Ionenleitung an Grenzflächen von LiNi0.5Mn1.5O4-Dünnschichtkathoden und Lithiumionenleitern

Im ersten Teil der Arbeit wurde die Ionenleitung in der LICGC, einer kommerziell erhältlichen Lithium-Ionenleiterkeramik, untersucht. Mit Hilfe der Hochfeld- Impedanzspektroskopie wurden nichtlineare Effekte untersucht, die an Grenzflächen auftraten. Nach Entwicklung eines empirischen Modells zur Da...

Full description

Saved in:
Bibliographic Details
Main Author: Gellert, Michael
Contributors: Roling, B. (Prof. Dr.) (Thesis advisor)
Format: Doctoral Thesis
Published: Philipps-Universität Marburg 2015
Online Access:PDF Full Text
Tags: Add Tag
No Tags, Be the first to tag this record!

The first part of the thesis deals with the ionic conduction in a commercially available lithium-ion conducting ceramic. High-field impedance spectroscopy was used to study non-linear effects, that occurred mainly at interfaces. The thickness of the grain boundary was estimated with a newly developed model that describes the ionic conductivity of the grain boundaries as a function of the applied field. A calculation of the applied field per single grain boundary indicated that a space-charge layer could not be the reason for their relatively low ionic conductivity. Space-charge barriers should disappear when a sufficiently high-voltage is applied and what should have led to a disappearance of the grain boundary resistance. Consequently multiple barriers have to exist in the grain boundaries. To shed more light on their nature, high-resolution STEM has been performed. The obtained results were used to develop a model describing the charge-transport across the whole glass-ceramic. The relatively high grain boundary resistance was found to be caused by non-ideal grain to grain interfaces. Grain boundaries of neighboring crystals with similar crystal lattice orientation showed dislocations, where one lattice orientation passes over into the other orientation. This was most likely the reason for the slightly higher activation energy of the grain boundary conductivity. Grain boundaries of neighboring crystals with strongly dissimilar crystal lattice orientation were found to be highly resistive and amorphous, which in combination with Al(PO)4 precipitations decreased the total conductivity. The second part of the thesis is about the development of a thin-film high-voltage cathode for lithium-ion batteries on gold plated stainless steel substrates and the characterization of their electrochemical properties. Several processes could be identified via electrochemical impedance spectroscopy that indicated the presence of an additional layer that increased the total resistance of the cathode. ToF-SIMS measurements and cross-sectional STEM-EDX images of the cathode showed two different additional layers that had formed at the interface of the gold plating and the cathode material. These layers could not be attributed to resistive processes found in impedance data. Another layer was found at the surface of the cathode by ToF-SIMS. The thickness of this SEI-type layer perfectly matched the impedance data. Consequently, both of the two additional layers found in cross-sectional STEM-EDX measurements must exhibit very high electrical conductivity. In the third part of the thesis, LNMO thin-film cathodes have been prepared on gold substrates and subsequently coated with LiNbO3 (LNO), Li4Ti5O12 (LTO) or ZrO2 (ZRO). The electrochemical characterization of the coated electrodes revealed slightly increased cell resistances, but also improved capacity retention of LNO and LTO coated cathodes compared to uncoated LNMO cathodes. In contrast, the oxide ZRO with low ionic conductivity caused very high internal resistances and lower discharge capacities. The investigation of the cathode morphology by means of SEM–EDX and ToF-SIMS revealed impurities in the LTO and the ZRO coating by cations that originate from the LNMO active material layer. A newly formed solid electrolyte layer with unknown chemical composition covered the remaining LNMO material. Nevertheless, he LTO coated cathode showed good electrochemical properties. Thus, the newly formed electrolyte layer and a potentially existing SEI should exhibit high electrical conductivity. To elucidate the chemical composition of the cathode surface and the electrolyte layer, HAXPES has been performed. The measurements showed a very thin SEI layer on top of the cathode consisting of electrolyte decomposition products and cations from the solid electrolyte layer. This SEI layer showed no impedance response. The newly formed solid electrolyte also exhibited the original LTO spinell structure with the octahedral sites filled up by manganese and nickel ion nearly to a value of 50%. The variety of transition metal ions in several oxidation states presumably increased the electronic conductivity and caused the very low resistance in the newly formed spinell solid electrolyte layer. Although the cell resistance increased slightly when a cathode with LNO or LTO coating was used, the impedance spectra did not show a separated process caused by ion conduction across the solid electrolyte coating. It was likely that the impedance and capacitance of both, the charge-transfer process and the ion conduction across the electrolyte layer, were too similar to result in separated impedance processes. Consequently, additional cathodes with considerably thicker LNO coatings were prepared in order to separate the processes in the impedance spectra. A linear correlation between the overall cell resistance and the solid electrolyte layer thickness was found. A linear fit was used to separate the solid electrolyte layer resistance from the charge-transfer resistance. It can be concluded that a LNO coating can greatly reduce the charge-transfer resistance cell resistance. Additionally, the coating can act as a chemical barrier in solid state batteries for chemically incompatible components.