While in particular the SEI has been intensively studied by spectroscopic methods, first and foremost by x-ray photoelectron spectroscopy (XPS), little is known about its local distribution and potential structural inhomogeneities due to the lack of spatial resolution of these methods.
Both the binder as well as the SEI form structures which are typically just a few nanometer in thickness and are therefore not accessible to light microscopy or even standard scanning electron microscopy (SEM). While high resolution SEM (HR-SEM) or transmission electron microscopy may in principle be able to resolve SEI or binder structures, their investigation by electron microscopy is far from straight forward. This is to one part due to their composition. As electrode binders typically polymers like PVDF or CMC/SBR are used while the SEI consists of a mixture of inorganic salts of low z-elements like LiF, LiCO3 or LiOH and organic lithium salts. The low electron density of these materials in combination with a mostly amorphous structure makes detection and distinction highly challenging. This is even further complicated on the negative electrode side, because it mainly consists of carbonaceous materials, and therefore posses a very similar scattering contrast. Further, spectroscopic methods available to electron microscopy cannot easily detect lithium and carbon and oxygen are part of the carbonaceous host matrix and therefore not suitable for distinction. Another issue when studying the SEI by electron microcopy is its high sensitivity to beam induced radiation damage.
While the energy dispersive x-ray (EDX) signal of fluorine may be used to detect gradients in the binder distribution for polyvinylendifluoride (PVDF) of a fresh electrode it falls short in resolving local binder structures due to the large scattering volume contributing to the EDX signal which limits resolution. For the SEI, which also contains fluorine in form of LiF, EDX is even less applicable, since the fluorine signal of LiF cannot be discriminated from that of the PVDF binder. Therefore, despite a rather similar initial problem, different approaches need to be followed to investigate binder and SEI structures by electron microscopy.
In order to study the SEI layer we used the reactivity of some of its components towards osmium tetroxide fumes (OsO4). OsO4 is a strong oxidising agent and may therefore react with parts of the SEI layer. The osmium compounds formed significantly increase the local scattering contrast. The contrasting and fixation properties of OsO4 has beneficially been used for preparing biological tissues for electron microscopy. However, its application in material science is scarce. In this contribution we demonstrate how OsO4 can beneficially be used to study the SEI layer by electron microscopy. Further, we will shed some light on how OsO4 reacts with parts of the SEI and how the selective reaction of OsO4 may be used to improve our understanding of its structure.
For the binder distribution a different approach was chosen. Energy selective imaging of backscattered electrons (ESB) is a rather new electron microscopy technique. Under certain conditions it may help to increase the contrast in samples with weak elemental contrast at a certain combination of acceleration voltage and filter grid voltage. Further, the low energy of the ESB electrons make it a very local probe with high lateral resolution. We were able to identify appropriate imaging conditions to clearly distinguish PVDF binder structures on carbon electrodes. The technique was applied to study the influence of the drying process on the binder distribution and new insights in how the drying process influences the distribution of binder and carbon black inside the electrode layer could be obtained.