Thread by @enn_nafnlaus

Thread: Looking At Tesla's New Raw Materials

While most media coverage seems focused on Plaid, this was a battery event, and among the flood of fundamental rethinks about the way batteries are made are the changes in raw materials.

Anodes: traditionally these are milled graphite (natural or synthetic) optionally alongside a *small* amount of silicon. Tesla appears to be using *no* graphite. All of those graphite resources people have been trying to secure? Worthless. It's metallurgical silicon instead.

I've discussed how much of a step-up silicon is previously - ~3600mAh/g vs 372mAh/g for graphite, and nearly as good as Li metal (3860mAh/g, albeit at a slightly higher voltage). Note that Tesla won't *actually* be in the ~3600 range, as the polymer encapsulant adds mass.

But what's remarkable is that they're doing this with plain metallurgical silicon - not even semiconductor grade. Only ~98% pure. Currently a ~1,5Mt/yr market. If that were all used for batteries, it'd be ~20TWh/yr capacity *already*. Scaling potential is limitless.

Production is typically done in a fairly typical carbothermal reduction process, but using an arc furnace to melt the silicon. That said, you could also do hydrothermal reduction or carbon capture if you wanted to produce it without carbon emissions.

Caveats: while their polymer encapsulant is designed to prevent electrical isolation of cracked silicon and destruction/reformation of the SEI from swell/contract cycles, we have no real-world data about how effective their approach is. One *assumes* cycle life is enough, but...

Regardless, they're basically reducing the anode mass down to near irrelevance. This isn't happening on the cathode side. Nor do they appear to be using the rumoured single-crystal cathode approach.

They seem to be implying that it's cobalt free, yet I was left with a feeling of some vaguery on this front - and their pilot plant included cobalt sulfate on the list of materials they'll be handling. I'm not sure whether they're all the way there at this point.

"Novel coatings & dopants" likewise could not be more vague.

The interesting discussion however is with respect to nickel: that they're ditching battery-grade nickel sulphate and instead going to "metal". Initially I read this as ferronickel, but it appears not to be.

Ferronickel (pic) = only about 35% pure nickel, ~65% iron, for use in making stainless steel; they seem to be using class 1 or high-grade class 2 nickel. That said, the main late stage impurity in pyrometallurgical is Co, and they should be able to skip it's solvent extraction.

A pyrometallurgic product, it's easily scaled, and indeed already at mass scale, about 2,1Mt/yr for Class 1 supplies, enough for ~2,5TWh/yr. Class 1 suppliers have been hurting with oversupply for the past few years, so are surely quite happy with this scaleup news.

(Indeed, Tesla's challenge right now is to convince producers to start projects to produce *even more*, at a time where the market has been oversupplied for years, and only recently starting to recover)

It was interesting to me to see Tesla's diversified cathode approach to help overcome scaling supply limits (e.g., nickel). I've not been much of a LFP bull because the lower voltage means buying more of "everything else" in the cell - but now that "everything else" is so cheap..

(I'm *assuming* that LFP is what's being referred to as "iron based", although there are other iron-bearing cathode possibilities)

On the lithium front, one can't help but notice the mention of hydroxide and the bullseye right near Piedmont in North Carolina.

At 27,9Mt reserves, Piedmont has 3 1/2 times as much lithium as the fabled (but expensive) Salar de Uyuni deposit in Bolivia. A spodumene resource, its direct output is lithium hydroxide, which is normally sought after for high-nickel cathodes.

But Tesla is rewriting the game.

We now know that Tesla has bought a >10k acre (>4k hectare) lithium clay claim in "Nevada". Let me be more specific: it would be in the Clayton Valley, near Silver Peak, which contains a salar that Albemarle runs as the US's only currently operational lithium producer.

It was initially thought that lithium collected in the salar from inflowing water, and leached into the clay from the surface. It's now believed to be just the opposite: lithium is liberated into the clays from decomposing bedrock and seeps up to the surface. Now they're sought.

The size of the lithium resource is staggering, and Tesla wasn't kidding about how it could replace every vehicle in the US and then some. The valley is underlain by ~100m of clay made of ~0,1% lithium, give or take. Here's Noram's Zeus deposit.

~100m at 0,1% ~= 20 Model 3 LRs worth of Li under every square meter, trivially mined. The total exploitable area is unknown, as it keeps expanding as exploration continues, but is probably in the lower hundreds of square kilometers. Tesla's claim is in the low single digit %s.

A new resource, lithium clay ventures have been striving to prove that they can *economically* extract Li from the clays at below current market prices. This is generally done with a sulfuric acid leach, but Tesla intriguingly reports a new NaCl (table salt)-based process.

In this process, Na+ would substitute for the Li+ in the clays, yielding a LiCl output, which apparently Tesla plans to use directly. While details are sparse, it's certainly appealing; lower corrosion, cheaper raw materials, and no need to neutralize the tailings.

Overall, Tesla's announcements entirely shuffle up the battery world. If their process takes over, graphite resources become irrelevant. All of that hard work on HPAL for nickel sulphate production is pointless. Anyone betting on high raw material costs, loses.

But let's not lose sight of the fact that we're still early. This line has only produced ">10000 cells". That's a couple dozen packs, tops. Will they iron out production kinks? Will the cells meet demanding standards? Tesla thinks so, but time will tell.

Prepare for a slog.

You may now honk your horns.