Ultrahigh-nickel layered cathode with cycling stability for sustainable lithium-ion batteries

Layered oxide frameworks serve as high-energy cathodes in Li-ion batteries but suffer from anisotropic lattice strain and stress during Li insertion/extraction. In high-Ni oxides, the lowering of the transition metal (TM) Fermi level upon delithiation leads to overlap of TM 3d and O 2p bands, invoking anionic (oxygen) participation in charge compensation. Deep charge compensation can generate highly reactive oxygen species, which may evolve as O2 and react with electrolyte to produce CO2/CO, risking thermal runaway. Oxygen loss also creates vacancies that promote TM migration, coupled with strain accumulation, driving irreversible phase transitions (layered → spinel → rock-salt) and performance decay. Stabilizing lattice oxygen during cycling is thus crucial for high-capacity Ni-rich layered oxides. Prior strategies include bulk doping, surface coatings, and gradient/single-crystal architectures to mitigate strain and side reactions, yet a comprehensive mechanism at the local structural scale remains elusive. There is a need for dopants that simultaneously tune particle morphology and local atomic/electronic structure to suppress irreversible phase transitions and enhance stability. Ordered structures offer low thermodynamic energy and stability. Tellurium, with suitable ionic radius, lattice matching, and unoccupied 5d orbitals, is proposed to form a Ni–Te ordered structure that enhances oxygen stability and refines particle morphology. The study introduces an ultrahigh-Ni layered oxide LiNi0.94Co0.05Te0.01O2 (NC95T) featuring refined grains and an intralayer Te–Ni–Ni–Te ordered superstructure to mitigate lattice strain and stabilize oxygen, targeting enhanced capacity, cycling stability, and high energy density.

Enhancement of Ni-rich layered oxide stability has focused on: (1) bulk doping to improve structural robustness, (2) surface coatings to suppress surface reactions, and (3) gradient compositions and single crystals to improve stress–strain tolerance. High-valence dopants such as Ta5+ and W6+ have been shown to refine grains and produce radially distributed primary particles, mitigating lattice strain over long cycles. However, links between electronic structure, oxygen evolution, and the local mechanism of doping remain incompletely understood. Ordered superstructures can lower thermodynamic energy and improve stability. Prior works report ordered motifs (for example, Ni6-ring and Li@Mn6 superstructures) that stabilize layered oxides. Te-containing layered oxides and Te–O systems show compatible chemistry and structure with high-Ni oxides, suggesting Te as a viable dopant to induce ordering. The gap identified is a strategy that simultaneously engineers particle morphology and local TM–ligand electronic structure to suppress thermodynamically driven phase transformations and oxygen loss.

Synthesis: NC95T (LiNi0.94Co0.05Te0.01O2) was produced via coprecipitation of NiSO4·6H2O, CoSO4·7H2O, and H2TeO3 (2 M total metal salt solution) using 4 M NaOH as precipitant and 1 M NH3·H2O as chelating agent at pH 11.0, 55 °C, 900 rpm for 30 h plus 10 h aging. The Ni0.94Co0.05Te0.01(OH)2 precursor was washed, dried (110 °C, 12 h), mixed with LiOH·H2O (1:1.02), calcined under O2: 450 °C (4 h, 2 °C min−1), then 740 °C (12 h, 1 °C min−1). Reference LiNi0.95Co0.05O2 (NC95) was synthesized similarly. Electrode preparation and coin cells: Cathode slurry (active:Super P:PVDF = 8:1:1) in NMP coated on Al foil, dried (110 °C), punched (14 mm, ~3 mg cm−2). Coin cells assembled in Ar glovebox with Li metal anode, Celgard2320 separator, and 1.2 M LiPF6 in EC:EMC (3:7) electrolyte. Cycling on Neware BTS-4000 at 30 or 55 °C; 1 C = 200 mA g−1. Pouch cells—Li metal batteries: Cathode loading 25 mg cm−2 (94:3:3 active:Super P:PVDF), Li strip (20 µm) on 4 µm Cu as anode; 21 cathodes/22 anodes; electrolyte 1.2 M LiPF6 in EC:EMC (3:7) + 2% VC, total 10.6 g. Formation: 0.1 C to 3.8 V, then 0.3 C to 4.3 V; first cycle 2.5–4.45 V at 0.05 C; subsequent cycles 2.5–4.3 V, discharge at 0.2 C. Pouch cells—Li-ion batteries with Si–C anode: Si–C anode (1,600 mAh g−1) loading 4.09 mg cm−2 on 6 µm Cu; cathode 25 mg cm−2 on 10 µm Al; stack of 15 cathodes/16 anodes; electrolyte 1 M LiPF6 in DMC:EC:EMC (1:1:1) + 5% FEC + 1% VC, 10 g total. Aged 48 h; conditioning at 0.1 C to 3.8 V then 0.3 C to 4.3 V. Initial cycle 2.7–4.45 V at 0.05 C; cycling 2.8–4.2 V (vs Si) at 1 C. Characterization: XRD (Bruker D8-Advance, Cu Kα), neutron diffraction (China Advanced Research Reactor), Rietveld refinement (GSASII). TKD for grain morphology (Helios 5 FIB/SEM; EBSD Nordlys II). 3D atom probe tomography (CAMECA LEAP 5000 XR; laser 355 nm, 20 pJ, 40 K). HAADF-STEM and EELS (FEI Titan Themis Z with Gatan GIF 965); SEM cross-sections and EDS mapping. In situ XRD in Be-window cell (0.2 C, 2.7–4.6 V) and DEMS (0.05 C, 2.7–4.4 V). XAFS: Ni/Te edges at BSRF and SSRF; O K-edge XANES (total electron yield) and RIXS mapping (BL20U, 527–532 eV excitations). Computations: DFT (VASP, GGA-PBE + U) on Li36Ni36O72 and Li36Ni34Te2O72 supercells; PAW, 600 eV cutoff, 3×3×2 k-mesh, convergence 10−5 eV, forces <0.05 eV Å−1. Oxygen vacancy formation energies computed via ΔE(O2 release) using reported O2 chemical potential.

• A Te-doped ultrahigh-Ni layered oxide LiNi0.94Co0.05Te0.01O2 (NC95T) with a Te–Ni–Ni–Te intralayer ordered superstructure was synthesized; Te exists as Te6+ and is uniformly distributed. • Electrochemical performance: 239 mAh g−1 at 0.1 C (to 4.6 V); 94.5% capacity retention after 200 cycles at 0.5 C. At 55 °C, 87% retention vs 33% for NC95. With 4.4 V cutoff, ~99% retention after 100 cycles at 0.5 C. With Li7Ti3O12 anode, >83% retention after 1,000 cycles at 10 C. • Energy density: Li metal pouch cell reaches 545 Wh kg−1; Li-ion battery with Si–C anode achieves ~404 Wh kg−1, retaining 91.2% after 300 cycles at 1 C (vs 24% for NC95). • Kinetics/voltage behavior: Broadened H2–H3 dQ/dV peaks (width ~2.2× vs NC95) reduce stress concentration; NC95T shows negligible voltage decay and low average polarization (0.1585 V) after 200 cycles, while NC95 shows 0.2568 V decay and ~4× polarization increase. • Structural stability: In situ XRD shows reduced (003) 2θ peak deviation at identical delithiation (0.98° vs 1.31°, 25% reduction). Lattice strain fluctuations in H2–H3 transition are ~−0.05% for NC95T (~3× lower than NC95). Lattice mismatch ratio along c stays below zero, indicating weak tendency to rock-salt formation. • Defect/phase evolution: Post-cycling EELS and STEM reveal rock-salt reconstruction limited to ~2 nm for NC95T vs >30 nm for NC95; fewer defects and near-zero strain by GPA in NC95T; EXAFS shows preserved local order for NC95T vs increased disorder for NC95. • Oxygen/electronic structure: DEMS detects CO2 release for NC95 above 4.3 V but none for NC95T up to 4.6 V. Ni K-edge XANES shifts monotonically to higher energy for NC95T (no oxygen release signature). O K-edge XANES indicates reduced surface byproducts (Li2CO3 peak near 534 eV minimal) and altered t2g/eg occupancy consistent with suppressed parasitic reactions. RIXS at 4.6 V shows no strong reversible O-redox in NC95; NC95T maintains similar spectral features to 3.8 V and exhibits stronger d–d excitation (~2.4 eV), indicating increased TM–O ionicity and stabilized lattice oxygen. • DFT: Ni–Te ordering lowers unoccupied states near EF by ~0.08 eV, reduces O p-band centre during delithiation (especially at Li0), minimizes Ni–O band overlap, increases O vacancy formation energies, passivates oxygen activity (Bader charge discreteness), and reduces bandgap, implying higher electronic conductivity. • Morphology: Te6+ induces grain refinement improving stress accommodation without preferential radial orientation; combined with ordering, it mitigates lattice strain and structural degradation.

The research addresses the instability of lattice oxygen and structural degradation in ultrahigh-Ni cathodes by introducing a Te–Ni–Ni–Te ordered superstructure that modifies TM–ligand electronic structure while concurrently refining particle morphology. Experimentally, NC95T demonstrates high capacity and markedly improved cycling stability, minimal voltage decay, and superior high-temperature and high-rate performance compared to NC95. In situ and ex situ structural probes (XRD, STEM/EELS, XAFS/EXAFS) show suppressed lattice strain, reduced rock-salt reconstruction, and preserved local order, correlating microscopic stability with macroscopic performance metrics. Gas analysis (DEMS) indicates suppressed oxygen-related gas evolution in NC95T, supporting the passivation of lattice oxygen activity. Surface-sensitive O K-edge XANES and Ni L-edge trends imply reduced catalytic activity (less Ni4+ at the interface after charging via reductive coupling) and a more stable cathode–electrolyte interphase. DFT elucidates that Ni–Te ordering lowers the O p-band centre during delithiation, reduces Ni–O band overlap, raises oxygen vacancy formation energy, and increases TM–O ionicity—collectively delaying anion redox onset and preventing lattice oxygen loss and cooperative distortions. While grain refinement aids stress management and Li diffusion, the absence of preferential radial orientation suggests that the major stability gains arise from electronic/structural modulation by the ordered superstructure. These findings demonstrate that achieving ordered TM-layer motifs is an effective pathway to reconcile high energy density with cycling stability and safety in Ni-rich layered oxides.

An ultrahigh-Ni layered cathode, LiNi0.94Co0.05Te0.01O2, featuring a Te–Ni–Ni–Te ordered superstructure and refined grains, delivers 239 mAh g−1 at 0.1 C and 94.5% capacity retention after 200 cycles to 4.6 V. Pouch cells achieve energy densities of 545 Wh kg−1 (Li metal) and ~404 Wh kg−1 (Si–C anode), with 91.2% retention after 300 cycles at 1 C. The ordered superstructure tunes TM–ligand energy levels, lowers the O p-band centre, suppresses lattice oxygen loss, mitigates lattice strain at high potentials, and limits rock-salt reconstruction. This multi-effect design strategy demonstrates that local ordering within TM layers can deliver high performance without compromising sustainability and safety. Future work should clarify the detailed mechanisms of anion redox onset and reversibility, optimize ordering motifs and dopant concentrations, and generalize the approach to other high-Ni and Co-free compositions and scalable manufacturing.

The study indicates that detailed properties and behavior of highly reactive lattice oxygen require further investigation, as reversible bulk O redox signatures are weak below 4.4–4.6 V and complex above. The exact mechanistic pathway linking surface electronic changes (Ni valence evolution via reductive coupling) to interphase stability warrants deeper probing. Although grain refinement is achieved, the absence of preferential radial orientation suggests morphology’s contribution is limited; disentangling morphology versus ordering effects would benefit from controlled microstructural variants. Additionally, the origin of the superior cyclic stability, while supported by multiple probes and DFT, is not exhaustively resolved and merits further mechanistic studies.