Understanding and Comparing the Stability of Water- Versus NMP-Based Tin(IV)Sulfide Electrodes Using Post-Mortem Analysis

Abstract

The market for lithium-ion batteries (LIBs) has experienced a tremendous growth since 1991, when the first LIB was commercialised by Sony, and it is expected to grow even further. The reason for this upward trend is on the one hand the improvement in performance, specifically in energy density and cyclability, and on the other hand the world’s increase in energy demand. LIBs are now able to power not only small electronic devices but also electric vehicles and stationary energy storage systems.In addition to the increased need for batteries with high energy densities, environmental aspects must also be considered. Firstly, commercial LIBs contain critical raw materials (CRM) such as Li, Co, and graphite which are often concentrated in a few areas of the world and are sometimes mined using techniques which exploit both humans and the planet. Secondly, the use of toxic chemicals such as N-methyl-2-pyrollidone (NMP) in electrode manufacturing raises concerns on the degree of sustainability of the product. Furthermore, in order to reduce the CO2 emissions and achieve climate-neutrality by 2050, the EU (BATT4EU) has set goals regarding energy densities, battery life time and safety. Desired values for gravimetric and volumetric energy densities for Generation 3 LIBs are considered 350-400 Wh/kg and 750-1000 Wh/l respectively for 2000+ cycles. These energy values can be achieved by either developing cathodes with higher voltage and/or by developing cathodes and anodes with higher specific capacity. The latter approach together with sustainability aspects are addressed in this work. Tin(IV)sulfide (SnS2) has been investigated in the last few years as one of the most promising anode materials for Li-ion batteries due to its high theoretical specific capacity of 1209 mAhg-1 compared to commonly used graphite (372 mAhg-1). However, its theoretical reversible capacity is 644 mAhg-1 due to the conversion of SnS2 to Li2S and Sn in the first discharge. Furthermore, SnS2 can be processed using water as a solvent in place of previously mentioned NMP. This way, not only the electrochemical performance, but also the environmental friendliness, as well as costs during fabrication can be improved.This study focusses on the optimization of water-based slurry processing using SnS2 as active material, and water-soluble binders such as sodium carboxymethylcellulose (Na-CMC) and styrene butadiene rubber (SBR) in comparison to NMP-based electrodes using Polyvinyl difluoride (PVDF) as a binding agent.In this work, four different electrode compositions were investigated, namely SnS2/C45/CMC:SBR(1:1) in a 90/5/5 ratio and a 80/10/10 ratio, as well as a comparative system consisting of SnS2/C45/PVDF in the same ratios as the previous electrode. All electrode compositions showed an initial specific capacity of ~550 mAh/g, however the cycling stability was in the order of H2O-80:10:10 > H2O-90:5:5 > NMP-80:10:10 > NMP-90:5:5 (Figure 1), with the best water-based electrodes reaching 80% state-of-health (SoH) after 200 cycles.In order to understand the degradation mechanisms occurring during cycling and to elucidate the impact of H2O- and NMP-based processing on electrochemical performance, techniques such as Raman, SEM, XRD, XPS, and EIS were performed on pristine electrodes, on electrodes after first discharge (formation cycle), as well as on samples at 80% SoH. Furthermore, in-situ dilatometry was conducted to measure the linear change in length of the electrodes in the direction perpendicular to the electrode surface during cycling. Changes in crystal structure, morphology, surface chemistry (i.e. SEI formation), interface and charge transfer resistance, and volume expansion were correlated with the electrochemical performance.(Attached: Figure 1: Comparative graph of galvanostatic cycling with potential limitation (GCPL) of water- and NMP-based electrodes with varying ratios of active material/conductive agent/binder showing the development of capacity and coulombic efficiency as a function of cycle number) Figure 1