Green Giants – Episode 119

David Mackanic (01:36)

You know, I’ve always thought that batteries are amazing systems and amazing technology. It’s really elegant that batteries have this contained ability to store energy and it’s a nice mixture of chemistry, physics, mechanics, and electrical engineering on top of that. And so there’s a lot that’s captivating inside the battery. And then to me, I always thought that…

we’re in the golden age of battery adoption. Slowly everything that can electrify does. And so I really wanted to be part of that transition. And so I went to Stanford to study that.

new category of battery materials. And, we were specifically focused on polymers, but we were really focused on a range of techniques to help solve problems inside batteries. So problems with degradation, problems with capacity fade, problems with mechanical fracture. And the more I thought about this, the more I saw that we had all these different chemistries that were degrading and having all sorts of issues with stabilization. And it was clear that, you know, we could spend all the time at Stanford research

these polymer materials to kind of solve problems with the existing electrolytes and the existing chemistries or we could try to reinvent the electrolyte and so that’s what I ultimately decided to do because I saw its applicability across all of these exciting new developments that were happening in the battery ecosystem.

Wes Ashworth (02:57)

I love that. And I think so important as you said, I think batteries are so central to everything right now. It’s hard to have a conversation in energy without going directly to batteries and bringing that into the equation as well. And so working on all facets of that is really intriguing. What was that point we you know when this stopped me feeling like an interesting PhD project and started feeling like this is something I need to build around and need to kind of become a company out of it.

David Mackanic (03:22)

Yeah, I mean, there’s a few things it’s so hard to wrap your mind around. You know, when I got to Stanford in 2015 to when I left in 2020, the battery industry size had between doubled and tripled during that time. I mean, the growth in this industry has just been massive. And it’s doubled or tripled since then as well. So we’re talking about a 10x growth over the past 10 years, roughly, right. And so that’s something that was really exciting is some of the stuff that we were working on might have been considered niche

at the beginning, but it was becoming more necessary as the demand for these batteries were growing. Specifically, what happened is that we had published a paper on some of these advanced materials that we had developed, specifically these advanced polymer electrolytes, which we can talk about later. And we had a company reach out to my PhD advisor saying, Hey, can we buy some of this and use it in a prototype battery that we’re developing? And that was pretty interesting to me because

you know, stuff you do in the lab is not necessarily meant to be shipped out to companies to start working with, but that gave us the opportunity to say, hey, let’s try to get this into a package that we can actually sample to this external party. We ended up not doing that for various intellectual property reasons, so on and so forth, but it did open up that avenue of, this type of material that we’re working on is something that the industry is looking for.

Wes Ashworth (04:41)

the appetite was there, obviously, and the interest coming in even from that perspective. And, you know, there’s always that huge leap between proving something can work and deciding, you know, it matters enough to really start building a company around it. So we’ll get more into that. But you know, I think that original insight leads us directly into the part of the battery most listeners probably understand the least. I want to spend a little bit time there. So most people hear better battery and they think new anode, new cathode, or entirely new chemistry. what’s everybody missing about the electrolyte specifically?

David Mackanic (05:10)

It’s a good question.

Historically, the electrolyte has been somewhat overlooked in battery innovations. so the conventional electrolyte inside the battery is the material that allows for ions to flow in the battery. So it allows for that battery to charge and discharge. Colloquially for things like AA and AAA batteries, this is sometimes referred to as battery acid. So that stuff that kind of can corrode or leak out of a AA or AAA battery. Granted, it’s a different kind of chemistry in a lithium ion battery, but the concept is the same.

And a lot of the problems are the same. So that battery electrolyte hasn’t been reimagined really for the past 30 years or so. And it’s a flammable, toxic, corrosive, volatile substance that continuously degrades inside the battery cell. In fact, the battery chemistry is designed to

function during a system of controlled degradation of these electrolyte materials, which is why your battery fades over time, which is why there’s these battery safety issues that get reported. But perhaps most importantly is why higher energy chemistries are fundamentally limited because they’re reacting with this electrolyte. And so what

we’ve tried to do is to develop an electrolyte that solves a lot of these fundamental chemical instabilities in a much safer, a much more stable, a much more mechanically robust form factor. And we can talk more about our specific technology, but that’s a little bit of background of why the electrolytes are such a problem in today’s batteries.

Wes Ashworth (06:34)

Yeah, it’s great. And we’re gonna get more into it and kind of dive into to some of the specifics of that. But it doesn’t really change the mental model because the electrolyte, you at first kind of sounds secondary until you understand kind of everything it influences. And it’s a useful reminder as well that the battery is a system. Changing one material can unlock very different behavior elsewhere. so that’s what we’ll get into. But we’ll kind of take it back to sort of simple, you know, understandings. For someone who has never taken a battery apart, it’s probably a lot of people. give us the movie version of your product, Proteus.

what goes into the cell as a liquid? What happens during formation? And what is different after it solidifies? Like help have a basic understanding of this.

David Mackanic (07:10)

Yeah, so that’s a great question. Our technology that we’ve branded Anthro Proteus is a new class of phase change electrolyte. So it’s an electrolyte that undergoes a liquid to solid phase transition inside the battery cell. So some of the listeners might have heard of something called a solid state battery, which is a type of battery that can solve a lot of the issues with the liquid electrolyte that we talked about previously. But as of today, there’s not a…

large-scale commercial party that can manufacture these solid-state batteries at scale. The manufacturing process for solid-state batteries remains extremely difficult. And so we’ve taken a different approach with this phase change process. And so what happens with our technology is a battery gets built exactly like normal. You can use any of the standard battery components in any battery form factor that exists today. We then inject a liquid precursor material exactly like you inject that normal liquid electrolyte. But then instead of

having it stay as this liquid material, has a bunch of issues we talked about previously, we use a process called battery formation to induce a liquid to solid phase transition inside the cell. So that liquid actually polymerizes, creating a mechanically robust solid inside the cell, forming really strong interfaces with the cathode and anode and kind of gluing the cell together. So you can imagine the inside of the battery, which has all of these layers of energetic materials are kind of now adhered together. The inside

ends up looking something like you could imagine like a composite structure the way that skis or surfboards might be kind of constructed with a fiberglass and then kind of this epoxy gluing it together. And so that’s what’s going on inside the battery and the thing that’s awesome is that this is allowing us to solve a lot of the problems with the conventional liquid electrolyte but not change the manufacturing process for the battery at all.

Wes Ashworth (08:55)

Yeah, I think it’s so important to understand that. And we’re gonna dig a little bit into that as well, too, and sort of the benefits of that. And but I think, you know, ultimately that liquid first, solid later idea makes the manufacturing logic easier to visualize. And you know, it does sound deceptively simple when you describe it. There’s a lot happening inside that transformation. And know, one thing you brought up, so people throw around, you know, solid state battery, you know, as though it describes one technology. where does Proteus kind of actually sit on the spectrum from conventional liquid lithium ion to a fully solid state architecture?

David Mackanic (09:25)

It’s a good question and unfortunately the term solid state in the industry has lost a lot of meaning because people use it to say a lot of different things and so the

A more appropriate way to think about Proteus would be a semi-solid state approach. You can think about these things as completely dry, which means they use no liquid components, would be a pure solid state or pure dry battery. Obviously, that doesn’t apply to us because we are injecting this liquid into the cell as the first part. And so we’re somewhere in between these definitions. But…

So are most companies that are producing solid state batteries today. And so we try not to get too much into the solid state versus not solid state. And we try to focus on what it means for the end battery. Is it safer? Does it cycle longer? Is it more mechanically robust? Is it higher energy density? So on and so forth. And then the other thing to call out about the Proteus technology is that there’s a wide range of formulations that use different sets of molecules, some which could be considered more conventional solid state and some which could be considered more

more semi or quasi solid state. So again, the terminology is a little bit less important to us. What’s important to us is that this material becomes a mechanical solid inside the cell to solve a lot of these issues with degradation, volatility, flammability, so on and so forth.

Wes Ashworth (10:39)

I appreciate the clarification because again, that distinction matters that solid state has become that sort of catch-all phrase for some very different technical approaches. So I think good to clarify there. and here’s a claim I want to pressure test a little bit. So you can get many of the benefits people want from solid state without rebuilding the factory altogether and no, serious new equipment. and that’s a little different from completely no process change, but what actually changes on the manufacturing line when you’re looking at your product?

David Mackanic (11:06)

This is a good question. 100 % drop-in has also become a little bit of a buzzword. And what we really mean when we say that is no new equipment. So there needs to be zero changes to any of the process equipment inside an existing battery facility.

Every new cell design, every new electrode design, every new chemistry does require adjusting to some of the process steps. things like the injection pressure of the liquid, things like the wet out time of the liquid, things like the specific formation voltage and specific formation time that the battery gets subjected to in this specific formation temperature if there’s a temperature control step. So these are normal tweaks that happen within battery manufacturing.

When you switch to our electrolyte, there is a chance you have to tweak some of these process parameters on the existing equipment. And so we say drop-in means no new equipment, but there might be some minor process changes that need to get made, all of which are within the range of kind of normal lithium-ion battery process conditions.

Wes Ashworth (12:04)

Absolutely. And I think that’s the distinction manufacturing people are going to care about immediately, not having to look at all new equipment and those sort of things. But some, you know, minor process tweaks and changes. I think that’s a completely doable and understandable, and what makes it so easy to adopt and also scale. so a material working in the you know, hundredth cell is very different from working in the millionth cell. what does become brutally difficult when you try to make this reaction happen consistently at industrial scale?

David Mackanic (12:29)

It’s a good question. We have definitely seen that as we have moved to larger and larger production batches as well as actually larger quantities of material as well. The thing that we have really needed to emphasize that we haven’t had previously, and I think we’ll get into this later as well, is an emphasis on

quality and process for all of the materials that go into the actual manufacturing process, as well as all of the unit operations that we’re performing, and implementing additional inline diagnostic tools that allow us to monitor if things are going well throughout the process. And so this has been one of the things that we have implemented on our production facility here in Alameda, is the ability to have a number of different metrics that ensure that this chemical reaction is happening to its completion, that we can perform on

standard battery cyclers, for instance, or standard formation equipment. And so we’ve had to implement a number of checks along the way that allow us to monitor in real time if things are going according to plan. And then again, making sure that we are really buttoned up in terms of all of the materials that are going into the manufacturing process, tracing them back to the origin of where we source them and making sure we’ve done quality control on all of those components. That’s been really essential. So developing high throughput ways to screen all the input materials.

and then developing mechanisms to test the quality and the success of these operations in real time so that we can catch them before things get too far along in the process and make any changes that we need to.

Wes Ashworth (13:59)

I love it. It’s always fascinating seeing how all that comes together. And as you go from kind of lab to commercialization and continuing to scale, solving, solving those issues and all that, how that all works. And you know, I think here we’ve established sort of the basics of how this works, you know, how the battery works and some context there. And I think the harder question is what the customer actually gets from putting it into a battery. And that’s what I was spent a little time on. So you can improve energy density, cycle life, swelling, mechanical strength, and safety.

which one do customers actually care enough about to pay for today? And if it’s that’s customers like individuals, but also you know there’s these manufacturing companies of people that are actually bringing your product into their process.

David Mackanic (14:35)

Yeah, it’s this is a really good question. And one of the things that we view a little bit differently at Anthro is we don’t view these as necessarily separate things. These things are highly coupled things like the energy density, the safety, the cycle, like the swelling behavior. And so while, you know, it’s possible to make a battery that has really high energy density, if it falls short on some of those other issues, it’s simply not going to be commercializable. It’s not going to meet the standards that the end user expects. And so we don’t focus

on a specific energy density number. We focus on enabling people to push energy density while maintaining or improving all of the other safety, cycle life, swell, mechanical strength features that are critically important. And you one of the things that we are finding a lot of initial success with is that virtually every battery manufacturer today has a silicon anode material on their roadmap. Silicon anode is a material that promises to increase the energy density of batteries so making them

up to 40 % thinner, up to 40 % lighter. But the problem is when you start to put more silicon into the battery, the batteries become very unsafe. They do not cycle as well, they start to swell up. And these create all sorts of issues when you’re integrating into a device. And so we’ve seen that our ability to stabilize silicon is something that customers and end users are willing to pay for. You don’t have to look too far to see how much people are willing to pay for kind of these thinner, higher performing batteries. You can look at consumer electronics, you know,

iPhone Air is one or two millimeters thinner, it’s more expensive, right? People want that thinner device. And then of course with EVs, if you can get kind of longer range at the same cost profile, that’s gonna make the vehicle more valuable to the customer overall. So that’s where we’re finding the ability to, you know, bring value to both the manufacturer and end customer is solving some of these issues with swell cycle life safety to unlock that improved energy density.

Wes Ashworth (16:27)

Yeah, I like

that. I always like a multiple choice where somebody’s all the above. It’s all important, it’s all

David Mackanic (16:31)

Hahaha

Wes Ashworth (16:33)

connected. A lot of those benefits are coupled and really important. So I like that a lot. I’ll touch on just one of these pieces. So when people hear like more energy density, they may assume the electrolyte itself is storing dramatically more energy. where does Proteus directly improve performance and where is it actually you know kind of enabling a more aggressive chemistry that provides the extra energy?

David Mackanic (16:53)

That’s a good question. So to be clear, the Proteus material itself does not change the energy density of the battery. So we use roughly the same mass as you would with a normal liquid electrolyte, put the same amount of that mass into the cell.

In terms of where it directly improves any battery cell, you put that material into a cell and immediately you will see the improved safety, the improved mechanical strength, the reduced swell. And then depending on the chemistry, you’re going to get all those other advantages, whether it’s notably improved cycle life, better high temperature cycling, higher energy density that system provides. There’s all sorts of other benefits depending on the cell chemistry that our material can impart. But those are…

for a given chemistry might not necessarily be applicable to kind of any sort of chemistry you put this material in because every battery chemistry behaves a little bit differently and so it’ll interact with the electrolyte differently. But we’re ultimately solving these core problems with degradation which manifests themselves in different ways for different chemistries. You know a different example we can talk about other than the silicon anode is there’s high voltage cathode materials that have ability to store more energy for lower cost. But today they suffer from a lot of electrode particle cracking and

gassing issues. So the failure mode there is they start to generate a bunch of gas. If you can stabilize those particles to prevent them from cracking, you can reduce that gas generation inside the cell. So again, kind of the end effect is that you’re solving this chemistry, but the thing that’s actually happening is a little bit different at the atomic level.

Wes Ashworth (18:16)

a good distinction. And you touched on this. I just wanna make sure I clarify, make sure we get it all out there. So with silicon, I think a lot of people have heard about that. It’s been around promising higher energy batteries for years, but to your point earlier, it’s like then swelling and degradation keep getting in the way. what is physically happening inside a silicon rich cell that proteus changes?

David Mackanic (18:36)

Yeah, this is a good question because this is one of the things that during my PhD we really focused on is silicon anode topic has been a problem for some time now.

So a silicon particle, when it’s storing energy inside a battery cell, it can actually store six lithium atoms for every silicon atom. And so when you’re charging that battery up, that silicon particle is increasing in size dramatically, up to three times its original volume. So it’s really growing. And then when you pull the lithium out, it shrinks, but as it shrinks, kind of fractures and cracks and pulverizes itself. And so then you end up with this material in the battery that’s

it’s shattered and that can create all sorts of issues. There’s now these sharp particles, these particles now react with the electrolyte, the liquid electrolyte in the cell. They expand the volume that they take inside the cell. So the cell starts to increase in size and it creates all sorts of issues. With Anthro’s Proteus technology, we’re actually polymerizing this material around each silicon particle. And so while that particle still expands when it’s charging, it now has this elastic rubber like structure, which is Anthro’s Proteus pushing on

it. And so then when you discharge the battery, Proteus pushes it down to compact it so those particles stay much more intact. They don’t have any sort of exposure to the reactive electrolyte because it’s this polymer around it and it makes it so that volume doesn’t increase as much. And so you can’t prevent that silicon from expanding and contracting, but you can make it happen in a much more controllable fashion, which then results in the extended cycle life, the reduced swell and the kind of less degree of reactivity inside.

that silicon anode. And so, you know, there’s other approaches to enabling silicon anodes, but they typically require, hey, we’re only going to use a certain amount of silicon, or we’re going to use some kind of complex geometry in the electrode or in the cell to constrain it, whereas we’re using the electrolyte itself to provide that constraining source at the particle level.

Wes Ashworth (20:30)

Absolutely. I appreciate going into that a little bit further and further clarifying. And I just think that mechanical side of electrochemistry is really fascinating because most of us instinctively think of batteries as just a purely a chemistry problem. But you know, it’s a good reminder that materials can physically move, crack and lose contact, even when the underlying chemistry looks promising. So how that all comes into play.

David Mackanic (20:49)

There’s a lot going

on inside a battery, that’s for sure, and a lot of it that

David Mackanic (20:52)

is invisible to the end user, but those trade-offs are not invisible to the end user, right? So we’re doing a lot to make those trade-offs not come to the surface, but with another technology, we think we can expand the realm of what’s possible.

Wes Ashworth (21:05)

And kind of speaking of all that goes into the batteries, the battery industry is kind of just full of this absolute safety language, you know. giving the more useful answer, what risks does a solidified, non-flammable electrolyte materially reduce and what failure modes can it never make disappear? Like what still exists?

David Mackanic (21:24)

This is also a great question.

you know, no battery of any type will ever be completely totally safe, right? This thing is storing energy and it’s, you know, if it releases that energy in uncontrolled fashion, there could be a safety incident. But what we think about and where, where Anthro Proteus provides an advantage, we think about this in two, two kind of categories. So safety, can have the severity of the safety incident and then the frequency of these safety incidents. that can, you know, frequency times severity can kind of provide some sort of safety factor. And when you think about the severity, Anthro Proteus

is non-flammable. And so we’re changing this flammable liquid electrolyte out for a material that’s non-flammable, that’s significantly less volatile, that’s, you know, just much less reactive than the conventional liquid electrolyte. So if there is sort of an overheating or thermal runaway event, the amount of energy that’s released is reduced and the time scale of which this energy is released is also reduced. And so the severity of that event is significantly decreased. Where we also really shine is on the frequency

of these thermal runaway events. And so these can be induced by battery short circuiting where there’s an internal contact between electrode materials in the cell. Using Anthro Proteus, we’re providing this mechanically strong structure, which helps prevent those shorts from happening in the first place. So whether it’s something like an impact test or a crush test or a test where the battery gets physically bent or penetrated, we’re less likely to have those short circuits. And then similarly, our batteries are much more temperature tolerant. So they’re much less likely to go into thermal runaway.

when they are overheated. And so this material can improve both the frequency and severity factors that lead to battery safety, helping all chemistries exhibit a higher safety profile, even though, you know, for any battery of any chemistry, for any type, there’s always going to be some risk of safety incidents happening.

Wes Ashworth (23:11)

I love that additional context. I think it’s a much more credible way to think about safety than pretending any, you know, high energy system becomes risk free. That’s always there, but we’re talking about, you know, safer and acceptable and it’s always interesting, those safety arguments. People get up in arms and they’re like, I can name probably ten things that are around you that are less safe than what you’re having an issue about. So I always love getting additional context out there

David Mackanic (23:33)

it’s a real problem, right? When you go

on an airplane and they’re like, hey, if you drop your phone, don’t try to go get it. Because if you bend that battery and it goes into thermal runaway, it’s big problem. There’s actually about 60 open flames on aircrafts in the United States every year that come from battery fires. And so this is an issue that happens even with something as simple as a phone. But

I agree with you, we need to define what the safety features should be. This is where the industry could really continue to progress and we’re excited to see more kind of safety certifications and safety standards be put forward for what needs to happen in these tests. And then technologies like Anthro Proteus and to your point system level designs can help make sure that we have kind of the right degree of safety for a given application.

Wes Ashworth (24:18)

I’ll move to a different piece. I’m sure people are starting to wonder now too. So you’ve been clear that Proteus is more expensive than a conventional electrolyte. when does paying more for the material create kind of a cheaper, more valuable battery, if we’re looking holistically? And where does that argument still maybe fall short? Like what’s the full context there?

David Mackanic (24:34)

It’s a great question. I should be clear you are correct. Anthro Proteus is more expensive today. We view at

massive scales, you know, once we’re broadly distributed on hundreds of gigawatt hours, there’s no reason that this cost cannot come down to be very cost competitive conventional electrolytes. I imagine we’ll talk about our AI platform in a second, but you could also imagine us working to optimize cost of electrolytes using that platform. So there’s no fundamental reason why Proteus is more expensive than conventional.

battery electrolytes. That said, we are targeting markets where we are adding a lot of value. We are in the production ramp phase of the company, and so we are intending to charge more for this electrolyte to a modest extent. And the way we think about this is the electrolyte is a really great lever to

add performance without adding significant cost to the battery. And so the electrolyte by rough rule of thumb makes up between five and 10 % of the cost of a battery cell. So let’s call it 10%. If you can change that electrolyte and pay 10 or 20 or 30 % more for the electrolyte, but that gives you 10, 20, 30 % more energy density or 50 % or 2X longer cycle life or the ability to pass a safety certification that you otherwise couldn’t, that is extremely valuable for the end

application and it’s creating a, you know, one to 3 % increase in costs at the cell level. And so we’re actually seeing the electrolyte as a really exciting opportunity to deliver a product that might be on its surface more expensive, but actually drives outsize value for the end application. And that’s only possible because our technology is completely dropped into the existing manufacturing process. So we do not view the cost of the electrolyte as a gating factor given that we can realize kind of these

performance benefits in the end application that make it obvious that this thing will pay for itself. It becomes a little bit more challenging if you think about the more expensive components in the battery changing those out. Of course then there’s an even higher burden of proof on kind of cost increase on some of those.

Wes Ashworth (26:37)

Yeah, it’s interesting, obviously, and focusing on that value. Obviously, you get a lot more out of it long term, and that’s easily justifiable. And to your point, it’s more expensive today. That’ll keep coming down as well, too. Performance going up. that’s a you know winning, winning solution there. So I’ll switch a little bit, you know, the over kind of like focusing on a little bit on AI, we got a touch there. Obviously, it’s a hot topic always, but that value proposition gets especially interesting when the customer’s not, you know, simply trying to

Build another EV factory, but a machine that has to carry its own intelligence and its own energy and thinking about that. And I know you’re starting with defense, you know, robotics, autonomous systems and other demanding markets rather than waiting for e you know years for automotive adoption. what do those customers need from a battery that the mainstream battery market has not yet been kind of optimized to deliver?

David Mackanic (27:24)

Yeah, there’s a lot of interesting factors at play here on this industry, this physical AI, this need for edge compute, this emerging robotics and autonomous systems is quite exciting. There’s a performance piece which is, you know,

manifesting a lot of the things that high-end EVs would want, but in sort of a different form factor and not quite as large of a scale. And so what these applications really need are long runtime. So you want these batteries to be really high energy dense. And so you can do either lot of compute or a long time mission duration. You want these batteries to cycle for a really long time because you don’t want to have to go and replace batteries in these applications. They’re going to be cycling a lot more given the compute workloads and the autonomous workloads. And you also can think about

kind of safety, lights out operations, structural integrity, things where as you’re putting more energy, whether it’s a consumer electronic device is doing more compute or robot, you want this to be extremely safe to be able to operate without, you human intervention or, or harming any humans, whether it’s in a workplace or just something that you’re carrying around. And so these demands are really pushing the limits of battery technology as well, because the form factors are much smaller, typically than what you’d see in kind of an EV.

or large-scale mobility application. So you’re getting a lot of those same power demands in a small form factor. so this is sort of a…

unique intercept between it’s not quite an EV battery scaling down, it’s not quite a consumer electronics battery scaling up, it’s somewhere in between. So that’s one thing. The other thing that’s really interesting about a lot of these applications is there’s increased emphasis on the supply chain and on the country of origin of these battery cells. And so for a lot of these applications that are dealing with intelligence, that are dealing with critical security applications, there is a demand for having these batteries produced domestically in whatever country they’re being used in. And that’s something that

definitely an interesting place to be in the United States battery manufacturing ecosystem right now. We’re seeing this emergence of these applications that not only have high performance requirements but also do have these strict domestic sourcing requirements or desires.

Wes Ashworth (29:28)

It makes a lot of sense. Very smart intelligent strategy and kind of where to start and how why those applications make so much sense. and we’ll we hit on this, but everybody is talking about physical AI right now. Like what breaks if intelligence keeps getting dramatically better, but the energy system inside the robot does not?

David Mackanic (29:44)

Yeah, I mean, it’s a good question. know, there, again, we’re here talking about kind of the smaller scale intelligence, right? But there’s a sort of direct analogy to what’s going on at larger scale AI infrastructure. Right now, there’s a lot of talk about compute being limited. You want to have more of these servers, more of these processors.

But at some point, those are going to become limited by the amount of energy that’s being produced, the amount of energy that can be fed into these data centers. And so there’s going to be constant desire for more of both of these. It’s kind of the same thing that’s happening at the smaller scale. So you could imagine having a really complex, really energy intensive algorithm on your smartphone. But if every time you talk to Siri, it takes 10 % of your battery, that’s not useful. And so you’re fundamentally limited by how much intelligence you can put on any of these applications, by how much power that consumes.

And of course, you also need the power to move and pick things up and take pictures and do whatever else you need the power for.

there is this intrinsic trade off between how much intelligence can be on these devices and how much power they’re going to be able to consume. And so this really is sort of an arms race in terms of the ability to put power on these devices. Right. If you have a robot, you can’t have it switch between dumb low power mode and smart mode. It needs to be able to interpret its environment in real time and respond accordingly. Right. Doesn’t know when it needs to be smart. And so again, I think there’s a great analogy between, hey, right now we might not be explicitly limited by

batteries in terms of the compute, but there’s always going to be more need for compute on these devices, which is going to intrinsically require bigger batteries. And so that can always help improve the performance of these applications.

Wes Ashworth (31:19)

Yeah, it’s super interesting. And it, you know, connects the two technology curves we normally talk about completely separately. And we spend so much time talking about intelligence that’s easy to forget. Every physical action still has an energy cost, so it’s always great to make that connection back. and you’ve talked about batteries eventually taking these kind of like radically different shapes and even becoming structural, and that’s a really interesting thought. what’s the strangest or kind of most surprising product request that made you realize the battery itself could stop being a box inside the device and maybe take on some

Different form factors.

David Mackanic (31:48)

Yeah, we’ve had a lot of these interesting requests, just given the unique mechanical properties our batteries have and given that people are just really trying to put energy everywhere. You know, when we were coming out of Stanford, we’ve got a lot of requests for integration into wearables and kind of these unique shapes. And there were some interesting applications, but it always kind of made sense. There was an aha moment for us early in Anthro where we were talking about our technology and this automotive customer said, I want to make a doorframe or a bumper out of this battery and have it be actually a structural

component because we can pass this certain test then it allows us to you put more energy into this battery and make it less you know kind of basically make a lighter pack design and so Just seeing all the ideas that people had which I thought were kind of I don’t know about too far-fetched But we’re a little bit beyond my comfort zone knowing that kind of pack design of EVs was really interesting and so seeing this pull for structural Integrity was a pretty interesting we had a different customer pull us in to actually embed batteries into the wings of this drone

which was also a pretty cool project. So we actually kind of laid them up in this curved fashion. And so it was just fun to see people getting really creative with these ideas of batteries that can bear structural load, that don’t need to be kind of baby and protected because of safety issues and seeing where people have kind of demonstrated that. We’ve actually done some cool demonstration projects showing these unique applications of structural batteries. Still a little bit kind of in development phase compared to our electrolytes being deployed on existing manufacturing lines, but a really exciting opportunity for the future.

Wes Ashworth (33:14)

Yeah, it’s so cool. And I think it’s where this gets really imaginative because the battery starts influencing the architecture of the product itself. And you know, once the energy sword can participate in the structure, the designers suddenly get some freedoms they didn’t have before. So it gets really interesting to see that and they the appetite for that out there as well, too. And what’s some of those requests are coming to you.

David Mackanic (33:32)

That’s right, and that’s never going to be possible with conventional electrolyte material,

so it’s something that we’re definitely excited about as well.

Wes Ashworth (33:38)

absolutely. And you touched on this kind of like mentioned it earlier. I want to get into it, but just kind of AI at the level of molecules. And I know Anthro is also trying to change the speed at which these materials get invented, and which takes us from the kind of physical battery into the discovery engine behind it. So, you know, everybody has an AI story now, and yours goes all the way down to molecular design. And guess give me one thing, Atlas, and you can tell us a bit about that, just can discover or predict that maybe a great team of

chemists could not, you know, practically find the right brute force experimentation.

David Mackanic (34:12)

Yeah, that’s a great question. So we’ve developed a technology called Anthro Atlas, which is our in-house AI materials discovery platform. And what this platform essentially enables us to do is generate a database of thousands of chemical precursors, use those to describe trillions of different molecular combinations for our electrolyte formulations, predict all sorts of properties of these electrolyte formulations that we can validate experimentally, and then take those electrolyte properties

apply them to a physical cell model so that we can then model from the molecular level how will these different formulations influence the end cell design. This is a really awesome platform because often it takes months or weeks in order to go from a new formulation to actually tested battery data. This allows us to not only predict different formulations from a design space that’s again has trillions of combinations to something where we can get kind of early evidence of life or function before actually step

foot into the lab. So it’s been great at really accelerating our development time and also enabling us to expand our design space of new molecules quickly because we can learn from experimental data that feeds back into this model.

Wes Ashworth (35:20)

Yeah, it’s so cool and interesting. And I know Anthro talks about nine times faster development and greater than ninety percent predictive accuracy. tell us a bit more about that. Like what is exactly being accelerated? What counts as an accurate prediction and anywhere the model still gets humbled by the real world? Like where’s where does it stand today?

David Mackanic (35:37)

Yeah, the

You know, 90 % or nine times faster development, it’s an interesting metric because there’s some degree of, if you can do this in zero tries, again, like I mentioned, without going into the lab at all, and the first set of molecules you test are good, that’s, I don’t know, 100 % faster. I don’t know what the metric is, but we’ve

had examples where we can really predict new high-functioning formulations just from our simulation data. The things that we’re able to experimentally validate in terms of, let’s go predict

a few dozen and then let’s go measure them in the lab. There’s some fundamental material properties that our electrolytes have in terms of their conductivity, in terms of their mechanical strength, in terms of their various other transport properties that we are really great at predicting over 90 % accuracy. The model works quite well based on the way we’ve parameterized it. There are still some kind of longer term performance profiles of cells that we are working on refining, which we do a

quite good job at, but probably are not in this 90 % accuracy threshold. So things like high temperature cycle life or the exact swelling characteristic of a battery were directionally correct in all of these things right now, but predicting cycle life to within 10 % is still a challenge across the whole industry.

Wes Ashworth (36:50)

Yeah, it’s

it’s just all so cool. And you know, if anybody’s head’s going here, I’ll ask this, so the computer can, you know, maybe recommend the perfect molecule, but there’s also factors like what if it’s toxic or impossible to synthesize or too expensive or impossible to manufacture? You know, it could be viewed as worthless. Like how do you close that loop and balance that between sort of AI prediction, chemistry and industrial reality?

David Mackanic (37:11)

It’s a great question. And this is a big kind of design choice in any accelerated materials discovery platform. We have made the decision only to include a subset of molecules that we believe are commercially viable in this training platform. So we’re not actually predicting new chemical strings at this point. We take a database of molecules that we believe are functional. We parameterize them in a way that a computer can understand. We’re not having the computer rearrange these atoms yet. We do want to do that down the line. We already

again, like I mentioned, have thousands of compounds, trillions of different formulation combinations. So there’s no reason to kind of start predicting new molecules. That said, we hope to get to that degree of sophistication down the line. And that goes back to what we talked about previously in terms of cost, you can actually imagine having this system formulate for cost or formulate for earth abundant materials or formulate for a specific business or economic profile you want rather than just a performance profile.

Wes Ashworth (38:04)

great context, super interesting. And I’ll as we get a little closer to time, I want to spend some time just talking about that sort of breakthrough to becoming, you know, industry and the company is really crossed from discovery into production. And I think that fundamentally changes the challenge. So you’re no longer just proving chemistry, you’re becoming a manufacturer. what did scaling Anthro teach you that no academic paper could have taught you?

David Mackanic (38:26)

I think, you know…

When you’re in your PhD or doing research, you were talking about the scientific method, which has its own rigors. But the rigors are, in my experience, nothing compared to running a kind of manufacturing scaling operation. And so this is definitely much more of a measure twice, cut once approach than some chemistry R&D where you can say, I’m just going to go try something and see what happens. Right. And so that’s been a learning is just really making sure we have good process control, really good quality management systems, really good income and quality control. control

on all of our materials, I mentioned, working with different vendors and suppliers and making sure they’re reliable. These are all sorts of things that I don’t think I fully appreciated when we were coming out of Stanford five years ago, but have certainly gotten much more detailed appreciation for now is just how much good process and good quality processes can result in either good or bad manufacturing and scaling outcomes.

Wes Ashworth (39:17)

And as you’ve gone, I know you’re building much more serious US electrolyte manufacturing capability while partnering downstream on cell production. What has to go right between today’s production and true industrial scale before you can say the technology is no longer the experiment, the factory is and it’s happened.

David Mackanic (39:35)

Yeah, I mean…

We have big ambitions and so we’re excited that we’ve taken this first step of scale up. We do have a production facility here in Alameda, California, where we are selling products from and do have relatively small volume production here. We’ve announced that we are building a facility for large scale electrolyte production in Kentucky. And this will really be kind of the gigawatt hour scale production of these battery materials. And so there’s some scaling opportunities there where we have to work with larger

reactor sizes, we have to work with larger quantities of incoming materials, we have to be really conscientious of inline process control and inline monitoring. But it’s, you know, at the beginning, a 2x scale up from our reactor size here. So it’s a pretty measured approach to eventually scaling up to, you know, a factor of 50 from what we’re doing here. But starting with kind of this, you know, go first factor of two, then factor of 10, then factor of 50, so on and so forth. And so we’ve got a pretty good path forward. But you know, that’s also

just the start, right? So then there’s eventually beyond Kentucky, there’s, hey, we want to have manufacturing of this either ourselves or with partners across the world, where we’re producing this at much larger scales, know, hundreds of gigawatt hours across the world. So, you know, excited for that scale up journey, we think we have a really great approach, we’re grateful for the support of the DOE, which is allowing us to take this next step into large scale production. But really, from there, it’s going to be ability to copy and paste across the world, which will measure the extensibility of the technology.

Wes Ashworth (40:59)

Yeah, I love that. I love seeing companies kind of go through that growth and scale and it’s exciting, you know, and seeing what’s ahead and that ability to make it kind of repeatedly economically and at serious volume and what that’s gonna happen. So I want to zoom out a little bit and kind of conclude in terms of just thinking about what batteries become next. And finish this by really widening that lens because the bigger story is how differently we may think about batteries 10 years from now. and a lot of battery conversations still sound like

You know, there will eventually be one winning chemistry. but what does the industry believe about the future batteries that you think is maybe wrong?

David Mackanic (41:32)

Yeah, I mean, I of course do not believe in the one winning chemistry view. And that’s why I’m so excited to work on the electrolyte. We don’t believe that the way to success is building kind of one single material and having it be static and exist forever. You know, we talked about compute and semiconductors earlier, right? There’s never been a time where we were like, yep, the transistor is good enough, right? We’ve always wanted smaller transistors. We’re always going to want better batteries. No one’s ever going to say the battery is good enough. And so what

we are really excited about is technologies like Anthro and other technologies that are platform technologies coming online to redefine battery performance possibilities. But then the existence of our electrolyte can enable people to start thinking differently about both battery chemistry and battery design. so getting these technologies to market is really important, not only for the immediate impact that they have, but for the designs and capabilities that they’ll unlock. So we may be thinking about battery chemistry totally different 10 years from now than we’re thinking. thinking

about it today. Again, we talked about 2015 when I started my PhD, LFP would have been considered not viable back then, right? And look at how big it is today. We think there’s going to be continued evolutions in this space enabled by things like Anthro Proteus, which don’t exist right now, that unlock the future of lighter, longer lasting, lower cost batteries at massive scale.

Wes Ashworth (42:49)

Yeah, I think a much richer way to think about the market overall. let me ask you this, from your opinion perspective, ten years from now, what will surprise us most about batteries and what becomes possible when the battery stops being the limiting constraint in the product?

David Mackanic (43:02)

You know, I think the…

real goal in 10 years is that people will not think about the battery at all, right? That’s the ultimate dream. And you might not think about your battery every day right now, but you probably do if your phone starts to run out, your laptop starts to run out, you start to plug your car in, both on the chemistry side, on the system side, and then on the infrastructure side in terms of being able to recharge these batteries, we imagine that it’s just going to totally fade into the background, there’s going to be electricity everywhere, and it’s not going to be something that even crosses people’s mind of this idea of my battery running out of capacity. So that’s the

ultimate goal to get to and we’re excited to be a part of the technology stack that makes that possible.

Wes Ashworth (43:39)

Yeah, it’s exciting to think about. I like that sort of like the thought shouldn’t be about the battery. It should be kind of the behind the scenes kind of like it’s working, no longer that huge limiting factor or whatever it is, but I love that. final question, I’ll just make a really an open one. Anything else you would want to share, final words of wisdom, advice, things you would want us to know about the company.

David Mackanic (43:56)

No, Wes, this has been a great conversation. I hope that we were able to educate people a little bit about what’s going on inside the battery, why battery materials are so important. And, you know, I think I’d just say one thing that I’m really proud of that we haven’t touched on is that, Anthro is really focused on both developing and scaling this technology here in the United States. And so we’re really excited for this kind of reemergence of the domestic supply chain for these critical energy materials. We’re excited to be part of that. And, you know, we want to be a provider of these advanced materials

domestically allowing us to not only be a technology leader but also a manufacturing leader. So that’s one final point that I’d like to get out there as part of our conversation.

Wes Ashworth (44:34)

Yeah, love that. Great, great way to wrap it up as well. And I know we’ve touched on a lot of kind of things at a high level. I would encourage people to go check the show notes, check the links, or check out Anthro, read more into it, be curious and ask more questions. But David, this was truly fantastic. I really appreciate the way you brought us all the way from what’s happening at the molecular level to the manufacturing at scale and ultimately, you know, what better batteries could make possible in the world around us. For everyone listening out there, thank you for spending your time with us today.

If you enjoyed this conversation, make sure you subscribe to Green Giants wherever you listen so you don’t miss an episode. And if the show is valuable to you, I’d really appreciate it if you would take a moment, rate the podcast, share this episode with someone in your network. Thanks for listening, and we will see you next time.