Thread by @mjlacey

Seems my threading earlier on @QuantumScapeCo went a bit bananas, so I'll collect some general post-presentation post-dinner thoughts/questions/observations under here... 1/

https://twitter.com/mjlacey/status/1336340656923365382

So finally we got to see some cycling data: 1 hour charge/discharge, practical capacities, >800 cycles achieved seemingly quite readily - for a so-called anode free Li metal cell these results are very impressive! 2/

https://twitter.com/mjlacey/status/1336347732160761861?s=20

Key caveat though - these are single layer pouch cells, which need to be brought up to many (20+) layers to commercialise. This is presumably not easy else I'd expect a company at this stage to show proof of concept multilayer cells... 3/

A reason for this might be QS's use of a catholyte (liquid electrolyte on the cathode side only) which probably can't be added last to a multilayer pouch as normal. It's solvable though, is my gut feeling 4/

https://twitter.com/mjlacey/status/1336353576793354248?s=20

One eyebrow-raising result was the reported plating of Li metal at >100 mA/cm2! Since a 15-minute fast charge would need "only" ~15 mA/cm2 this sounds like a big deal. But this plot has prompted much debate over in the #BMWS chat... let me explain... 5/

https://twitter.com/mjlacey/status/1336346573161328644?s=20

My interpretation - which so far seems to agree with available details - is that this is a Li | Li symmetric cell where the current goes through increasing current steps, and the voltage is measured. For each current step, about 0.1 mAh/cm2 (by my eye) is passed 6/

So very interesting as this is, this *shouldn't* be interpreted as evidence of a 2-minute fast charge. It's closer to saying the anode can handle 2% of full charge in about 3 1/2 seconds. Not criticising, just don't assume this extrapolates to cells and full charges... 7/

Worth mentioning, there's been much talk about this plot because it is not clear exactly what it shows or how it should be interpreted. The use of two y-axes confuses the relationship between voltage and current here, so students: don't use two y-axes, it's not good practice! 8/

Elsewhere, @steingart has been poking at the claimed necessity of Li metal to reach high energy density. Since the cells are still relatively small single layer pouch cells, we'll have to wait to see what QS will be able to achieve here 9/

https://twitter.com/steingart/status/1336344120776937475?s=20

I wrote a bit about this back in September. While I agree Li metal is theoretically best for gravimetric energy, Si can still be competitive, and on volumetric energy, could beat Li metal. So I don't think QS have blown the competition away... just yet 10/

https://twitter.com/mjlacey/status/1302313352333492226?s=20

Lots of chemistry questions - what is the electrolyte, how have they solved the interfacial problems - still seem to be a closely guarded secret. So questions are for sure still going to be asked until proper multilayer cells are demonstrated 11/

https://twitter.com/stevelevine/status/1336361277753610242?s=20

This also means that other big questions - especially things like cost - we're just going to have to wait a bit longer for more info/clues... 12/

So if I might sum up... @QuantumScapeCo's results are... tantalising, they look to be onto something very promising. But experience tells us it's easy to underestimate how hard it is to go from where they are now to >20 GWh/yr production 13/

I really hope the hype can be kept in check and the expectations don't become unrealistic. I still feel their timeline of commercialisation in ~4 years is very optimistic, but who knows - things move very fast in batteries these days. I'll be watching closely... /end, maybe