Abstract
Silicon is a promising anode material for next-generation lithium-ion batteries due to its exceptionally high theoretical capacity and natural abundance. However, its practical implementation remains limited by severe chemical-mechanical degradation caused by large volume changes during lithiation and delithiation. The understanding of how these localised processes occur and accumulate across length scales is crucial for the successful development of long-lasting silicon-rich electrodes.
This thesis uses a correlative, multiscale imaging approach that combines electron microscopy and synchrotron X-ray nano tomography to study degradation mechanisms in silicon-based anodes. Time-resolved three-dimensional imaging is applied to directly observe microstructural evolution, lithiation pathways and deformations under realistic cell operating conditions. Advanced image analysis, including deep learning-based segmentation, generative image enhancement and digital volume correlation, enables quantitative analysis of phase evolution, stress localisation and particle-level response.
The results reveal strong heterogeneity in both lithiation behaviour and mechanical deformation, starting from the very first electrochemical cycle. While only a small percentage of silicon particles follow the traditional core-shell lithiation pattern, the majority exhibit complex internal lithiation networks that are influenced by local microstructural constraints and processing-induced defects. The findings suggest that strain localisation has a greater influence on degradation than average expansion. The developed framework’s wider applicability to solid-state battery materials is further demonstrated by applying it to sulphide-based solid electrolytes.
This thesis uses a correlative, multiscale imaging approach that combines electron microscopy and synchrotron X-ray nano tomography to study degradation mechanisms in silicon-based anodes. Time-resolved three-dimensional imaging is applied to directly observe microstructural evolution, lithiation pathways and deformations under realistic cell operating conditions. Advanced image analysis, including deep learning-based segmentation, generative image enhancement and digital volume correlation, enables quantitative analysis of phase evolution, stress localisation and particle-level response.
The results reveal strong heterogeneity in both lithiation behaviour and mechanical deformation, starting from the very first electrochemical cycle. While only a small percentage of silicon particles follow the traditional core-shell lithiation pattern, the majority exhibit complex internal lithiation networks that are influenced by local microstructural constraints and processing-induced defects. The findings suggest that strain localisation has a greater influence on degradation than average expansion. The developed framework’s wider applicability to solid-state battery materials is further demonstrated by applying it to sulphide-based solid electrolytes.
| Translated title of the contribution | Multiskalige mikrostrukturelle Bildgebung und fortgeschrittene computergestützte Analyse zur Aufklärung des chemo-mechanischen Verhaltens in Batteriematerialien |
|---|---|
| Original language | English |
| Qualification | Dr.mont. |
| Awarding Institution |
|
| Supervisors/Advisors |
|
| DOIs | |
| Publication status | Published - 2026 |
Bibliographical note
no embargoUN SDGs
This output contributes to the following UN Sustainable Development Goals (SDGs)
-
SDG 7 Affordable and Clean Energy
Keywords
- Lithium-ion batteries
- silicon anodes
- electrode degradation
- lithiation mechanisms
- in situ X-ray nano tomography
- electron microscopy
- CNN-based segmentation
- image enhancement
- particle-level analysis
- digital volume correlation
- mechanical stress
- solid-state batteries
- 3D imaging
Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver