X-ray Raman spectroscopy probes redox dynamics in lithium-ion battery cathodes

With the growing demand for electric vehicles, there is an urgent need for lower-cost, longer-lasting lithium-ion batteries that offer enhanced charge storage capabilities and stability over thousands of cycles. Nickel-rich layered cathodes based on LiNiO₂ are being developed to facilitate greater lithium intercalation, while reducing dependence on expensive and toxic cobalt. However, the specific roles of different redox centres in LiNiO₂ and their connection to degradation processes remain unclear.

Recent studies have suggested that molecular O₂ may play a role in the bulk redox process at highly delithiated states [1,2], based on signatures detected in O K-edge Resonant Inelastic X-ray Scattering (RIXS) during high-voltage charging. However, soft X-ray techniques like RIXS can only probe less than 10% of the few micrometre-sized cathode particles used in these studies, making the attribution of bulk molecular O₂ redox processes uncertain. Therefore, understanding oxygen redox in charge compensation in LiNiO₂ requires methods that can provide comparable information across depths ranging from tens of nanometres to several micrometres.

To address this, XRS was employed at beamline ID20 as a bulk-sensitive probe of both the O K- and Ni L_3,2-edges, enabling a direct comparison with more surface-sensitive soft X-ray absorption spectra (XAS). By tuning the incident energy to 10 keV and recording inelastically scattered photons with energy loss events near the O K- and Ni L-edge, XRS can probe shallow O 1s → 2p and Ni 2p → 3d transitions. This provides information comparable to soft XAS but with a significantly greater depth of around 10 μm.

This study combined various X-ray spectroscopies to obtain a depth-resolved (ranging from 10 nm to 10 μm) analysis of the redox processes in LiNiO₂, distinguishing reversible bulk redox activities from near-surface degradation. XRS indicates continuous changes in both the O K- and Ni L_3,2-edges throughout delithiation, confirming that charge compensation occurs via Ni-O rehybridization, with a corresponding decrease in electron density at oxygen sites as lithium is removed.

As shown in Figure 1a, the O K pre-edge, associated with transitions from O 1s → O 2p-Ni 3d hybridized states, intensifies and loses its asymmetry, reflecting a reduction in electron density. Importantly, no significant feature is observed at ∼531.5 eV, where one would expect the molecular O₂ signature detected by RIXS [1,2]. 

Click image to enlarge

Fig. 1: Depth-resolved core-loss spectra of LiNiO₂ cathode at different states of charge (SoC). a) and b) XRS results (∼10 μm information depth) for the O K-edge and Ni L₃-edge core-loss spectra of LiNiO₂ electrodes at varying SoC. Experimental XRS data are marked by black dots and represented by smooth solid lines. Charge transfer multiplet calculations for formally Ni²⁺ (green), Ni³⁺ (purple), and Ni⁴⁺ (pink) environments are included. c) and d) FY-XAS results (∼200 nm information depth) for the O K-edge and Ni L₃-edge.

Focusing on the Ni L-edges (Figure 1b), XRS reveals that in pristine LiNiO₂ materials, three nickel environments coexist: Ni²⁺, Ni³⁺, and Ni⁴⁺ [3]. During lithium removal upon charging, Ni²⁺ first converts to Ni⁴⁺, followed by the conversion of Ni³⁺. By the time the voltage reaches 4.8 V, the XRS signal is mainly dominated by Ni⁴⁺ species. The continuous increase in Ni⁴⁺ and the corresponding changes in the O K-edge confirm that charge compensation involves rehybridization between nickel and oxygen centres; however, there is no evidence that molecular O₂ plays a significant role in the observed capacity.

In contrast, the near-surface region of LiNiO₂, probed using Fluorescence Yield (FY)-XAS (with an information depth of about 200 nm), exhibits features of trapped molecular O₂ in the O K-edge when charged to potentials above 4.1 V (Figure 1c). This is accompanied by an increased presence of Ni²⁺ in the Ni L₃-edge (Figure 1d). As the molecular O₂ signature and Ni²⁺ features grow in FY-XAS, the intensity of the O K pre-edge peak and the Ni⁴⁺ component are suppressed compared to the XRS results. This suggests that molecular O₂ formation occurs near the surfaces of LiNiO₂ and is associated with nickel reduction, leading to the development of a denser NiO-like reduced surface layer that hinders lithium transport.

Notably, bulk-sensitive XRS (Figure 1a, 1b) shows no indication of significant O₂ formation or nickel reduction, even after charging to 4.8 V. This highlights the critical connection between the nickel reduction to Ni²⁺ near the cathode surface and the formation of molecular lattice O₂ in that region, as supported by RIXS data [1,2].

Similar trends are observed with more surface-sensitive Total Electron Yield (TEY)-XAS measurements (with ∼10 nm information depth), and scanning transmission electron microscopy of cross-sections of electrodes confirms that the reduced surface layer forms only at the regions in contact with the battery electrolyte. This highlights that molecular O₂ formation in LiNiO₂ is linked to near-surface degradation, rather than to reversible redox behaviour.

In conclusion, this work emphasizes the importance of integrating bulk- and surface-sensitive techniques to distinguish between bulk redox and surface degradation processes. The results indicate that redox activity in the bulk of LiNiO₂ cathode particles occurs via rehybridization between nickel and oxygen centres across the entire potential range examined (3.0 – 4.8 V).

Above 4.1 V, molecular O₂ forms in the near-surface region of LiNiO₂ particles in contact with the electrolyte, but this process does not contribute to the reversible capacity. Instead, the emergence of molecular O₂ is associated with nickel reduction, which creates a reduced surface layer that obstructs lithium transport.

Overall, this research provides new insights into the surface instability of LiNiO₂ at high voltages, highlighting the need for strategies such as cathode coatings, composition gradients, and electrolyte formulations to stabilize nickel-rich cathode surfaces.

XRS, as a hard X-ray technique, can measure shallow absorption edges with bulk sensitivity and minimal sample environment restrictions. It is a powerful tool for revealing the bulk redox behaviour of electrode materials for next-generation batteries.

Principal publication and authors
Distinguishing bulk redox from near-surface degradation in lithium nickel oxide cathodes, L. An (a), J.E. N. Swallow (a), P. Cong (a), R. Zhang (a), A.D. Poletayev (a,b), E. Björklund (a,b), P.N. Didwal (a,b), M.W. Fraser (a,b), L.A.H. Jones (a), C.M.E. Phelan (a), N. Ramesh (a), G. Harris (c), C. J. Sahle (d), P. Ferrer (e), D. C. Grinter (e), P. Bencok (e), S. Hayama (e), M. Saiful Islam (a,b), R. House (a),(b), P.D. Nellist (a,b), R J. Green (c,f), R.J. Nicholls (a), R.S. Weatherup (a,b,e), Energy Environ. Sci. (2024); https://doi.org/10.1039/D4EE02398F
(a) Department of Materials, University of Oxford, Oxford (UK)
(b) The Faraday Institution, Harwell Science and Innovation Campus, Didcot (UK)
(c) Department of Physics and Engineering Physics, University of Saskatchewan, Saskatoon (Canada)
(d) ESRF
(e) Diamond Light Source, Harwell Science and Innovation Campus, Didcot (UK)
(f) Stewart Blusson Quantum Matter Institute, University of British Columbia, Vancouver (Canada)

References
[1] N. Li et al., ACS Energy Lett. 4(12), 2836-2842 (2019).
[2] A.S. Menon et al., PRX Energy 2(1), 13005 (2023).
[3] A.D. Poletayev et al., arXiv:2211.09047 [Online] (2024).