All-solid-state batteries (ASSBs) are expected to revolutionize the Li-ion battery landscape thanks to multiple benefits, such as improving both intrinsic safety and potentially enabling high energy anode concepts.[1,2,3] Amongst the variety of possible solid electrolyte (SE) materials for ASSBs, sulfidic electrolytes are considered particularly advantageous, as they possess high ionic conductivities and beneficial mechanical plasticity, allowing their pre-densification by uniaxial cold pressing at high fabrication pressures (p fab).[4] However, one bottleneck of sulfidic ASSBs compared to conventional liquid-electrolyte batteries is the necessity of a certain operating pressure to provide/maintain an intimate mechanical contact between the various materials during materials characterization measurements (e.g., SE conductivity measurements) and/or battery cycling. For these, a special cell setup is required that is hermetically sealed and facilitates a defined and, ideally, variable application of different compressions (operating pressure, p oper) on the sample/cell stack.[5] Typically, ASSB characterization cells possess a defined circular geometry, whereby the components are placed in an inert polymer containment (e.g. a PEEK tube) and are electrically contacted via metal dies.[2,5,6] However, a problem which is often neglected in the literature is the poor solid-solid contact of pre-densified SE pellets (done at a high p fab) and mechanically hard current collector dies to contact the sample for impedance measurements. This leads to impedance artifacts and potentially inaccurately determined solid electrolyte Li+-ion conductivities (σ Li).[2] In this contribution, we firstly describe a novel ASSB cell setup enabling the artifact-free measurement of the ionic conductivity of pre-densified SE-binder sheets in rigid cell matrices. By applying this setup, we could decouple effects from fabrication pressure and operating pressure, providing the basis for determining the correlation between the SE separator porosity and its Li+-ion conductivity, comparing it to the correlations predicted by the Bruggeman equation.We use sheet-type separators consisting of Li6PS5Cl (LPSCl) solid electrolyte and different amounts and types of binder, such as hydrogenated nitrile butadiene rubber (HNBR) and polyisobutylene (PIB), prepared by a slurry-based process.[5] By means of potential-controlled electrochemical impedance spectroscopy (PEIS), the impact of different p fab and p oper on the Li+-ion conductivity of the LPSCl/binder-sheet separators are investigated. As a result, we show that densifying the SE separator sheets at a high p fab, resulting in a low residual porosity, can be used to increase their conductivity at lower p oper. However, a phenomenon which is not reported in such detail so far, is the sensitivity of the measured σ Li of pre-densified LPSCl/binder-sheet separators regarding the p oper. While for pure pre-densified SE powder pellets only a negligible sensitivity regarding the p oper was observed,[2] scalable pre-densified LPSCl/binder-sheet separators show a strong correlation between p oper and measured σ Li.The extent of this phenomenon can be traced back to a “springback” effect, wherby the release of pressure after fabrication a dense binder-sheet separator leads to an apparent decrease in conductivity. We found a clear trend between the binder amount and the extent of this springback effect in LPSCl/binder-sheets and how this affects σ Li at different p oper. Furthermore, to deepen the knowledge about this springback effect, we also investigated the extent of the springback effect on the type of binder (HNBR vs. PIB) and its molecular weight (PIB OPPANOL N50 vs. PIB OPPANOL N150).Additionally, to fully understand different contributions of the pressures and springback to the overall separator impedance, we used PEIS measurements at a very low temperature (-70°C) and a measurement cell with a negligible stray capacitance to clearly distinguish between SE bulk, grain boundary, and springback effects as well as contact resistances, in the impedance spectra for such LPSCl/binder-sheets.Lastly, we evaluate how the springback effect contributes to an overall full-cell impedance in LiNi0.6Co0.2Mn0.2O2|LiIn half-cells and provide a guideline to overcome contact resistances for pre-densified sheet-type ASSB cell components. Based on our findings, it can be concluded that the springback effect contributes to the overall impedance not only in pure SE impedance measurements, but also in half-cells and cannot be neglected when characterizing slurry-coated SE/binder separators-sheets.[1] J. Janek and W. G. Zeier, Nature Energy, 1, 16141, (2016).[2] J.-M. Doux et al., J. Mater. Chem. A, 8, 5049 (2020).[3] S. Cangaz et al., Adv. Energy Mater., 2001320 (2020).[4] Z. Deng et al., J. Electrochem. Soc., 163, A67 (2015).[5] C. Sedlmeier, T. Kutsch et al., J. Electrochem. Soc., 169, 070508, (2022).[6] W. Zhang et al. ACS Appl. Mater. Interfaces, 9, 17835 (2017). Acknowledgements: This work was carried out as part of the project "Industrialisierbarkeit Festkörperelektrolyte", funded by the Bavarian Ministry of Economic Affairs, Regional Development and Energy and also supported by the BMW Group.
Read full abstract