Advanced Materials: IEST SPFT2000 Quantitatively Stress-Chemical Coupling Drives Multi-Step Phase Transition Degradation in Single-Crystal Ultra-High-Nickel Cathodes

1. Background: The High-Voltage Degradation Puzzle in Ni-Rich Cathodes

Ultra-high-Ni layered oxide cathodes (LiNixCoyMn1−x−yO₂, x ≥ 0.92) deliver exceptional specific capacity but suffer from accelerated capacity fade when operated at elevated cut-off voltages (≥4.3 V). The conventional understanding attributes this degradation to the formation of an electrochemically inactive rock-salt (Fm-3m) phase at particle surfaces and crack walls. However, the sequence of phase transformations leading from the pristine O3 layered structure to the final rock-salt phase — and the role of mechanical stress in driving this cascade — has remained poorly characterized.

Previous studies have faced two fundamental limitations: (1) polycrystalline samples introduce grain-boundary cracking as a confounding failure mode, obscuring the intrinsic lattice-level mechanisms; and (2) ex-situ characterization captures only the final degradation state, missing the transient intermediate phases that reveal the thermodynamic and kinetic pathway. The multi-step phase transition pathway in single-crystal NCM cathode materials has therefore remained incompletely defined — until this study.

2. Experimental Strategy: Isolating Intrinsic Degradation in Single-Crystal Ni-Rich SC-N92

The research team selected SC-N92 — a single-crystal Ni-rich (≥92% Ni) cathode material with 2–5 μm particle size — as the model system. This choice was strategic on two fronts: (1) the single-crystal morphology eliminates grain-boundary contributions, allowing the study to focus purely on intracrystalline degradation; and (2) the high Ni content maximizes the sensitivity to Ni³⁺/Ni⁴⁺ reduction instability and cation mixing at high voltage, making SC-N92 an ideal probe for the high-voltage cycling failure mechanism of single-crystal Ni-rich cathodes.

A multi-scale characterization suite was deployed:

• Electrochemical: Half-cells + pouch full cells at 4.3V, 4.4V, and 4.5V cut-off voltages; EIS, GITT, dQ/dV, rate capability, and long-cycle testing to quantify capacity fade, voltage decay, impedance growth, and D_Li⁺ degradation.
• Structural: FIB-SEM cross-sectioning + HAADF-STEM at atomic resolution + GPA strain mapping to visualize lattice bending, stacking faults, and cation mixing at crack regions.
• Mechanical: IEST SPFT2000 Single-Particle Mechanical Testing System to quantify particle-level critical fracture stress and establish the correlation between lattice distortion severity and mechanical strength.
• Dynamic: In-situ XRD to monitor c-axis lattice parameter evolution during charge/discharge.
• Computational: DFT first-principles calculations (DOS, charge density difference, migration barriers, COHP) and MD molecular dynamics simulations to quantify Li⁺ transport kinetics in each phase.
• Chemical: TOF-SIMS depth profiling, XPS etching, XANES/EXAFS, and LA-ICP-MS to trace electrolyte decomposition products, transition metal dissolution, and cross-contamination from cathode to anode.

Figure 1. Electrochemical data in half/fullcells for SC-N92 at 45°C. (a) Cycling performance in half cells, (b) Rate performance. (c) EIS Nyquist plots before cycling and after 200 cycles within the voltage range of 4.3-4.6V. (d) Cycling performance in pouch-type full cells, (e) Corresponding energy density, GITT diagrams during (f) charge and (g) discharge processes, (h-j) dQ/dV contour diagrams with a cut-off voltage window of 4.3-4.5V at 45°C.

3. The Complete Multi-Stage Phase Transition Chain

3.1 Atomic-Scale Visualization: Lattice Bending Initiates the Cascade

Figure 2. Morphology and crystal structure for cathode in pouch-type full cells at 45°C after 300 cycles. Continuous FIB-SEM of single particles with different sections for (a) SC-N92-4.3 V and (b) SC-N92-4.5 V. HAADF-STEM of (c) surface and (d) 200-nm deep positions for SC-N92-4.3 V, (e) surface and (f) 200-nm-deep positions for SC-N92-4.4 V, (g) surface and (h) 200-nm-deep positions for SC-N92-4.5 V with corresponding strain mapping using GPA patterns and enlarged lattice images.

Atomic-resolution HAADF-STEM imaging combined with GPA strain analysis revealed that lattice bending induced stress concentration is the primary trigger — not a secondary consequence — of phase degradation. Bending regions exhibit concentrated elastic strain that: (1) distorts the oxygen framework, facilitating transition-metal migration from octahedral to tetrahedral sites; (2) locally increases the c-axis lattice parameter, destabilizing the layered O3 structure; and (3) creates a positive-feedback loop where initial phase transformation further amplifies local stress, accelerating subsequent transformation steps.

3.2 The Five-Phase Degradation Chain

The O3→O1→LiNi₂O₄→Ni₃O₄→rock-salt degradation chain uncovered in this study represents the first complete mapping of the transformation pathway in ultra-high-Ni cathodes:

3.3 Ni₃O₄: The Missing Link and Multi-Phase Crack Region

A finding that fundamentally revises the accepted degradation model: the crack regions in cycled SC-N92 particles do not contain only rock-salt phase, as previously assumed. Instead, multi-phase coexistence at crack regions in degraded cathodes is the norm: LiNi₂O₄ (Fd-3m), Ni₃O₄ (Cmmm), and rock-salt (Fm-3m) are simultaneously present. EDS line scans across crack walls show a Ni/O ratio gradient consistent with progressive oxygen loss along the degradation chain, with Ni₃O₄ representing an intermediate stoichiometry between LiNi₂O₄ and NiO.

DFT and MD simulation for cathode phase transition kinetics provided the quantitative rationale: Ni₃O₄ exhibits a Li⁺ migration barrier that is approximately 3× higher than that of the O3 phase, effectively blocking the two-dimensional diffusion channels. The Cmmm orthorhombic structure lacks the continuous octahedral-vacancy network required for facile Li⁺ hopping, making even a thin Ni₃O₄ layer at crack walls a potent barrier to lithium transport.

Figure 3. XRD poles molecular dynamics simulation for cathode in pouch-type full cells at 45°C after 300 cycles. XRD poles depicting the (003), (101), and (006)/(012) lattice planes for (a) SC-N92-4.3 V and (b) SC-N92-4.5 V. Molecular dynamics simulation plots for the systems of (c) SC-N92-4.3 V, (d) SC-N92-4.4 V and (e) SC-N92-4.5 V.

4. The Stress-Chemical Coupling Feedback Loop

4.1 c-Axis Contraction, the H2-H3 Irreversible Transition, and COMSOL Stress Simulation

Figure 4. In-situ XRD: c-axis lattice contraction and anisotropic strain in Ni-rich cathodes. At 4.3V, contraction is limited to 3.91% with reversible H2-H3 transition. At 4.6V, contraction reaches 6.41%, triggering the H2-H3 irreversible phase transition in high-voltage NCM — the lattice cannot recover its original dimensions upon relithiation.

In-situ XRD monitoring during charge/discharge (Figure 4) quantified the c-axis lattice contraction and anisotropic strain in Ni-rich cathodes as a function of voltage. At a cut-off of 4.3V, the maximum c-axis contraction was limited to 3.91%, and the H2→H3 phase transition remained fully reversible — upon relithiation, the lattice recovered its original dimensions with no detectable residual strain. At 4.6V, contraction reached 6.41% — a 64% increase — and the H2-H3 irreversible phase transition in high-voltage NCM was triggered: the (003) diffraction peak showed permanent broadening and position shift after the first delithiation/lithiation cycle, indicating that the layered structure could not fully recover its original orientation and interlayer spacing.

This irreversibility is the structural root cause of the observed voltage and capacity fade: each high-voltage cycle introduces a small but permanent lattice disordering that accumulates over hundreds of cycles, progressively degrading the crystallographic template required for facile two-dimensional Li⁺ diffusion.

COMSOL Multiphysics Simulation provided the engineering-scale validation of these atomic-scale findings. The model coupled lithium concentration diffusion with mechanical stress generation, simulating the evolution of Li⁺ distribution and von Mises equivalent stress within an SC-N92 particle after extended cycling. At 4.3V, the simulation showed a relatively uniform Li⁺ concentration profile from the particle surface to the core, with moderate and evenly distributed von Mises stress. At 4.5V, a dramatically different picture emerged: a steep Li⁺ concentration gradient developed between the particle surface (near-fully delithiated) and the interior (partially lithiated). This gradient generated a correspondingly non-uniform stress field with stress concentrations at the phase-boundary regions where lattice parameters differed most sharply. The simulation directly demonstrates how the high-voltage phase transitions create self-amplifying mechanical stress that drives the progressive particle cracking observed in FIB-SEM — perfectly consistent with the stress-chemical coupling induced cathode particle fracture model.

4.2 DFT First-Principles and XANES/EXAFS: Electronic Origin of Kinetic Degradation

Figure 5. (a–d) DFT first-principles calculations: density of states (DOS), charge density difference, Li⁺ migration energy barriers, and COHP bond analysis across O3, O1, LiNi₂O₄, Ni₃O₄, and rock-salt phase interfaces. (e–g) XANES, EXAFS, and wavelet transform (WT) of the Ni K-edge for SC-N92 cycled at 4.3V and 4.6V.

DFT and MD simulation for cathode phase transition kinetics provided the electronic-structure explanation for the observed degradation. Across the five-phase chain, DFT calculations revealed a progressive deterioration of electronic and ionic transport properties:

• Density of States (DOS): The O3 phase exhibits a clear metallic character with substantial electron density at the Fermi level. As the phase transformation progresses to O1, LiNi₂O₄, and Ni₃O₄, the DOS near the Fermi level systematically decreases — Ni₃O₄ and rock-salt show a distinct band gap opening, indicating the transition from metallic conductor to electronic insulator.
• Charge Density Difference: Electron redistribution maps across phase boundaries show that at high-voltage O3/O1 and O1/LiNi₂O₄ interfaces, charge is substantially depleted from the oxygen framework and accumulates on transition-metal sites — weakening TM–O bonds and facilitating further TM migration into lithium sites (cation mixing).
• Li⁺ Migration Barriers: Nudged elastic band (NEB) calculations quantify the progressively increasing activation energy for Li⁺ hopping: O3 (~0.3 eV) → O1 (~0.5 eV) → LiNi₂O₄ (~0.7 eV) → Ni₃O₄ (~0.9 eV, ~3× the O3 barrier) → rock-salt (>1.2 eV, effectively blocked). This trend confirms Ni₃O₄ as the kinetic bottleneck predicted by the structural analysis.
• COHP Bond Analysis: Crystal orbital Hamilton population analysis across phase interfaces reveals that at 4.5V+, the excessively strong interfacial bonding at multi-phase boundaries (particularly LiNi₂O₄/Ni₃O₄ and Ni₃O₄/rock-salt interfaces) contracts the interlayer spacing, further elevating the Li⁺ migration barrier. This “over-bonding” effect at high voltage is the electronic-structure origin of the observed kinetic degradation.

XANES and EXAFS spectroscopy at the Ni K-edge (Figure 5e–g) provided the experimental validation. After cycling at 4.3V, the XANES white-line position and EXAFS Fourier transform confirm that a high proportion of electrochemically active Ni³⁺/Ni⁴⁺ is maintained, with Ni–O bond lengths close to the pristine values. In stark contrast, the 4.6V-cycled sample shows: (1) a pronounced shift of the absorption edge to lower energy — indicating substantial reduction of Ni³⁺/Ni⁴⁺ to inactive Ni²⁺; (2) EXAFS fitting reveals shortened Ni–O bond lengths in the first coordination shell, consistent with the formation of rock-salt-like local environments; and (3) wavelet transform analysis shows the disappearance of the characteristic second-shell Ni–TM scattering peak, indicating severe disruption of the layered ordering. These spectroscopic signatures directly corroborate the DFT prediction that high-voltage cycling drives irreversible electronic and structural degradation through the multi-stage phase transition cascade.

4.3 Chain Interfacial Failure: From Cathode CEI to Anode SEI

Figure 6. Characterisation of the morphology in pouch-type full cells at 45°C after 300 cycles. The TOF-SIMS concentration distribution and 3D rendering of the composition (NiF3, CoF3, and MnF2) of TM phase for (a) SC-N92-4.3 V and (b) SC-N92-4.5 V with (c) corresponding depth profile curve, the concentration distribution and 3D rendering of the composition (PF2O2, CF and CN) of CEI film for (d) SC-N92-4.3 V and (e) SC-N92-4.5 V with (f) corresponding depth profile curve. Depth XPS contour diagram of C 1s, O 1s, F 1s for (g) SC-N92-4.3 V and (h) SC-N92-4.5 V within the 0-300 nm range. A single-particle force test for (i) SC-N92-4.3 V and (j) SC-N92-4.5 V; (k) Curve graph for the single-particle force test. TEM images of the cathode for (l) SC-N92-4.3 V and (m) SC-N92-4.5 V. TEM images of the anode for (n) SC-N92-4.3 V and (o) SC-N92-4.5 V. LA-ICP-MS with Ni, Li, and P distributions for (p) SC-N92-4.3 V and (q) SC-N92-4.5 V.

The degradation initiated by lattice bending and phase transition does not remain confined to the cathode. TOF-SIMS depth profiling combined with XPS etching (Figure 6) provided a layer-by-layer chemical analysis from the SC-N92 particle surface into the bulk interior. Three key findings emerged:

• Electrolyte penetration depth: At 4.5V, electrolyte decomposition products — including LiF, LiₓPOᵧF₂, and organic carbonate oligomers — were detected at depths exceeding 200 nm from the particle surface, indicating that the crack network formed by phase-transition-driven mechanical failure provides continuous pathways for electrolyte infiltration deep into the particle interior. At 4.3V, these species were confined to the outermost ~50 nm.
• CEI thickness and uniformity: The 4.5V-cycled cathode developed a thick (≥100 nm), chemically heterogeneous CEI with alternating fluoride-rich and organic-rich layers — a signature of repeated cracking/exposure/passivation cycles. The 4.3V cathode formed a thin (~20 nm), uniform, and chemically stable CEI dominated by a single LiF-rich layer — consistent with controlled surface passivation that inhibits further side reactions.
• Transition metal dissolution: XPS quantification showed that the CEI on 4.5V cathodes contained substantial Ni, Mn, and Co — direct evidence of active-material dissolution into the electrolyte, driven by the acidic decomposition products (HF from LiPF₆ hydrolysis) concentrated in the crack-confined electrolyte volume.

LA-ICP-MS and cross-sectional TEM (Figure 6) traced the downstream consequences of this cathode degradation. Quantitative LA-ICP-MS analysis of the cycled graphite anode revealed significant Ni deposition — the dissolved transition metals migrate through the separator and electrochemically deposit on the anode surface. The corresponding anode TEM images showed a thick, irregular SEI on the 4.5V-paired anode, with Ni-rich inclusions that act as catalytic sites for further electrolyte decomposition. This creates the chain interfacial failure from cathode to anode in Li-ion batteries: cathode cracking → electrolyte penetration → TM dissolution → Ni deposition on anode → catalytic SEI growth → active lithium trapped in anode SEI → accelerated full-cell capacity fade. The bidirectional consumption of active lithium — at both the cathode CEI and the anode SEI — explains why the full-cell capacity fade rate at 4.5V is substantially faster than what half-cell data alone would predict.

4.4 Voltage Window: The Master Control Variable for SC-N92 Stability

Figure 7. Comprehensive degradation mechanism schematic. The full-chain chemo-mechanical failure loop: (1) lattice bending concentrates stress → (2) stress triggers multi-stage O3→O1→LiNi₂O₄→Ni₃O₄→rock-salt phase transition → (3) phase boundaries amplify internal stress → (4) stress exceeds particle fracture strength (IEST SPFT2000 quantified) → (5) intracrystalline cracks form → (6) electrolyte penetrates cracks → (7) thick, non-uniform CEI forms + TM dissolution → (8) dissolved TM deposits on anode → (9) catalytic SEI growth on anode → (10) bidirectional active-lithium loss accelerates full-cell failure.

Voltage window optimization for single-crystal NCM cathodes is not merely a matter of trading capacity for cycle life — it represents a fundamental binary choice between two degradation regimes:

• 4.3V “mild high-voltage” regime: All degradation drivers are partially but significantly suppressed. Lattice bending is limited to isolated regions; the c-axis contraction of 3.91% remains within the elastic recovery limit; the H2-H3 transition is fully reversible; phase degradation is confined to the early O3→O1 stage without progressing to spinel or rock-salt; Li⁺ diffusion remains uniform across the particle; internal stress is low and evenly distributed; and single-particle mechanical strength, measured by IEST SPFT2000, shows only moderate reduction.
• ≥4.5V “ultra-high-voltage” regime: The degradation cascade is fully activated. Lattice bending becomes pervasive; c-axis contraction reaches 6.41% — beyond the reversible limit; the H2-H3 transition becomes irreversible; the complete O3→O1→LiNi₂O₄→Ni₃O₄→rock-salt chain is triggered; ion channels are progressively blocked by Ni₃O₄ and rock-salt; a steep Li⁺ concentration gradient develops between surface and bulk; internal stress accumulates continuously and non-uniformly; and particle mechanical integrity drops precipitously — directly linked by IEST SPFT2000 data to the severity of lattice distortion.

Figure 7 provides the complete mechanistic schematic — the full-chain chemo-mechanical failure loop — synthesizing all characterization data into a single degradation model that connects lattice-scale phenomena (bending, phase transition) to particle-scale outcomes (cracking, strength loss) to cell-scale failure (CEI/SEI growth, TM crossover, capacity fade). This model fills the theoretical gap in understanding high-voltage cycling failure mechanism of single-crystal Ni-rich cathodes and provides atomic-scale guidance for practical engineering strategies including voltage window control, lattice-strain buffer doping, and conformal surface coating.

5. Practical Implications: Voltage Window Optimization and Materials Design

6. Conclusions

• Complete degradation chain established: For the first time, the full O3→O1→LiNi₂O₄→Ni₃O₄→rock-salt degradation chain is experimentally documented in single-crystal ultra-high-Ni cathodes. Lattice bending is identified as the primary trigger — not a secondary consequence — of phase degradation. While stress concentration in turn accelerates the irreversible phase transitions, forming a positive feedback loop that ultimately leads to the formation of through‑particle intragranular microcracks.
• Ni₃O₄ (Cmmm) identified as key intermediate: The Ni₃O₄ intermediate phase from spinel to rock-salt transformation is reported for the first time in cathode degradation. It lacks continuous Li⁺ transport channels and represents the kinetic bottleneck in capacity fade.
• Multi-phase crack regions: The study overturns the assumption that only rock-salt forms at cracks. Multi-phase coexistence — LiNi₂O₄ + Ni₃O₄ + rock-salt — is the actual degradation microstructure.
• Stress-chemical coupling loop: Lattice bending → stress concentration → accelerated phase transition → further lattice distortion → intracrystalline cracking → electrolyte penetration → chemical degradation → more cracking. This feedback loop explains why high-voltage degradation is autocatalytic.
• Chain interfacial failure: Cathode degradation propagates to the anode via transition-metal dissolution and cross-contamination, creating a full-cell failure mode beyond the cathode alone.
• IEST SPFT2000 validation: Single-particle mechanical testing provided the quantitative link between lattice distortion severity and mechanical strength loss — a dataset critical to establishing the stress-chemical coupling mechanism.

7. FAQs