Fundamental and Applied Studies Towards the Development of All-Solid-State Batteries Based on Sulfide-Based Alkali-Ion Conductors

Both Na-ion and Li-ion solid-state ionics show promise as underlying chemistries of energy storage technologies, which might play a significant role in future efforts to store energy more efficiently and safely, however, with slightly different scopes. ASS-LIBs will very likely be the key to introdu...

Full description

Saved in:
Bibliographic Details
Main Author: Duchardt, Marc
Contributors: Roling, Bernhard (Prof. Dr.) (Thesis advisor)
Format: Doctoral Thesis
Language:English
Published: Philipps-Universität Marburg 2020
Subjects:
Online Access:PDF Full Text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Both Na-ion and Li-ion solid-state ionics show promise as underlying chemistries of energy storage technologies, which might play a significant role in future efforts to store energy more efficiently and safely, however, with slightly different scopes. ASS-LIBs will very likely be the key to introduce renewable energies to the mass-market of lightweight electric vehicles. ASS-SIBs, being currently at a considerably earlier stage of development, might eventually contribute to cheap grid-scale energy storage causing lower costs compared to ASS-LIBs. In the first publication, Na11Sn2PS12 (NaSnPS) was reported, which exhibited an unexpected stoichiometry of Na11Sn2PS12. This finding contradicted the previously reported theoretical works, according to which a quaternary compound NaSnPS would possess a stoichiometry of ‘Na10SnP2S12’, corresponding to the quaternary Li counterparts, e.g., Li10SnP2S12 or Li10GeP2S12. In addition, a hitherto unprecedented crystal structure was found, considerably deviating from the one of Li10SnP2S12. Even though both structures constitute the same basic building blocks of alkali ions and unconnected [SnS4]4– and [PS4]3– tetrahedra, they are very different, with the one of NaSnPS being larger and more complex. Strikingly, the elemental ratio between Sn and P is also inverse, being 2:1 for NaSnPS and 1:2 for the Li analogues. Its grain conductivity amounts to a value of 3.7 mS/cm, making it, at the time, the fastest sulfide-based Na-ion conductor. The focus of the second paper lies on the discovery of NaSnPSe, hence the heavier analogue of NaSnPS. Whereas in the initial study, due to failed attempts at growing single crystals, the crystal structure of NaSnPS had to be solved from powder X-ray data, in the subsequent study, single crystals from both NaSnPS and NaSnPSe were obtained. This enabled a much closer look at the occupancies of the six different crystallographic Na positions that make up the cationic lattice and thus the underlying conduction mechanism. The previously developed idea that all Na positions take part in long-range ionic motion was corroborated, and two interstitial positions were identified as playing a crucial role in creating the necessary 3D pathways. NaSnPSe is the first example of a quaternary selenium-based alkali ion conductor. Similar substitution of Se for S in the Li analogues proved impossible apart from negligible partial substitutions. In contrast to the widespread belief that a softer anionic lattice necessarily leads to a higher ionic conductivity at r.t., NaSnPSe in fact exhibits a lower grain conductivity at r.t. than NaSnPS with only 3 mS/cm. This underlines not only that the activation energy of ionic motion has to be taken into account, which is indeed lower for NaSnPSe, but also the attempt frequency, which is generally lower if the lattice is softer. Interestingly, it was found that Na11.1Sn2.1P0.9Se12 repeatedly showed a composition slightly off from the expected stoichiometry analogous to that of NaSnPS, which indicates the tendency of forming solid solutions. In the field of Li-ion SEs, research is noticeably evolving from the sheer discovery of new materials to sophisticated ways of producing them, finally enabling their cost-competitive large-scale manufacturing. In the third report, an innovative synthesis strategy for the preparation of sulfide-based SEs was presented. Li3PS4·3THF is a readily available intermediate product that can be transformed into beta-LPS or Li6PS5Cl. It is generally obtained by the wet-chemical synthesis of stoichiometric amounts of Li2S and P2S5 in anhydrous THF. However, while P2S5 is very cheap, Li2S, the feedstock chemical for all syntheses of sulfide-based Li-ion SEs, is extremely expensive, because the common synthesis from Li2SO4 as precursor involves very high temperatures above 1000 °C. Therefore, in this work, a strategy was conceived, which allows for the in-situ generation of Li2S from the elements. To this end, lithium and sulfur are loaded into a flask together with THF as solvent and naphthalene (NAP) as electron transfer agent. The radical anion (NAP–) then reduces the sulfur to S2–, finally resulting in Li2S. If high loadings of NAP are used, the reaction proceeds quickly. This preceding step is synergistically coupled to the subsequent formation of Li3PS4·3THF upon the addition of P2S5, giving rise to a one-pot synthesis or flow-oriented synthesis of Li3PS4∙3THF. Even though elementary Li has to be used, which is, of course, a considerable energy toll, rough calculations show the approach to be promising from a cost perspective. -LPS or Li6PS5Cl generated from Li3PS4·3THF show their typical characteristics without any deterioration due to residual traces of NAP, which proves the practical applicability of the suggested strategy. In addition to the identification of promising SEs (especially regarding their ionic conductivity, electrochemical stability and plastic properties) and their efficient synthesis, a similarly important goal is the optimal adjustment of the SE’s morphology. In the fourth publication, morphological aspects of ASS-LIBs were therefore investigated and correlated with their electrochemical performance. Two model cells were constructed that were based on the so-called sheet-type ASSB concept and incorporated state-of-the-art NMC-85|05|10 as CAM and either beta-LPS or LPSI as SE. It could be demonstrated that beta-LPS with its peculiar, microporous morphology enables a much more favorable cathode morphology than LPSI. The latter with its larger particles leads to a lower contact area between CAM and SE, more void space and thus more serious constrictions. Numerical transport simulations indicated a roughly doubled tortuosity for the cathode based on LPSI as SE component. The four times higher intrinsic ionic conductivity of LPSI, though, overcompensates this effect and enables an overall superior battery performance. This shows that the charge transfer plays only a minor role. The actual bottleneck for a better battery performance is the ionic transport within the SE phase inside the composite cathode. Even though the ionic conductivity is the most crucial prerequisite for an SE, careful attention should be paid to the morphology of the SE and thus the morphology of the resulting cathode, in order to fully benefit from the high ionic conductivities of state-of-the-art SEs.
Physical Description:88 Pages
DOI:10.17192/z2021.0074