Settling down for the much-anticipated #QuantumScape reveal. A little taste of this already on their website, link below. Feels like a thread coming on... ir.quantumscape.com/news/news-deta…
Interesting to hear from the start a clear disclaimer about this showcase presenting "forward looking statements", making clear that reality might differ from expectation outside of QS's control
Recap of investment: they're well-funded by now, having raised ~$1.5bn so far, about ~$300m has gone into R&D so far. VW are committed to the tune of $300m, and have a joint venture with QS for manufacturing. Now they're listed, market cap was ~$15bn last I checked...
More recap of basic concept: solid state electrolyte with Li metal anode, but the cell is fabricated "zero excess" or "anode free", i.e., there is no Li metal in the cell from the start, it is plated out of the cathode on 1st charge, for maximum energy density
Conventional knowledge is that this means big capacity losses because Li metal generally means bad efficiency. QS is arguing that not only have they solved that, but going without an anode host means eliminating a transport bottleneck, enabling fast charge as well...
Interesting to note already that QS seem to indicate already that the "garnet" class of solid electrolytes is classed as an "unsuccessful attempt". Most of us guessed that they were using the garnet "LLZO"... maybe not? Wonder if we'll get some clues... 
An important point - so far it seems they are only presenting results on single layer pouch cells. Going from single layer to multi-layer (e.g. tens of layers in practice) adds complications (pressure, inhomogeneity etc) which could be just one challenge for them from here
Big new result: they are claiming they can reach >100 mA/cm2 Li plating without failure - this would correspond to around 2 minutes full charge. Crazy impressive if true, but I'm sure many will be super skeptical without seeing more data (me included tbh)...
Some impressive looking drive cycle data as well and comparison to a C/Sia anode, although it is hard to know at first glance how this translates to "real" cycle numbers that we're used to. Going to need to look at all this much more carefully later on
This is more conventional and looks very impressive. Showing many repeats as well, showing decent consistency (important!). Zero excess, sensible capacity, very typical test for Li-ion batteries. But still relatively small cells, so interpret with caution
This is interesting too. DSC curves of lithium + electrolyte - no exothermic reaction as the lithium melts, an indication of stability of the electrolyte/separator towards lithium metal
Now we have the panel discussion. At this point I thought I'd dig up my thread from Sep on QS: big Q was on how thin could they make their separator? This is the big challenge for energy density. So far, we still don't know
https://twitter.com/mjlacey/status/1302313352333492226?s=20
Big endorsement in the panel from (NP laureate) Stan Whittingham, hailing QS's achievements and data - "just got to make the cells bigger and get them into cars"
Endorsement from @venkvis - the key result is that all the challenges are solved at the same time, "it passes the 'and' test"
Very interesting thought from @alexlouli in the #BMWS chat. QS currently use liquid electrolyte on the cathode side (confirmed by @stevelevine!) to make that interface work. This would pose a problem for making their cells multi-layer...
In normal production electrolyte is added after stacking/winding - here presumably you'd want to add it before, to not get liquid into the anode side where it can cause problems. It's solvable, but probably needs to be a new production step they have to develop
As @jordi_sastr3 notes this is probably the way they need to go. FWIW I think it's sensible, even if it is not "truly" solid state it would still eliminate most of the liquid component. Whether this still offers big safety advantages is yet to be seen
https://twitter.com/jordi_sastr3/status/1336355623118790657?s=20
Jagdeep addresses something I was thinking of - volume expansion of Li, definitely an issue for making thicker cells. Making the argument this can be designed for e.g. on module level - it's more a problem if you have 3D volume expansion, which you do with Si anodes
Some electrolyte details - "Our electrolyte is different - but it is on the same order of $/kWh as ordinary polymer separators". And that it can be processed roll-to-roll to keep costs low
And that's the call over. Time for me to get something to eat and digest this... plenty to think about and come back to later!
• • •
Missing some Tweet in this thread? You can try to
force a refresh
