particle microstructure of the NM9505 hydroxide precursor was generated by two-dimensional calculations, which was composed of spherical particles with an average particle size of about 350 nm (Figure 5A). The microstructure of the primary particles calcined by NM9505 at different temperatures and LiTM ratios was simulated computationally, and some examples are given in Figure 5b-d. The final results are summarized in Figure 5e and compared with experimental observations. It is evident that when the LiTM ratio is less than 1 (LiTm 1), the particle size growth rate is lower, while when LiTm is greater than the stoichiometric amount (Litm 11), the particle growth rate is significantly larger. The authors propose that the growth of primary particles is mediated by mass transport through two mechanisms, namely lithiation-induced crystallization and liquid-phase sintering. As shown in the vertical line at LiTM = 1 in Figure 5E, there are two distinct regions in which lithiation-induced crystallization is predominant and liquid-phase sintering plays a major role. This provides a plausible explanation for the experimental observations, i.e., the size mutation at Litm = 1. Figure 5f and Figure 5g simplify the particle growth mechanism dominated by lithiation-induced crystallization and liquid-phase sintering. When the LI source is just at or below the stoichiometric value (LI TM 10 or 10), most of the LI is incorporated into the particles, where it participates in phase transition and crystal growth during calcination, forming a smaller-sized composite composed of Rs and layered phases (Fig. 5f). When additional lithium (Litm 1.) is added0), not all LiOH reacts with the cathode precursor during calcination, and additional lithium salts exist in a molten state around the primary particles (Figure 5g). The liquid-phase sintering mechanism is activated due to faster mass diffusion through this liquid phase. As a result, small primary particles tend to merge into larger primary particles, similar to the synthesis of single-crystal NMC cathode materials by the molten salt method, sometimes described as the Ostwald maturation process. As a result, the overall primary particle size eventually becomes larger under lithium overload conditions. Bibliographic informationchen, k., barai, p., kahvecioglu, o. et al. cobalt-free composite-structured cathodes with lithium-stoichiometry control for sustainable lithium-ion batteries. nat. commun. ,