Lithium metal has been used as an anode in lithium batteries for more than 40 years. Its stability in organic electrolytes is linked to the spontaneous formation of a passive surface layer, which effectively prevents further electronic interaction with the electrolyte. On the other hand, this layer conducts lithium ions and is therefore commonly known as the Solid Electrolyte Interphase (SEI). The detailed composition and morphology of the spontaneously formed SEI is strongly dependent on the electrolyte used. The SEI working model assumes that the structure of the layer is at least double. Closer to the metal anode, a thin (few nm) compact SEI layer is formed, which is electron-impermeable but allows the conduction of Li+ ions. This compact layer is followed by a thicker (some 100 nm) porous SEI layer. The pores of this part of the SEI hold the electrolyte through which the active species are transported to the electrode. This model is based on measurements of impedance spectra taken in combination with a systematic variation of the cell parameters, which allowed a more accurate understanding of the processes. Currently, there is no established consensus in the research community on the presence and magnitude of the impedance contribution for the charge transfer reaction, which takes place at the phase boundary between lithium metal and compact SEI, and the contribution for Li+ ion desolvation, which takes place at the boundary between porous and compact SEI.
As part of the thesis work, we have attempted to determine and distinguish between the contribution of the charge transfer reaction and the desolvation of the lithium ion by impedance spectroscopy measurements on symmetric cells with lithium-metal electrodes. For this purpose, lithium electrodes were pre-passivated in different mixtures of electrolytes and solvents. The electrodes prepared in this way were then used to assemble symmetric cells with different salt concentrations in the electrolyte. The impedance spectra of the cells were measured at open circuit potential. These spectra were analysed to determine the variation due to a lower electrolyte concentration at the same SEI pre-composition. We hypothesised that changing the electrolyte concentration significantly changes the impedance contribution of desolvation, but not the contribution of the redox charge transfer reaction, since this occurs at the phase boundary to which the electrolyte has no access.
|