Homogeneous SWCNTs
A light topic for discussion today – SWCNT as a conductive additive for lithium ion battery systems. It’s no new news that SWCNT or MWCNT are applied as conductive additives even in industrial applications as there has been numerous literatures that back up its effectiveness is improving the z-axis electron flow in both anodes and cathodes. The high aspect ratio of CNTs compared to other carbon conductive additives make it a sort of electron highways and even act as scaffolds that keeps the overall electrode intact throughout the treacherous volume expansion/contraction that is long term cycling. It is due to this clear advantage that CNTs bring that it has been extensively applied despite it being quite… quite.. expensive.
But the difficulty with high aspect ratio nanomaterials lies in its high surface area. High surface area increases a CNT strand’s affinity with other CNT strands, which makes homogenous dispersion incredibly difficult. And really, what good is a clump of CNT in one 100um x 100um area of an entire 200mm x 100mm electrode sheet?
Again, this is all common knowledge but, physical and chemical dispersion methods have been applied for CNT dispersion. Common dispersion methods include acid treatment and simple sonication but the expensive nature of effective sonicators (i.e. tip sonicators) in tandem with the incredibly reversible nature of these CNT dispersions make it not so ideal. Chemical dispersion methods refer to dispersion agents (surfactants) where it aims to reduce the affinity of individual CNT strands with other CNT strands. The two that are most common – based on my experience and limited research – are SDS (sodium dodecyl sulfate) and PVP (polyvinyl pyrrolidone) respectively.
As a Ph.D. candidate, cost is usually not an issue – whatever gets your research into top tier journals is selected and applied. This usually meant that I could even consider increasing the CNT wt.% to shockingly high values or even consider dip-coating CNTs onto fabrics (1) if it got the job done. Now this is a luxury that is simply not practical. The art of industrial engineering is to get the job done while making sure it’s not over-engineered – the act of simply increasing the amount of CNT when an alternative, optimized method can be achieved is exactly what can be considered as ‘over-engineered’.
So, this got me thinking – which method between SDS and PVP is better?
The simple answer is that while both dispersion agents have a trade-off point where the amount of surfactant (SDS) or polymer coating (PVP) becomes too much as to start disrupting electron flow, SDS is considered better. The work by Kang et al. (2) explores and compares the two – it’s in Korean so here’s a quick summary.
The research team attempted to create a polymer-CNT nanocomposite using PS scaffolds and infiltrating CNTs uniformly between the scaffolds and freeze-drying to remove the solvent to yield the PS-CNT nanocomposite.
PVP and SDS were compared by comparing the morphologies and electrical conductivities of the two respective nanocomposites.
Results indicated SDS showed comparatively better dispersed PS-CNT nanocomposite. To achieve electrical conductivity of 10^-9 S/m, the PS/SDS-CNT required a minimum of 0.23 wt.% CNT content while the PS/PVP-CNT required a minimum of 0.90 wt.% CNT content.
This is attributed to the lower molecular weight of SDS (288 g/mol) vs. PVP (40,000 g/mol) as this means that SDS attaches to the CNT strands as ‘dots’ while the PVP attaches to the CNT strands as ‘layers’ thereby being more likely to disrupt connection between CNT strands.
Left) SDS acting as ‘dots’ on CNT strands and Right) PVP acting as ‘layers’ on CNT strands
In conclusion, SDS should be the preferred method for dispersion when considering CNT applications. Personally though, I’m not quire convinced the true impact of both SDS and PVP on the overall long-term performance of LIBs in a wide range of temperatures. My experience with SDS involved countless washing with de-ionized water to get rid of the excess SDS (endless soap bubbles) and I imagine degradation in high voltage environments and/or organic solvents will yield by-products.
In an industry where competitive/comparative advantages are won with an accumulation of baby-steps, it’s these kinds of small differences that I think will be crucial and consider fascinating as it links directly to performance and cost.
References
1) Tae Gwang Yun, Donghyuk Kim et al, 2018, Advanced Energy Materials, 8 (21), 1800064
2) Myung Hwan Kang et al, 2013, Polymer(Korea), 37 (4), p.526-532
