Lithium-Ion Battery Solid Electrolyte Interface (SEI) Analysis Using the fibTOF

Valentine Riedo-Grimaudo, James Whitby, Lex Pillatsch
TOFWERK, Switzerland

The Solid Electrolyte Interface (SEI) in lithium-ion batteries is critical for determining both the battery’s capacity and longevity. During multiple charge cycles, the SEI behaves dynamically and its composition changes over time. This evolving nature of the SEI, particularly evident during galvanostatic cycling, significantly affects the battery’s performance and lifespan.

Much battery research today focuses on analyzing the chemical composition and molecular structure of the SEI to develop optimized interface designs. However, current instrumentation often lacks the spatial resolution and sensitivity needed to analyze lithium at interfaces that are typically around 100 nm thick or less. The FIB-SIMS technique (Secondary Ion Mass Spectrometry using a focused ion beam microscope), combined with the advanced mass analyzer of the fibTOF detector from TOFWERK, addresses this challenge. This application note highlights a FIB-SIMS study of SEI composition in a lithium-ion battery cell with LiNiₓMnyCo1-x-yO₂ cathodes. It also explores Transition Metal (TM) dissolution from the cathode and its deposition on the anode during galvanostatic cycling, as well as the impact of molecular additives on SEI properties and TM shuttling.

The TOFWERK fibTOF Detector

The addition of a mass analyzer to a Focused Ion Beam (FIB) – Scanning Electron Microscope (SEM) expands the range of possible investigations on microscope platforms by enabling FIB-SIMS. The fibTOF provides rapid real-time 3D visualization of the spatial distribution of all elements, including hydrogen and lithium, as the sample is milled by the FIB beam [3,4]. The spatial resolution, which can be better than 50 nm laterally and 10 nm in depth, together with limits of detection in the ppm range (e.g. for B, Li, F), make the fibTOF an ideal instrument to probe the solid electrolyte interface in lithium-ion batteries.

Figure 1.  Positive secondary ion elemental maps showing the distributions of the major lithium isotope ⁷Li⁺,  Li₂O⁺ and Li₂F⁺ molecular ions, and the isotope of manganese ⁵⁵Mn⁺ for a lateral projection vs depth. The number of frames (Z-axis) is the number of completed raster by the FIB beam, and is directly related to the fluence received, and indirectly related to the depth. Projections obtained with the same additive are arranged in columns. BE stands for baseline electrolyte and is the control sample where no additive was applied during cycling [5].

Experimental

In this work, a gallium FIB was used with the beam normal to the sample surface. The FIB bea was set to 30 keV and 1 nA, and the field of view was adjusted to 30 µm x 30 µm. This work investigated six different graphite anodes LiNi_0.6Mn_0.2Co_0.2O₂: one pristine, four cycled to the end of their life (50 % state of health) in were vinylene carbonate (VC), fluoroethylene carbonate (FEC), chloroethylene carbonate (ClEC) and vinylethylene carbonate (VEC).

Solid Electrolyte Interface Composition

It is possible to select arbitrary slices averaging and regions of interest from the 3D data set collected with the fibTOF. In this case, the elemental distribution along the SEI was of particular interest, which is along the erosion direction (Z) of the FIB beam. Therefore, X-Z projections were compared as presented in Figure 1. These projections were created by the X values for every probed Y position. This reduces the available information from 3D to 2D. The chemical composition distribution of the investigated lithium species (Li, LiF and Li₂O) and manganese shows significant differences along the depth of the electrodes and the respective SEIs. These findings become even more pronounced when comparing the results by means of depth profiles, as illustrated in Figure 2. Depth profiles are obtained when averaging per removed layer (frame) all acquired data points from the same X-Y plane, thus reducing the data set to 1D.

Figure 2. Depth profiles of several lithium- and manganese- containing secondary ions. The peak position and the extent of the distribution depend upon the additive used. In addition to the BE control sample, the results for the original sample, the graphite anode that is not cycled (pristine), are plotted, showing the absence of the species under investigation.

For the case of the BE (Baseline Electrolyte, no additive added) sample, as well as for the anodes cycled in the presence of the additives FEC or VC, the majority of lithium is found in the very top layers of the electrodes. This indicates the formation of an effective SEI, preventing further electrolyte decomposition on the anode. In contrast, the anodes cycled with ClEC and VEC, have a deeper penetration depth of the lithium species, indicating the formation of an ineffective solid electrolyte interface, which leads to continuous electrolyte decomposition and the deposition of lithium deeper in the electrode. A different trend can be observed for the manganese depth profile. While the majority of manganese is detected in the top layers for the BE, VC and VEC samples, manganese is found also in deeper layers of the electrode for the FEC and ClEC samples. These findings suggest earlier dissolution of the cathode active material for the FEC and ClEC samples, due to the formation of hydrofluoric acid (HF) and hydrochloric acid (HCl), etching transition metal from the cathode active material [6].

Conclusion

This application note demonstrates the use of the TOFWERK fibTOF for the imaging of lithium species in the solid electrolyte interface of lithium-ion batteries. These light elements are difficult or impossible to image with comparable spatial resolution by other techniques. Further, the study confirmed irreversible phase changes and metal dissolution of the cathode.

References

[1] U. S. Meda et al., Journal of Energy Storage, 47, (2022), p. 103564.  https://doi.org/10.1016/j.est.2021.103564

[2] Y. Chu et al., Electrochemical Energy Reviews, 3, (2020), p. 187. https://doi.org/10.1007/s41918-019-00058-y

[3] L. Pillatsch et al., Progress in Crystal Growth and Characterization of Materials, 65, (2019), p. 1. https://doi.org/10.1016/J.PCRYSGROW.2018.10.001

[4] J. A. Whitby et al., Advances in Materials Science and Engineering, (2012), p. 180437. https://doi.org/10.1155/2012/180437

[5] F. Pfeiffer et al., Advanced Energy Materials, (2024), p. 2402187.
https://doi.org/10.1002/aenm.202402187

[6] J. C. Hestenes et al., ACS Materials Au, 3, (2023), p. 88.
https://doi.org/10.1021/acsmaterialsau.2c00060

Acknowledgment

We thank Dr. Baghernejad and his team from the Forschungszentrum Jülich, Germany, for collaboration and proving us the samples for analysis.