Reliable TEM specimen preparation is essential when transmission electron microscopy is used to examine lithium-ion battery materials. For silicon-based anodes, however, preparation can change the very features an engineer or materials specialist needs to observe: lithium atoms in formed lithium silicide may be removed by ion and electron beams. A combined focused ion beam and ultramicrotome method offers a way to target a specific region while reducing ion-beam damage to the final thin specimen.
Why silicon anodes are difficult to examine
Silicon is of interest as a lithium-ion battery anode because it has a high theoretical capacity and a low working potential. Its major challenge is substantial volume expansion and contraction during lithiation and delithiation. These changes can lead to material fracture, capacity loss, and poor cycling performance. Understanding how the material’s microstructure changes during charging is therefore important to evaluating its behavior.
The study examined an alloy anode made of active silicon and an inactive TiFeSi2 matrix. After 20 charge-discharge cycles, the battery was dismantled in an argon-filled glove box. The recovered material was cleaned with dimethyl carbonate to remove remaining lithium salts, then dried. The charged alloy had fractured as it expanded, leaving shards from which a region could be selected for microscopy.
Preparation is a critical part of this analysis. A thin, well-positioned specimen allows a defined cross-section to be examined, but the preparation process itself can introduce artifacts or alter sensitive phases. In this case, the lithium-containing phases were particularly vulnerable: ion-beam milling and electron-beam observation could remove lithium from the material.
Comparing FIB and mechanical sectioning
Focused ion beam (FIB) preparation is useful when a specific area of a sample must be extracted and thinned. It can produce a cross-section from a chosen location, but the gallium ions used in the reported method damaged the silicon structure. The authors also found that specimens thinner than 70 nm were affected by ion-beam damage, and selected a thickness of 70 to 100 nm for their conventional FIB cross-sections.
Ultramicrotomy, by contrast, uses a cutting blade to produce thin sections mechanically, without milling the final specimen with an ion beam. Mechanical preparation has its own risks: force from cutting can cause plastic deformation or tears. In addition, conventional ultramicrotomy can make it difficult to isolate a precise region of interest, potentially requiring many sections to obtain a useful observation area.
The combined FIB-ultramicrotome technique was designed to address these opposing constraints. FIB identifies and extracts the target area; an ultramicrotome then cuts thin sections from that extracted material. This divides the task between the two methods: FIB provides location-specific sampling, while mechanical sectioning avoids ion-beam milling of the final slices.
How the FIB-ultramicrotome method works
First, FIB was used to extract a cube approximately 10 × 10 × 10 µm from a selected area of the charged anode. The cube was attached to a copper lift-out grid and embedded in an epoxy block. After curing, the block was trimmed around the sample and mounted for ultramicrotomy.
A diamond knife with a 35-degree angle was used to cut sections at a set thickness of 70 nm. Water was used to collect the sections. The researchers reported that anhydrous ethanol caused specimens to sink, while another organic solvent solidified along the knife and block, making cutting and collection difficult. To minimize reaction with lithium, the collected sections were dried promptly and prepared for TEM analysis.
The approach produced successive slices below 100 nm thick. Because FIB had already isolated the region of interest, the area requiring mechanical trimming could be kept small. That made it possible to obtain continuous thin sections while retaining the ability to examine a selected part of the fractured material.
What the microscopy results showed
Selected-area electron diffraction (SAED) and energy-dispersive X-ray spectroscopy (EDS) were used to compare specimens made by conventional FIB and by the combined method. Both sets of observations showed the TiFeSi2 matrix. However, their diffraction patterns differed: the FIB-only specimen showed silicon in an amorphous state, consistent with ion-beam damage, while the FIB-ultramicrotome specimen showed crystalline silicon and a lithium-silicide phase identified as Li21Si5.
Lithium silicide is sensitive to electron-beam exposure. In this analysis, the researchers reported that it remained stable during SAED when a very low-intensity beam was used in a widely expanded state. This observation highlights that specimen preparation is not the only consideration: microscopy conditions also matter when examining beam-sensitive phases.
EDS mapping showed silicon and TiFeSi2 matrix in both types of specimen. Gallium was detected in the analyzed area of the FIB-only specimen, but not in the FIB-ultramicrotome specimen. The authors interpreted this as evidence that FIB affected the surface of the extracted cube without significantly affecting its interior, which became the mechanically sectioned specimen. These findings support using the combined method when ion-beam damage could compromise analysis of a target region.
Practical implications for materials analysis
For laboratories investigating lithium-ion battery anodes, the choice of preparation method should reflect both the material’s sensitivity and the analytical question. FIB can be valuable for precise site selection, while ultramicrotomy can provide sections without directly milling those final slices with gallium ions. The combined method is therefore relevant when researchers need location-specific sampling and want to preserve crystalline silicon and lithium silicide for examination.
The technique does not remove every preparation concern. Mechanical cutting can introduce deformation or tears, and the reported procedure depended on careful embedding, trimming, sectioning, and prompt drying. Sample thickness and collection conditions also require attention. The results apply to the tested Si-TiFeSi2 anode material and preparation conditions; the report proposes the approach as potentially useful for other materials that are sensitive to ion beams and require observation of a specific area.
Conclusion
Microscopy results are only as informative as the specimen that reaches the instrument. In silicon-based anodes, conventional FIB preparation can alter silicon and lithium-containing phases, complicating interpretation of the charged material’s microstructure. Combining FIB-based site selection with ultramicrotome sectioning offers a practical alternative: it retains access to a defined region while reducing ion-beam exposure to the final TEM slices. For materials teams, matching preparation and observation conditions to the sensitivity of the sample is a key step toward more dependable microstructural analysis.