This Machine Could Be A US Manufacturing Breakthrough

It's the early morning in Knoxville, and so the pilgrimage to Matt Robb's Biscuits has begun. You have from 9 to noon to acquire a biscuit, but they're usually all gone before that window expires. Such is life in the hard streets of Knoxville, which, okay, are not actually hard at all. You've got graffiti giving way to parks, quaint streets packed with restaurants and bars, and in an ultimate flex, you've got a sun sphere. Powerful. Our interests, of course, are away from the city center and in this office-cum-laboratory, because this is the home of Perseus Materials, and we have come to watch them cook.

The goal of Perseus is to simplify how we build enormous structures. One of the things that we really focus the tech development towards is wind turbine blades that can be up to 150 meters.

So something like a wind turbine, is that made with a mold?

Yeah, the mold is even bigger than the part itself. And so you need special factories, you need special— just to move it around. The upfront cost and lead times are a huge problem. So we developed a combination of the chemistry processes and the new automation capabilities together to invent a kind of mold-free process. You no longer need the lead times and the upfront cost of the tooling.

The first key to this technology is some chemistry Dan helped develop while studying at Stanford.

We started talking to blade makers and composite manufacturers, and we kind of identified core constraints that seem to exist in every composite manufacturing process. And then we traced it to a limitation in chemistry. And we didn't use that to then invent new chemistry. We had the chemistry, we just didn't realize what what this phenomenon could be used for. And then when we realized, now we have really weird freedom in how we design and process.

What Perseus has built is just straight up bizarre. It wants to make huge structures out of composites like fiberglass and Kevlar by using this thing. It's a machine that has some of the flexibility of a 3D printer, but does jobs usually associated with more traditional manufacturing techniques like stamping and casting.

The goal is to do any fibers, but today we're using glass.

Here we go, we're pouring some resin onto glass fiber, and then we're gonna cure it and shape it at the same time. The second big leap forward takes place on the machine itself. The material goes down, and the machine shapes and cures it at the same time. Zach, what are you doing? Just checking to see if it's cured. Okay.

You wanna tap it?

Like a little bit hard. Feel back here, tap it hard.

Yeah.

And then feel right here. Oh yeah, yeah, way more solid. A little gooey on one side and firm on the other. Yeah, so that's— I'm just checking to make sure it's curing as we're going through.

We kind of redid the process where we're actually curing it as we're pulling. That gives us flexibility with pulling infinitely. We can decide whatever shape we want on the go. We composite, 3D printing, pultrusion.

Resin is the heart of the thing that you're controlling.

Yeah.

And a resin is what, like a plastic?

Yeah, it's basically a plastic precursor. Yeah. So it's a liquid form that then hardens.

So you're putting the resin on the sheet.

Yeah.

And then you're triggering that process through heat and pressure?

I'm going to leave that a little bit vaguer, but yes.

Okay. Okay. So that's where the magic happens is how you're triggering this process.

If we want to revolutionize how we build big things, let's say there's a 12-mile bridge and you want to build the entire skeleton of the bridge on site so you don't have to transport it, components, and do the assembly. What does manufacturing have to look like? It has to be extremely fast. It has to be faster than any manufacturing technology that exists today. It has to be extremely reliable and it has to be very low cost. And so that is only capable with, like, with the advanced synthetic chemistry that we've developed and worked on. We can manufacture faster than steel.

That's cool. So, like, in your future vision, you might have these machines right by whatever you're trying to make.

Yes.

I've seen just about every new stab at manufacturing there is, but have never run across anything quite like this. It's chemistry meets materials science meets computing in a totally new form. Perseus is in Tennessee in part because it struggled to win over venture capitalists. This forced it to chase some government grant money that came from the Department of Energy and to set up shop in Knoxville.

So we're funded by ARPA-E, NSF, and Department of Energy AMTO office. And part of that last one was a partnership with Oak Ridge National Lab.

Which is just down the road.

Which is down the road. So we opened up shop nearby and started working.

I can understand the doubts around the technology. It might work okay on these small prototype parts, but then struggle to make big stuff with enough precision and speed. Still, I like that this is new and different. And it's not hard to see a bigger, more automated version of this with lots of actuators cranking things out. If you're not here for bridges being printed on demand, I don't know what to do for you. Build your company faster and better. Switch to Brex to upgrade your finances. After enjoying my time huffing fiberglass with the boys, I needed to make a pilgrimage to Knoxville's ultimate star attraction, the University of Tennessee. I mean, it's no University of Texas, but the surroundings are all right. Inside this university building is Eonyx. It's a startup run by Don DeRosa, and it's trying to make better batteries faster.

My name's Don DeRosa. I'm the CEO of Eonyx. We're an accelerated materials discovery company for lithium-ion batteries.

We are in a new electricity-rich era. As cars, planes, and all sorts of other vehicles are now running on lithium-ion cells. Making new battery chemistries that are safer and more efficient is a must, but this process tends to require huge investments and many years of research and development. And so, EONIX has decided to automate the R&D process using a mix of software and robots.

Everything is running by itself. Formulations are automatically sent to our robot.

Okay.

So we have the ability to use 12 different additives, 4 different solvents, or 2 different salts at a time. And you're able to get up to about 32 solutions done in just a few hours.

Okay.

Now for context, we used to have 2 lab techs who would work at this station and they'd get about 11 or 12 solutions done per day. So we have a significantly higher output. We don't require 2 humans to be at the station at all times. And you don't have to work in a glovebox, which is a huge upside.

Ionix tries to blast through as many promising battery chemistries as possible. Then it measures and analyzes the chemical reactions in ultra-fine detail.

Once we automatically formulate those suggestions, they go into a series of machines to evaluate thermal performance, conductivity, viscosity, things we want to know before they should even be considered for a battery. So we're able to see like this one doesn't freeze all the way down to -80°C, that's probably a good candidate for low-temperature applications like space or eVTOLs or Arctic conditions for military applications. We've made a lot of nonflammable lithium-ion batteries. We've made one that doesn't suck.

The major goal is to find chemistries that last longer and are less flammable and that can slot right into existing battery designs.

So we evaluate different chemistries, different electrolytes, and we pick out which one works better. And then we drop it right into a commercial cell. So this is a cell that's really 99% of the way there from a manufacturing perspective. Only thing it's missing is the liquid electrolyte. We go over here. This is just a glove box. We fill it, we seal it. So it's crimped right here. And then we have a battery and that's it. Getting something to just drop into batteries that are already manufactured, that's the real trick.

We're gonna go blow a battery up.

We're gonna blow a battery up, yeah. So we're gonna do a totally real test called the nail pen test where we take a fully charged lithium-ion battery and we drive a nail through it.

Let us in! I wanna blow stuff up!

Alright, so they have the tool open so you can see the chamber.

Oh, cool.

It's an accelerated rate kilometer. Allows us to determine basically the thermal performance of a battery on failure. So we're going to drive a nail into that battery. We're going to short it internally. And, well, this one's going to explode.

Okay.

Like a lot.

So much anticipation. This is the most awkward part. I hate this part. It scares me every time. Are you guys going to count down?

Just be like, any minute now. You're going to hear it.

Yeah.

Yeah. Oh, nice.

Okay. Oh, shit. Yes. Oh, that's a nice fire. How hot did this one get?

This was just under 800.

800 Celsius.

800 Celsius. Wow. Feeling satiated by my daily dosage of explosions, it was time to push on across Tennessee and through its weather. It was time to get moody and make a religious pilgrimage of sorts. It was time to go to Oak Ridge.

City of Oak Ridge, Tennessee, while a vast army of workers kept the secret of a new age.