Structural Degradation Studies of High Voltage Lithium Ion Battery Materials

High voltage Li-ion battery materials are being developed for high energy density applications, such as electric vehicles, and grid scale storage. However, high voltage cathode materials, such as LiCoPO4, typically undergo severe capacity loss during cycling, preventing commercialisation.

Degradation within high voltage cathode electrodes can occur due to parasitic surface reactions with the electrolyte at the electrode/ electrolyte interface, and via structural degradation within individual cathode particles. Parasitic reactions at the electrode/ electrolyte interface can result from a lack of passivation by a cathode electrolyte interphase (CEI) layer when operating outside the electrolyte stability window. Structural degradation within individual primary particles can occur due to defect formation, or unstable phase formation. If unstable phases form near the electrode/ electrolyte interface degradation can be accelerated by parasitic reactions from the electrolyte.

Improving the electrochemical properties of high voltage cathodes requires detailed understanding of the underlying degradation mechanisms. However, characterising degradation in Li-ion cathode materials is difficult because Li is hard to detect, and complex local environments exist within a single electrode. Ideally, techniques used to characterise degradation should correlate the underlying electrode microstructure with degradation induced chemical and structural changes.

Here, novel electron and ion microscopy techniques are developed to study degradation phenomena at the electrode surface, within individual primary particles, and during cycling. The techniques aim to correlate chemistry and microstructure, and enable understanding of the role of local environment in degradation behaviour. The techniques are validated by conducting a comprehensive study of degradation in high voltage Li-ion battery material LiCoPO4.

The CEI layer is challenging to image and characterise because the CEI layer is thinner (10-20 nm) than the anode equivalent SEI layer (up to 1 µm), and CEI layers typically contain Li. He-ion microscopy (HIM) and secondary ion mass spectrometry (SIMS), a mass spectrometry technique integrated into some helium ion microscopes, are suited to CEI layer studies because HIM is a highly surface sensitive imaging technique, and SIMS is capable of detecting Li. Here, He-ion microscopy (HIM) and in-situ Ne-ion time-of-flight secondary ion mass spectrometry (SIMS) has been used for the first time to study CEI formation resulting from electrolyte oxidation on the cathode.

HIM is used to image the CEI layer on LiCoPO4 at length scales enabling the thickness and morphology of the CEI to be related to the underlying microstructure. HIM imaging showed that the CEI forms on LiCoPO4 agglomerates. SIMS mapping and depth profiling characterisation identified that the CEI layers are composed of oxyfluorophosphates, and a layer of uncycled Li exists on the surface of charged electrodes. Using HIM SIMS to analyse cathodes at different states of charge and after different cycle numbers, enabled direct imaging of the partial dissolution of the CEI layers across the electrode. Partial CEI dissolution is most significant on larger LiCoPO4 agglomerates. The HIM SIMS technique provides new opportunities for correlating underlying electrode microstructure with heterogeneous CEI formation.

Li is hard to detect with most chemical spectroscopy techniques. Co in LiCoPO4 oxidised from Co(II)-Co(III) as Li is removed. Electron Energy Loss Spectroscopy has been previously used to measure the change in Fe valence state in LiFePO4 to study lithiation mechanisms. Here, the de-lithiation mechanisms of LiCoPO4 are visualised post-mortem using an electron energy loss spectroscopy (EELS) technique to map changes in the valence state of Co across the electrode.

Using Co valence state EELS, the shrinking-core de-lithiation mechanism of LiCoPO4 was directly visualised. The unstable CoPO4 phase forms on the outside of LiCoPO4 particles, leaving it susceptible to damage from the electrolyte. Damage from the electrolyte is observed as a layer of Co(II) on the outside of particles, along with a region of trapped Li on the surface of electrodes charged to 5.1 V vs. Li/ Li+. Results from valence state EELS characterisation, in conjunction with the HIM SIMS results indicate that the cyclability of LiCoPO4 could be improved by improving the stability of CoPO4, or developing methods to shield CoPO4 from electrolyte degradation.

To prevent the impact of sample damage, and ensure transient states during charging could be characterised, an in-operando TEM technique was investigated to image changes in primary particle microstructure during cycling of active materials. The initial set-up aimed to create a Li-ion half-cell in the TEM. LiCoPO4 was successfully isolated on the nano-cell working electrode, however, plating of Li from the electrolyte onto the reference electrode was unsuccessful. The results presented suggest a pseudo-reference electrode cell, using Pt as the reference electrode, would be a better set-up for in-operando TEM.

As part of the in-operando TEM set-up, the electrolyte was irradiated with the electron beam. It was found that the electrolyte formed a series of precipitates with increasing electron dose, which post-mortem energy dispersive X-ray spectroscopy identified as consisting of carbon, fluorine, oxygen, and phosphorous. The precipitates formed as a result of electrolyte reduction from the electron beam. The results present an opportunity to use a closed liquid cell set-up in the TEM to study reduction mechanisms of different lithium ion battery electrolytes.

The novel electron microscopy techniques presented in this thesis offer future routes to characterising complex degradation mechanisms in high voltage Li-ion battery materials. The methods enable correlation of chemical and structural changes with electrochemical history, whilst maintaining an understanding of the complex local environment.