For years, perovskite solar cells have been the Next Big Thing in solar. In this episode, Joel Jean, co-founder and CEO of Swift Solar, explains what exactly perovskites are, the performance and weight advantages they promise, the challenges they face, and when they might actually reach market.
Text transcript:
For years, perovskite solar cells have been the Next Big Thing in solar. Several hype cycles in, however, there are still no perovskite products on the market. That’s been my excuse for procrastinating on learning exactly what the heck “perovskite” means.
These days, though, just about every big player in solar is chasing them — the biggest US solar company, First Solar, bought a perovskite startup last year — and several are promising that commercialization is imminent. So I guess it’s time to figure this stuff out!
Long story short, perovskite semiconductors have a different crystalline structure than the silicon variety, and can be designed to capture parts of the spectrum of sunlight that silicon doesn’t absorb, more efficiently than silicon. Perovskites can also be stacked atop silicon in “tandem” cells, which promise to substantially increase efficiency for fairly cheap. (Perovskite materials are low-cost and abundant.)
To get into the details, I’m talking to Joel Jean, co-founder and CEO of Swift Solar, a US startup working on commercializing perovskites. Jean’s been a solar researcher for a decade and started Swift to unlock what he sees as solar’s full potential. We’re going to talk about what exactly perovskites are, how to stack them, the performance and weight advantages they promise, how to overcome the stability challenge, and when they are finally going to reach market.
All right then, with no further ado, Joel Jean of Swift Solar, welcome to Volts. Thank you so much for coming.
Joel Jean
I really appreciate the opportunity.
David Roberts
So, you know, I've been hearing about perovskite panels for years and years now. I feel like this hype cycle has sort of come and died down and come back and died down a couple of times.
Joel Jean
A few times, yeah.
David Roberts
So, I'm really hoping you can convince me that this is the real time, that we're actually going to see some of these things soon. So, let's start at the beginning. I suspect most people are as technically ignorant as I am about the nuts and bolts of how solar panels work. So, just to sort of frame this, as I've been reading about it, what I've picked up is, silicon, crystalline silicon, the standard material for solar panels, has a certain crystalline structure that pulls photons out of certain wavelengths of light. And the significance of perovskites is that they have a different crystalline structure.
So, it's not that perovskite itself is a particular material. Perovskite is a term for the crystalline structure. Is that accurate?
Joel Jean
That's very accurate.
David Roberts
Yeah, I got there.
Joel Jean
That's accurate. Yep. So, silicon has a cubic crystal structure, right? So, it's an example of that kind of crystal structure. Perovskites are a new kind of semiconductor material that forms in a different crystal structure. So, perovskite is a very general term for this ABX3 crystal structure, naturally occurring as calcium titanate. But no one's making solar cells out of that mineral. We actually make a synthetic perovskite where you basically can swap in different elements or different molecules on those different sites, those A, B, and X sites. And that includes some very cheap and abundant materials on those different ionic sites.
And that's what allows you to make a perovskite solar cell.
David Roberts
Right. So one of the questions I had is, could you make perovskites out of just any material, or are you restricted somewhat? Is there a certain set of materials, or is it literally just the structure that matters? And, like, in theory, any material could be squeezed into that structure.
Joel Jean
Somewhere in between there. So, material scientists will probably tell you much more about this. I think you can think of it as like: there's a certain class of materials. If you look at the periodic table, there are certain elements, certain families that can go into each of those sites, right? So, for your A, B, and X sites, you need them to have different kinds of ionic charges. So, you need that to balance out to make the material. So, you can think of it as like, you basically have a few different options for the A site, a few different options for the B site.
Usually, two different options. The B site is a metal, and the X site is pretty much always a halide. So, when we talk about perovskites for solar cells, we're referring to, usually, always a metal halide perovskite. That's the metal at the B, halide at the X, and then you can swap out the A site to tune the properties of the material.
David Roberts
So when you're swapping out different materials, the effect is that you capture somewhat different parts of the wavelength of light, is that the idea? So if you tweak the materials, you can capture different swaths of the spectrum, is that right?
Joel Jean
Yeah, yeah. So you're getting to kind of the important properties of this material. So, absolutely. By tuning mostly on the A and the X site, by changing out those materials, you can change what we call the "band gap," so you can change the range of the solar spectrum, you know, wavelengths of light that you can absorb and absorb efficiently. So that band gap is a fundamental property of any semiconductor. It's the defining property of a semiconductor. And every solar cell is made out of a semiconductor. Right. So we really use that to make what we call a tandem solar cell, and that's what makes perovskite so compelling.
David Roberts
Yeah, we're going to get to tandems. We're going to get to tandems in a second. But for the crystalline structure of silicon, is that similarly flexible? Like, can you tweak the materials involved in that, too, to get different, or is that constrained to just kind of one material and one structure?
Joel Jean
The latter. The short answer is no. You can't just change out things in silicon. You can do something called doping. And we do that to make a silicon solar cell, but it's nowhere near the level of tweaking. You're changing out one in a million atoms versus, you know, one in every three, right.
David Roberts
Got it. So, there's a particular part of the bandwidth of light that crystalline silicon can get, and it's pretty much stuck getting that swath, whereas perovskites, you can deliberately construct to get different parts of the light wavelength. Is that roughly correct? And if that's so, are you taking advantage of materials science? I just interviewed someone on the pod the other day who's using nanoscience and AI to test thousands of new materials at a time for their properties, and it sort of made me wonder, is something similar going on here? Are you trying out a bunch of different kinds of materials in these spots, or have you kind of settled on the best ones at this point?
Joel Jean
Yeah, that's a great question. So, your first question, that's roughly correct, that silicon and perovskites absorb different parts of the spectrum. The more important thing, though, is silicon actually has a broader spectrum that it absorbs, but it can't cover that whole spectrum very efficiently. So, perovskites actually absorb a smaller subset of that spectrum, but it does it more efficiently.
David Roberts
Got it. Interesting.
Joel Jean
And that's really the key. So, we'll get to tandems, like you said, in a bit, but that's the key. So, just that nuance.
David Roberts
I'm curious about the material science involved here. Is there a lot of work switching out different kinds of materials still, or have you sort of, like, converged on a small set of the best materials? I'm just wondering how big a part material science is playing in this right now.
Joel Jean
Yeah, so over the last 15 years, since perovskites were first developed in 2009, we've seen a tremendous number of materials that have been tried by all kinds of academic labs all over the world, thousands of papers being published on different materials, on different device structures. So at this point, we're no longer doing in industry right, as a company commercializing this technology, we're no longer going out there and trying thousands of different compositions and types of perovskites. But there are academic labs doing that. What we do now is figure out how to take the perovskites that are kind of best in class that we've developed over the years and make them more stable, more efficient, and more manufacturable.
David Roberts
Got it, got it. Okay. So, if you got perovskites that are capturing a somewhat narrower band, but more efficiently, the obvious thing that comes to mind to do here is to put a layer of perovskites on top of a layer of silicon. That way, the perovskites are absorbing their part of the bandwidth, and the silicon is absorbing its part of the bandwidth. And thus, you are, in net, capturing a much wider part of the spectrum. You're capturing much more, and that is the tandem cells. So, just a tandem solar cell is just two layers of different material on top of one another. Right. Is it that simple?
Joel Jean
Exactly. Brilliant. You got it. You're a physicist.
David Roberts
Bite your tongue. So, a couple of questions about that, beginning with, why use silicon at all? Why not just custom design perovskite one and custom design perovskite two and use two layers of different kinds of perovskites on your tandem cell? Why is silicon still playing a role here?
Joel Jean
Yeah, great question. Maybe I'll take a step back and kind of reiterate what you said to lead into that.
David Roberts
Sure.
Joel Jean
So when you're making a tandem, right, you stack those two solar cells just as you said, and generally, you choose the top one to absorb the high energy light. So that's the visible and the UV light, and transmit the low energy light into the bottom cell, which then converts that part of the spectrum with a higher efficiency. Right. So you have the top one absorbing high energy, converting that efficiently, and then you have the bottom cell absorbing low energy light and converting that efficiently. And silicon turns out to be really, really good at absorbing that low energy light and converting that.
So, if you kind of take that broader view of this entire technology landscape, there are sort of two classes of solar cells, right? There are wafer-based technologies, and there are thin films. And over the years, right, silicon has become the very best of the wafer-based technologies, and that's why it's over 95, maybe 97% of the global market today.
David Roberts
Yeah, there's barely any thin film left. It's just First Solar doing it anymore.
Joel Jean
Exactly.
David Roberts
Just cause wafer-based silicon panels just got ludicrously cheap, I think, is the simple explanation of that.
Joel Jean
Yeah, exactly. With scale, with continued technology development over those decades. Yeah. It's gotten so good. But to take that, like, towards perovskites, I mean, what the perovskites really offer here is a thin film technology that, for the first time, is actually more efficient than silicon, not just cheaper. Right. More efficient. And that gives you this bridge between the dominant wafer-based technology and this really exciting, but always hyped, kind of thin film category. Right. You can make it more a hybrid cell that's actually more efficient than either silicon or perovskite alone. So perovskites make silicon better, and we see it as essentially taking the current industry, leveraging that entire infrastructure that's been built on the backbone of silicon to a whole new S curve.
Right. A whole new, whole new technology.
David Roberts
Okay, so step one here is then just adding a layer of perovskites onto traditional wafer based silicon solar panels. Is that correct? Like, the outcome of that will look like a solar panel we're familiar with. Is that right?
Joel Jean
Exactly, yeah. It'll produce more power. It'll look maybe a slightly darker. Right. It'll look a little bit different, but basically the same.
David Roberts
But, like, form factor-wise, it's going to look like a solar panel.
Joel Jean
That's right.
David Roberts
But there are people doing perovskite on perovskite tandem cells. Are there? Are there not?
Joel Jean
Yeah, there are. Actually, our company. So, my co-founders were the first people in the world to actually demonstrate an efficient perovskite on perovskite tandem, what we call an all perovskite tandem. So, we actually started the company working on that technology.
David Roberts
Oh, interesting. And you switched to perovskite on silicon?
Joel Jean
Yeah, and I'm happy to dive into why that is.
David Roberts
I am curious, was that out of commercial considerations or technological considerations?
Joel Jean
A bit of both. So, to make up all perovskite tandem, kind of to go back to that, the physics of this, you need a bottom cell, this lower energy absorbing cell, to be really, really efficient and stable. And like I mentioned, silicon is really good at this. You know, it's got lifetimes of decades, and it's very, very efficient in that range of infrared wavelengths. And making a perovskite that has what we call a low bandgap, that absorbs that lower energy light very efficiently and can be very, very stable, is incredibly hard.
David Roberts
You're kind of trying to simulate silicon in a way?
Joel Jean
Not just for the hell of it, right? We do it because it allows you to use much less material. It allows you to make something that can be very lightweight, very flexible, and, eventually, ultimately much cheaper. So that's the promise of the all perovskite tandem. But it's hard, really hard to make.
David Roberts
Are there operative perovskite on perovskite tandem panels out in the wild now? It's like, is anybody selling them, or is this just a lab thing so far?
Joel Jean
There aren't any commercial products right now. There are folks who are putting things in the field, doing field testing, and there are starting to be rumors of commercialization, but I wouldn't say there's any mature product out there on the market.
David Roberts
This is maybe speculative, but if you're looking down the road, like, I don't know, 10-20 years, do you think that silicon is going to fade out of the picture eventually? Because perovskites are so much lighter and involve so much less material and they're so much cheaper, that if you can mimic silicon performance with perovskite, ultimately, that seems like it's going to be preferable. Do you think perovskites are going to win out in the end, and silicon is eventually just going to be sort of put in the rearview mirror?
Joel Jean
That's a great question. I hope so. Innovation never really stops. So, I'm not going to try to call out that perovskites are going to be the winner, but I think they are a good candidate. They're the most promising new technology that we have. I think if you look at perovskites in the context of this seven-decade history of solar, we've never seen a technology that could be both more efficient and lower cost, that combination of higher efficiency and lower cost. So, that's what makes me excited and hopeful, optimistic that we have a technology here that could take the whole industry and take humanity to a new era, a new S curve of solar technology.
That's not to say that silicon is going to fade away anytime soon. It's amazing. It's still improving. It's clearly the cheapest today. So, I don't want to say, like, I'm not a silicon naysayer. I think we should be deploying silicon as fast as we can, and we should be working on this next-generation technology to build the future.
David Roberts
You think it's safe to say that the first commercial perovskite panels will be perovskite on silicon tandem? You think that's safe to say?
Joel Jean
I would say the first compelling solar product based on perovskites will be a perovskite on silicon panel. The reason I say that is that you could make a single junction perovskite panel.
David Roberts
Single layer, basically, right?
Joel Jean
Just a single layer. Yeah, exactly. Very much like First Solar's cadmium telluride panels. You could make a perovskite panel, you know, just on a sheet of glass, and use that as a solar panel.
David Roberts
Yeah, but you don't expect that to get to market before perovskite and silicon?
Joel Jean
I don't expect it to be a better product. I expect that people will try to, especially there are folks, companies in China, who are trying to do this as a first step.
David Roberts
Oh, interesting. Why do one layer instead of two? Like, why even try that? Is it just cost-saving or manufacturing efficiency, why not do two?
Joel Jean
Yeah, I mean, making just one is faster, right? You're making one cell instead of two. So, you're not dealing with the challenges of bringing those together into a tandem.
David Roberts
Right.
Joel Jean
So, if you just want to make that one and you can sell that, great. Right. More power to you. I think the challenge here for bringing that to market successfully is that tandems really give you a new step up. They give you a step up in performance. So, it's a superior product. If you're making a perovskite single junction panel, you're competing with cadmium telluride, you're competing with silicon. Your efficiency is not dramatically better than either of those, maybe not even better than the best silicon out there. So, what is your advantage? It's making it cheaper.
David Roberts
Right. So, cheapness is your only advantage in that respect. Okay. The other obvious question that comes up here is, why only two layers? Like, if you can design perovskites to absorb particular parts of the spectrum, why not three layers and get three chunks of the spectrum, or four layers and get four chunks? What's the kind of limiting — why two?
Joel Jean
Yeah, that's a great question. And you're pointing to the decades to come. So, three cells, four cells. Absolutely. That gives you higher efficiency, but that gives you diminishing returns.
David Roberts
This is on someone's radar, this is something that people are working on?
Joel Jean
People are working on these mostly in the academic setting right now, because you get diminishing returns in practice. It's hard to stack more and more and more cells. There is an opportunity there to keep boosting the efficiency up. But if you haven't optimized the two-cell structure, then there's no point, really, in jumping ahead and trying to make three cells before you've maxed out the gains of the two cells.
David Roberts
So, you're saying it's likely that the leap in efficiency you get from one to two, you'll get a smaller leap from two to three, and smaller from three to four, et cetera. Like you're chasing diminishing returns by adding more layers?
Joel Jean
Exactly, yeah. Let's put numbers to it. Right. So, with one cell, you can get up to a theoretical limit of about 30% efficiency. 30%. So, that's what silicon caps out at. That's what cadmium telluride caps out at, anything.
David Roberts
Just to be clear, silicon in practice is in, like in the mid-20s. 30, you mean, 30 is like the theoretical limit?
Joel Jean
Exactly. It's the theoretical ceiling. Thanks for clarifying that.
David Roberts
Yeah, yeah, yeah, yeah.
Joel Jean
That's the thermodynamic limit, it's right around 30%. The same limit for two cells for a tandem is about 45%.
David Roberts
Interesting.
Joel Jean
So about 50% more power relative.
David Roberts
Yeah.
Joel Jean
Then that single layer. Pretty big jump. If you add another cell, so three cells, multi-junction or triple junction, you're around 50%. Right. Just over 50%. So you're not getting a huge jump beyond that. Right. And if you look at a tandem today, the world record's around 34%, so you still got plenty of headroom on the two junction before you want to go chasing the harder technology.
David Roberts
Got it, got it, got it, got it. But, like, someday in our bright future, when all of this is scaled up and it's all super cheap, much cheaper, and the materials have gotten cheaper and manufacturing's gotten cheaper, there could be a three-layer commercial panel at some point. You think so?
Joel Jean
Absolutely, absolutely. And that would be amazing. And I think that's not unprecedented. So, there's actually this core physical concept of making a tandem, or what we call a multi-junction cell, has been used successfully to break through efficiency limits in space solar.
David Roberts
Oh.
Joel Jean
So, sometimes you'll hear about a solar cell record at 38% efficient. Right. Or even 40% plus. You know, that is breaking through these limits. And the way they do that is by painstakingly developing very, very thin layers. Perfect, atomically perfect layers of semiconductor. Many, many of them.
David Roberts
Yeah, space solar is extremely expensive. Expensive is the main thing.
Joel Jean
Yeah, yeah, yeah. So, you're making a ton of layers. So, 3, 4, 5, 6 junctions, and that's how they get to those very high efficiencies. But, they cost hundreds of dollars per watt or thousands.
David Roberts
That's funny. Okay, so my next question was, how much more efficient? But you just answered that. So, like, 30% is the thermodynamic limit on a single junction silicon cell. And after, as you say, I don't know, 70 years or however long we've been working on single junction silicon cells, we've gotten up to — what's the record now? It's like 24% or 25%, am I making that up? Like, what's the like in the field now? What's the best that's been achieved?
Joel Jean
Yeah, it's a little bit different. So, the record in the lab is about 27%. In the field, the best, like, commercial cells that are being made are something like 25% to 26%, and most cells are 24% to 25%.
David Roberts
Interesting. So, we have gotten. I mean, that's pretty remarkable. Like, we've gotten in the ballpark of the limit. We've gotten in the ballpark of about as best we could do on silicon panels.
Joel Jean
It's pretty cool. Yeah, it's come a long way.
David Roberts
So, as you say, 30% is the limit there. 45% is the limit for a two-junction. Is that perovskite on silicon or just any two layers? 45% is the limit?
Joel Jean
Any two layers are about 45% or 46%. Same with perovskite on silicon. They happen to match up quite well to get close to that theoretical limit.
David Roberts
And then, same question, like, what's the best that's — I guess you can't say in the field, since none of these are commercial yet. But in the lab, what's the highest actual efficiency that's been achieved with a double junction?
Joel Jean
The highest efficiency for small cells in the lab is about 34%. So, it's broken through the limit of a single junction silicon cell.
David Roberts
I see. And this is skipping ahead a little bit, but I'm just sort of wondering, like, the first panel you sell on the market commercially, what efficiency do you expect that to have? It will be a tandem. It will be a perovskite on silicon tandem. What efficiency do you expect in your actual manufactured product? Do you know yet?
Joel Jean
Yeah. So, we're not in production yet, but we are moving steadily towards that. For us, we expect the first product to be somewhere above 25%. Right. So, if you look at a commercial panel today, a typical panel is in the 20% to 23% range. What I was mentioning earlier were cell efficiencies. So, these are panel efficiencies. Now, 20% to 23%, the high end might be a 24%, approaching 25%. So, if we're not stepping up above that, we're not making a very compelling product, right.
David Roberts
Yeah.
Joel Jean
So, you can think of that as sort of the minimum bar. And for us, we expect to be in the 27% to 28% efficiency range. And that's where you start to really see compelling advantages over silicon alone.
David Roberts
Interesting. Here's an even more unanswerable question, but how long till you hit 30? You said 45's the limit. 30 would be a very big deal, a serious advance over existing panels. When do you guess you'll be hitting 30?
Joel Jean
So, 30% is sort of a psychological threshold as well. Right.
David Roberts
It's a nice even number.
Joel Jean
It's a nice even number. People are very excited about the 30%. There's certainly no question that if you can make a 34% cell, you can eventually make a 30% panel. Right. The amount of loss you get from a cell to a panel is not so large that it doesn't translate. Now, there are challenges. There are huge challenges in getting this from small to large. So, going from that lab cell, that's 1 cm², you know, hundreds of square centimeters for each cell in production, into a large module. And then also, you have to make it very, very stable. Right. And we'll get to that probably.
David Roberts
Yeah, we're definitely going to get to that.
Joel Jean
But that combination challenge means that the first products are not going to be 34% efficient at the cell level. Right. Even though that's the world record. So 30%, to your question, I think we can expect that in the next, let's say, five to seven years on the market. I don't have a crystal ball, so I can't say for sure, but I think I would say safely in that range, we would expect above 30% products.
David Roberts
Interesting, interesting. So, going from 20 to 23 to 30 in five years, that's just a real big deal. That's not a question, but I'm just reiterating, that's a really big deal. Like, that would be, I feel like if you had a rooftop solar installation, that would almost be enough of a performance boost to make it worth replacing your panels.
Joel Jean
It very well could be. And I'll add also not just rooftop. Right. If you have, let's say, a transmission interconnection on an existing utility scale plant, maybe you're going to go ahead and upgrade that because it's hard to get interconnects. Right?
David Roberts
True. Yeah. Make as much use as possible of the existing interconnects. Just switch out panels for much, much more efficient panels.
Joel Jean
That's right.
David Roberts
Yeah, and like, if you're looking at a giant solar farm, like a 30% boost in the efficiency of a giant solar farm, that's just, that's meaningful. That's financially meaningful. So let's talk about cost then. This is the other thing that's on everyone's mind that everyone wants to know. The cost per watt is kind of the metric people use. People have been talking for ages about getting silicon solar down under a dollar a watt, which I think happened a while back. Like, that's in the rearview mirror now, isn't it?
Joel Jean
Quite a while back.
David Roberts
What are we down to? What's the latest number on silicon cost per watt?
Joel Jean
Yeah, it's down in the sub-twenty cents a watt range.
David Roberts
Holy crap!
Joel Jean
It's come a long way.
David Roberts
Ridiculous. It's so hard to keep track of this. We're down to sub twenty cents a watt. There's no simple answer for how much per watt perovskites are going to cost, since, as we've been discussing, there are different flavors here. There's single layer perovskite, there's tandem perovskite on silicon, there's tandem perovskite on perovskite. There's triple junction someday. But like, for the first, let's just say the first panels you end up selling commercially, a perovskite on silicon tandem cell made into a panel. What's the cost per watt when you start selling it? Are you going to be able to beat silicon only panels in cost, or is it going to be comparable?
Tell us a little bit about how you think about cost.
Joel Jean
How I think about cost is like, there's actually two metrics that really matter. You care about the cost of the module. Absolutely. Dollars per watt. That's what our customers care about. But even more importantly, you know, when you think about solar energy, you care about the cost per kilowatt-hour, right? You care about that end cost of the electricity coming out of your solar panels. So those are two different things. So you could make a free solar panel, but if it's 1% efficient, then you need a whole lot of them to make electricity. So that installation, the racking, the wiring, the permitting, soft cost, all of that balance of system becomes very, very expensive if your efficiency is low.
David Roberts
Yeah.
Joel Jean
Right. So when you have a higher efficiency, on the flip side, you can actually save on those things per watt.
David Roberts
So, you could have a higher cost per watt. But if you are hyper-efficient, you could still come out with a lower cost per kilowatt-hour. Is that right?
Joel Jean
Yeah, that's right. So that's how I think about it, is like you can drive greater value at the module level by having a more efficient panel. That's not to say we can't actually be cheaper. So when you're making a module in terms of dollars per watt, right, if you make a more efficient module, you're getting more watts out. Now, of course, adding some layers, adding this perovskite on top of a silicon cell is going to add costs as well. So your dollars go up as well. So it's a question of do your dollars go up more or do your watts go up more?
David Roberts
Right.
Joel Jean
And there's kind of a techno-economic question there. We've done the cost modeling. We've looked at various different kinds of module formats, and for us, we expect that the cost per watt can be very competitive. Apples to apples, same scale can be very competitive between a perovskite silicon tandem and a silicon single junction panel.
David Roberts
But how? Because the tandem is just the single-junction cell with more stuff on it, with several more steps and more materials. So, how could it not be more cost? Is it just cheap to add that stuff?
Joel Jean
So, it is cheap to add that stuff. Most of the cost of a panel is actually in the glass. It's in the silicon, it's in the junction box and other stuff. But even that aside, right. So, yes, the cost is higher in terms of dollars per, let's say, square meter of panel that we make, but we get more watts out. So, you asked about dollars per watt for the panel. Right. So, that combination is really what matters, and we can get much more watts out, like we talked about, higher efficiency.
David Roberts
Right. So you think when you first start selling panels, you're going to be competitive basically on cost.
Joel Jean
I wouldn't say when we first start selling panels. That's certainly not — that would be naive to think that our first panels, our first factories are going to be cost competitive with ten gigawatt factories out of China. N ot going to happen. Right. I think that's a natural evolution of any manufacturing technology. Right. There's learning as you go. And with scale, the cost comes down. So what I'm talking about is if you're manufacturing silicon at the same scale as you're manufacturing perovskites, you should expect a tandem to have a comparable dollars per watt and a much lower dollars per kilowatt hour because of those savings on the balance of system at the end of the day.
David Roberts
I see. So, these are going to, once scale comes and the predicted costs fall and manufacturing kinks are worked out, etcetera, etcetera, you expect these ultimately to be cheaper in terms of electrical output than their competitors.
Joel Jean
That's right. It would not be worth it if it weren't.
David Roberts
That's a lot of effort to put into it. If it was only the same.
Joel Jean
Yeah, we started Swift Solar, right, to help stop climate change. And even if we're putting gigawatts out there, replacing all the silicon out there, if we're not decarbonizing, if we're not driving carbon emissions reductions on the grid, then we won't have succeeded. Right. So that's like fundamental to our mission as a company, to actually go and drive down the cost of solar, to actually bring the lowest cost solar energy in the world in the history of the world. Right. And that's what perovskite tandems allow us to do.
David Roberts
And so, the reason it's cheap to add a layer of perovskites on top of a layer of silicon is mainly because the materials themselves are cheap. Talk about the material intensity here and maybe the energy intensity of production of these things relative to the production of standard solar panels.
Joel Jean
Yeah, absolutely. So the cost comes down to a few different things. It comes down to the material cost and the cost of the energy and the equipment that goes into this and the facilities and so on. So on both fronts, perovskites, sort of at a fundamental level, can be more attractive and lower cost than silicon, and I'll tell you why. So perovskites actually use 100x less materials than silicon. Silicon is not at the core of a very good material for solar cells. We've engineered the hell out of it. We've made it really, really good.
David Roberts
Is it just a historical quirk that we ended up with? Like, why did it end up being the dominant? Do you know the history there? Is it just the first thing we discovered figured out, or?
Joel Jean
Yeah, well, most people know silicon from transistors, right? From integrated circuits.
David Roberts
Right.
Joel Jean
And that industry has driven technology and society as we know it, right? It's scaled tremendously. Silicon is the semiconductor that we've learned how to control best of anything out there. Almost pretty much any material in human civilization, silicon is the one we can control best, and we understand it. We can model it down to the very last atom.
David Roberts
Not physically the best, right? Not physically the best material?
Joel Jean
For solar, yeah, so you need a ton more silicon. You need 100x more silicon to absorb sunlight than you need of perovskites or of cadmium telluride or other kind of thin film material.
David Roberts
Ah. So you're saving on materials just in terms of bulk. And the materials themselves are cheap. Like, these are engineered materials, these are materials that have been designed to do this. What are they made out of? Like, what is the material? These metal halides, are they easy to find? Are they easy to produce? Are they cheap?
Joel Jean
Yeah, it's a great question. So, the good news is that we are using, you know, very small amounts of these abundant materials that are produced in very high volumes or can be produced in very high volumes. So, even though the supply chain today for these specific perovskite precursors, we call them, like, chemicals that are used to make the perovskite, even though those are, you know, not produced, there's not a very mature supply chain for that. The raw materials that go into that are actually produced in high volumes. They're abundant and they can be sourced from the US and friendly countries.
So the materials themselves can be very, very low cost. And that's why the panel made out of perovskite can also be very low cost.
David Roberts
I see. And is the panel like, is a comparably sized panel made of perovskites lighter, literally, physically lighter than the silicon because of how much less material it uses, or is it mostly the glass that's the weight?
Joel Jean
Yeah, there are two levels of that. So it's lighter on both levels. The perovskite itself as an absorber, as the semiconductor material that's kind of fundamental to the cell, is 100x thinner, and therefore, on that order lighter, that order, like factor order magnitude, kind of lighter than the silicon cell itself. Now, when you put it into a panel or a product, what dominates the weight is not the semiconductor, it's the other pieces of it. So it's things like the glass, it's things like the encapsulant, the steel frame, these other pieces. So what you really want is something that's very efficient, that can then make the weights or the watts per kilogram higher, or the kilograms per watt lower.
Right. There is another aspect to this, which is that you can make perovskites very flexible. So instead of forcing them to live on glass, you can actually make a perovskite on a thin plastic sheet.
David Roberts
You're segueing smoothly into my next question, which was precisely about form factors. So, like, I see that if you're just looking to get a foothold in the market, it seems to me like the smartest thing to do is just to make a tandem solar panel that looks and behaves more or less like traditional solar panels that can slipstream into the existing solar panel, whatever supply chain or, you know, workforce, etcetera, etcetera. But with wafer-style silicon, you have to make a big rigid panel. But as you say, perovskites do not need to be rigid.
You do not need a big, rigid panel. You can make, in fact, thin film out of them, which means you just spray it on like a little flexible piece of plastic. What is the backing for thin film? Is it plastic?
Joel Jean
It could be glass, but when you make it flexible, yes, it's often plastic. Or a metal foil.
David Roberts
Metal foil. Interesting. So, you could make almost any form factor then that you wanted in theory, that's correct. Is any of that on your radar? Like, are you looking into that? What would it be like if you were going to try to commercialize a non-panel form factor? Where would you start? Like, what would you go after first?
Joel Jean
Interesting. Yeah. So, I will say I've spent many years in grad school and as a nerdy grad student, MIT, basically making these little flexible, lightweight panels. And there's a lot of challenges with doing it. Right. It's not just saying, "Let's go make a perovskite solar cell on, you know, on this other surface." It's like, "How do you make it survive many flexes?" Right. How do you make it protected from moisture? The way, like, glass is an incredible barrier to moisture and oxygen.
David Roberts
Yeah.
Joel Jean
How do you make these plastics protect against that? So, it actually turns out that it's actually quite hard and expensive to make these very flexible and lightweight products, even though we've been working on it and talking about this kind of promise in the industry for many, many years.
David Roberts
Yeah, ages.
Joel Jean
Yeah, you don't see a cheaper, dominant, flexible product, even though it's so compelling. Right. It captures your imagination.
David Roberts
Yeah, yeah. Are they in use anywhere? Like, is there any, like, submarket where they are in use? They are being sold, like First Solar is making and selling them, aren't they?
Joel Jean
They are making and selling.
David Roberts
What are they getting used for?
Joel Jean
So, First Solar is actually really good at making traditional looking solar panels using thin films. They're cadmium, telluride, semiconductor on glass. They make a panel that looks very much like a silicon panel. There are smaller startups and companies building flexible plastic-based or metal foil-based thin film solar panels, and they are used in more niche applications, maybe very lightweight rooftops, maybe on boats or on vehicles. I think there's an extra twist here, which is that in a lot of these applications, people are freaking creative and are able to make silicon work for a lot of these applications.
Like, silicon is not meant to be very flexible. But, you know, if you engineer a panel that's kind of more resilient, it can bend, it can curve, and it turns out curving is pretty good, works well enough for a lot of these applications. So, I don't want to poo-poo too much on the flexible or, you know, lightweight applications. I think there's a lot of opportunity there in the long run, but I think silicon is particularly good at getting around these and it is very cheap. So, that's why we think that making a process silicon tandem can still open up these kind of alternative use cases.
And you don't have to just go for the super lightweight, super flexible thing on plastic.
David Roberts
Right. So, you're starting with a panel-looking panel. You're starting with a rigid panel. That's what you're focused on?
Joel Jean
We're starting with a rigid looking panel for traditional solar use cases.
David Roberts
Right.
Joel Jean
But in parallel, we're also working with partners, the folks who are able to make, let's say, a curved panel for a car or a custom panel for a satellite. We're working with the folks who are in those verticals, who are really, really good at making those products, and we can provide them the higher efficiency cell, the core building block that gives them more power out without changing their designs. So we're going both ways at the market, if that makes sense.
David Roberts
So, you can just substitute perovskite cells for silicon cells in those thin, flexible applications.
Joel Jean
Yeah, exactly. And they won't be as thin or as flexible as a true thin film, but we'll get there. We'll get there.
David Roberts
Yeah, I just wonder about all the places. One of the things that used to capture my imagination, but turned out to be practically much more difficult than people thought, was the hype about solar going to be a building material. It was going to be on the roof of the car, it was going to be in fabrics and backpacks. We got these thin, flexible panels that you can put anywhere now. So, like, solar is just going to be integrated into all materials around you all the time.
That has not really come true. That's not really played out. I'm not familiar with a single large-scale commercial product that uses those thin films in those applications. Has that just not happened yet, or is that ultimately a dead end? This is a little off our topic, but I'm just curious.
Joel Jean
Oh, no problem. I agree. At the top level, I agree. The operative word there, I think, is large scale. Large scale solar. We're talking gigawatts of solar, is what you need to get to very low cost, the way silicon has. There's no other market right now that has reached that kind of scale, and therefore the low costs. So it's not that we haven't seen solar backpacks or solar watches or solar flexible panels out there. They are still being used. They're more niche. And I think sometimes — I've been on the wrong side of this as well, sometimes we overhype these things — academics certainly have an incentive to kind of paint the picture of what's possible. I've done that myself and absolutely, like, there are a lot of things that are possible with future solar technology, and I hope that that happens. I hope we can drive that, but the practical reality is that silicon is much, much cheaper at much larger scale today. So we have to work within that.
David Roberts
Well, speaking of working within that, if I have a factory, you know, there are lots of these big gigafactories now making standard silicon solar panels. If I wanted to convert that factory to make perovskite on silicon tandem panels, how big of a deal is that? Am I gutting it and building a new factory? Am I just tweaking the machines? How big of a deal is it to slipstream into existing manufacturing?
Joel Jean
So, we're designing this so that it can, as you say, slipstream into the manufacturing process. The good thing here is that it's very much a complement. It's not like we're trying to replace the silicon. We're actually adding some steps at the back end to make a tandem on top of the silicon.
David Roberts
So, is making and depositing the silicon layer the same, whether you're doing a single layer or a tandem? Like, that first step of doing the first layer is the same regardless.
Joel Jean
It's the same. Yeah. So, we start with the silicon wafer. Right. And we convert that into a cell. And that — I say "we" like collectively, that's what the industry does. Right. We don't make silicon in-house ourselves, but basically, you can convert that to make a single junction cell, or you can omit a few steps at the end and turn it into a tandem, add some different steps.
David Roberts
So, you're just adding a few steps.
Joel Jean
Exactly.
David Roberts
So, you could theoretically convert a gigafactory to make tandem cells without too much disruption, without too much cost. Is that a thing that's actually going to happen, do you think?
Joel Jean
Yeah, so we work with a number of partners, and we're talking to a number of partners to make this perovskite on silicon tandem. Right. And everyone in the industry wants to make perovskites.
David Roberts
It's the next big thing. That much I picked up from reading around. Every company, now at this point, has either started or bought a perovskite operation.
Joel Jean
So, there's a lot of interest in this. And part of what people are doing now, what we're seeing, is folks who are building silicon factories are now leaving space to make the tandem part of the line.
David Roberts
Huh, hilarious. An empty corner.
Joel Jean
Exactly.
David Roberts
Empty corner of the factory.
Joel Jean
Yeah. Leave that for me. Right.
David Roberts
Perovskite. Perovskite ready, as they say.
Joel Jean
Yeah, exactly.
David Roberts
Oh, that's funny. So, this is not going to require building a whole new set or a whole new kind of factory from the ground up or anything like this. This is going to be a modification of existing manufacturing that won't be too disruptive?
Joel Jean
So, it's somewhere in between, because it doesn't have to be disruptive. I will say that because the industry is growing so fast. Right. It's doubling every three years. We're kind of looking at new factories every year. So, it's more likely. Right, like those new factories are the ones that are going to be tandem factories, rather than saying, " Let's go retrofit an old one."
David Roberts
The perovskite panels are higher efficiency and cheaper. So, that hasn't been the barrier. The barrier to commercialization, t he reason they've been taking so long to come to market is resilience, stability. So, perovskite panels traditionally start degrading more quickly than silicon panels, so you lose, like, 8% efficiency a year or something like that. And so, the lifetime of the cell is much shorter. That has been the main barrier keeping these things out of the market, I think it's fair to say. So, tell us where we are on stability. Are we anywhere close to saying, "Here's a tandem panel that will last as long as traditional panels?" Or, like, where are we?
Joel Jean
Yeah, that's a fair intro and fair framing. So, Perovskite's stability, I will say, has improved dramatically in the last five years. So, our cells today are at least ten times more stable than they were three years ago.
David Roberts
Oh, interesting.
Joel Jean
And even the last six months, you know, we've shown that our cells can go for, with no degradation over, like, 3000 hours of testing at high temperatures.
David Roberts
Yeah. What is — How like, I don't even know how to ask this question. Why were they so unstable? And what are you doing to them to make them more stable? Is it, is it about the materials you're choosing or the exact crystalline structure or manufacturing techniques? Like, what is making them more stable?
Joel Jean
Yeah, absolutely. So there are a lot of ways that perovskite cells can degrade. Right. So in contrast to silicon, which is literally a rock.
David Roberts
Yeah.
Joel Jean
You're making these synthetic materials that have a lot of components in them, and they're formed at low temperatures. So there's a lot of ways in which they can degrade. And I'll give you a few examples. Right. So the perovskite structure itself, we talked about ABX3. There's all kinds of different molecules in here. You know, it could be four or five different kinds of materials. And those can segregate, they can, under high temperature, they can move around and they can even react with other parts of the cell.
David Roberts
High temperature, like a hot day in July-type high temperatures?
Joel Jean
Um, like a stovetop or like, uh, a very like, you know, sort of boiling water kind of temperatures. So a little bit hotter. But generally, yeah, those kind of high temperatures combined with light tends to degrade perovskites. And those are two unavoidable things in, you know, for a solar panel, high temperatures and light.
David Roberts
Heat and light are pretty, pretty core to this.
Joel Jean
Yeah. So that's something that we've worked tremendously on and improved tremendous amounts. And the way we approach this is like, do the science before doing the engineering. Right. So we've spent the last decade plus understanding these degradation mechanisms and fixing them. So you really have to address all these issues simultaneously by engineering this device. So things like moisture ingress, things like thermal degradation we talked about. Chemical reactions that can happen, even mechanical breakage. You can have this cell peel off. Right. You can delaminate. And that's something we see when you thermally cycle these cells or panels.
So these kind of issues, they all have to be solved in parallel. And that's why you have to have a very deep understanding of the science to be able to address one without failing on another.
David Roberts
Right.
Joel Jean
And that's what we've been doing. Right. So we're getting closer and closer to commercial ready. And our latest results translate to decades of field lifetime. So, you know, I won't say that it's proven out like, it's for sure that we need to prove it out in the field. But I can say confidently that stability is no longer a fundamental blocker to commercialization.
David Roberts
Well, but there's a lot in that statement. When we say commercialization, I mean, where is the first place these are going to show up in the market? Because you could think of solar power applications that don't need to last 25 to 30 years, like standard solar panels, where perovskites could start first. So I'm curious, like, how you're going to ladder up, like, where you think, like, where, where you are on stability and what types of commercial applications that opens?
Joel Jean
Yeah, that's a great point. And we are going to aim for that 30-year life. Right. We're going to aim to make panels that can be 30-year plus kind of lifetimes. We don't change our technology pathway from going for that, even if we see applications that can use a shorter lifetime. So, for example, a satellite in low earth orbit might be up there for five to seven years. So if you're going after that kind of market, I imagine that's the kind of thing you're thinking about.
David Roberts
Yeah, yeah. Right, right.
Joel Jean
Yes. You can get by with a shorter lifetime. That doesn't mean we're gonna go build a panel that only lasts for five to seven years. It's just that they don't need to see as much proof. They don't need to see five years in the field or two years in the field. They just need the proof from space tests, for example, or the thermal cycling test. That will give them confidence that our thing will last versus saying, "Okay, we need to have many years of field data proven out that this panel is bankable, that banks will finance this thing before we go deploy it," which is the case for utility scale solar.
David Roberts
Right. So is it your commercial plan to start with some applications like that and then ladder up?
Joel Jean
I think ladder up or like stepping stones is sort of one way to look at it. But I do view it as a sort of all at once kind of strategy. And I know that sounds kind of silly. It's not that we want to divide and go in all different directions, but the way I look at it is if we want to help stop climate change, we need to be building something that is relevant, that can actually solve problems and make the cost of solar cheaper on the grid at scale. So we're going to build for that future, we're going to build for that destination.
And along the way we're getting a lot of customer interest. Right. A lot of customers from the space sector, from aerospace, from vehicles, basically are reaching out and saying, "Hey, can you provide us a more efficient solar panel?" And, yeah, we're gonna, we're gonna entertain those. We're gonna talk to them and work with them to make better products because we love all of it. We do have to stay focused on that long term vision, though.
David Roberts
Got it. So solar panels that last as long as today's solar panels is the destination.
Joel Jean
And I'm confident we'll get there. It'll take time.
David Roberts
You'll get there. But, but you're gonna start selling commercially before you get there, though, right?
Joel Jean
Yeah.
David Roberts
Your first commercial products are not gonna be standard panels that go on rooftops because you're not going to be at that 30-year stability mark yet. So I'm wondering what is your first commercial product going to be and who are you going to sell it to?
Joel Jean
Trey, it's a great question. I believe that we will have internal confidence before the market has confidence. We're going to have data before the market has confidence. Your question points to what is the go-to-market strategy? I will say from a historical perspective, every commercial solar technology today, silicon, cadmium telluride, CIGS, they all started out with major reliability issues, lifetimes well below five years. Right. And they, through extensive development, reached a 20 plus year lifetime in warranty. So history is in our favor here. It's going to take time, but we are on that pathway. So I can't tell you exactly where the very first products, our first sales and our first panels are going to go, but I will tell you that it's all those markets that we talked about.
There's compelling avenues to take this perovskite tandem and go serve that market.
David Roberts
When is that? When are you going to sell a product on the market?
Joel Jean
Great question.
David Roberts
This is the $6 million question. When I threw this out on Twitter, "What do you guys want to know about perovskites?" That, you will not be surprised to hear, was the number one question, which is like, "When can I buy one?"
Joel Jean
It depends on who you are. Right?
David Roberts
Right.
Joel Jean
If you're ready to take some risk with us and partner with us to make some early product. Right. If you, you being a manufacturer, if you're an end user who's excited about next-generation technology, you could probably buy one sooner. Right. So in the next two to three years, as we start production, if you're thinking like at large scale, where this is going to be a very compelling economic proposition for your rooftop. Three to five years, right. It's something that is going to take some time to mature. So with the way we see it, it's, we can already start making commercial size prototypes right now and start to deploy those.
And if you want to buy one, if you are the right person, we could entertain that, we can talk about it. But production at scale, I think three to five years is a safe timeline.
David Roberts
Three to five years. And here I'll push you to say something exciting, which is, do you at this point view perovskites taking over as sort of inevitable at this point? Has development come along enough? Like, you've answered enough of the questions, proven this out enough that basically this is at this point going to happen and it's only a matter of exact timing.
Joel Jean
Yeah, I'm biased.
David Roberts
I imagine you are.
Joel Jean
So, starting with that, I think that perovskites are the most exciting new technology in solar that we've ever seen. That we've ever seen, period. It's an opportunity here because we have seven decades of history in solar. We've never seen something that can both outperform and underprice the incumbent. So this is something that is truly something new. And I think we've done enough development that we feel confident in saying that this is going to get out there and it's going to be a superior technology. I think with our latest series A that we announced last month that there's a lot of proof points.
We have private investors really jumping into this, looking at our data. We have the government backing us most recently with about $10 million of grant funding in the last three months. So it's really a lot of momentum, I think, for people who are looking at this data and looking at it.
David Roberts
The government must, I mean, I'm guessing that the US government, part of what they're intrigued by with this is that supply chain question is that you can source your materials and theoretically do your manufacturing in the US. Is that right?
Joel Jean
Yeah, no, absolutely. That's the motivation. Right? Like, it's the opportunity to basically bring solar manufacturing back to the US. There's domestic jobs and energy security, like, parts of that. Right. But, like, we see, we've gotten funding recently from Department of Energy and the CEC. So California Energy Commissioners, thank you to those folks. But, you know, really they see it as like, if we seed manufacturing, if we continue to seed manufacturing.
David Roberts
Yeah, yeah.
Joel Jean
We seed innovation. Right. We see that next generation. And that's why it's so important for all these folks, like, you're saying, to bring this technology back.
David Roberts
Yeah, well, I was going to ask, like, I'm guessing with perovskites being this promising and being this quasi inevitable at this point, everybody, every solar company is looking into them. I'm just guessing that China, being China, is dumping billions of dollars on developing them. Where are we in that race? Like, is this another thing where China is just going to, like, spend out the wazoo and end up dominating, or do we actually have a reasonable chance here to get a foothold in something that China doesn't dominate for once?
Joel Jean
That's a great guess. Yeah.
David Roberts
I mean, who knows? Who knows? But tell me your best guess.
Joel Jean
Yeah. No, well, so it's an entirely new technology. Right? So I think we can win here. I'm confident we can win here. We do need strong government support. Right. So again, like, we've, we've had funding from the government. We've seen subsidies and tariffs and things. So I think that's important for leveling the playing field. So no question, that's key to opening up this window of opportunity. But to your point, like, China has a lot of capital going into solar. What I would say is just, it's not just about scaling. Right. China's really, really freaking good at scaling up fast, putting a capital in scaling up fast, and that does lead to technology improvements.
But in this technology, we still need innovation. There's a ton of upside left, and I'm confident that with our team and our technology, we here in the US can actually compete with the best in the world. So again, it's no guarantee. China's been incredible at scaling up silicon, scaling up EVs and other technologies, but we have an opportunity here to bring that back to the US. Know we're going to grab it.
David Roberts
Yeah. And it's a really big deal. Like, if perovskites are going to take over the market on some timescale, then it really matters who gets there first. Like, it's not a, this is not a niche issue we're talking about here. This is the whole solar market at some point.
Joel Jean
Yeah, yeah. Perovskites and future of solar.
David Roberts
All right, well, my final question, which you already sort of semi answered, but I'll just put it to you straightforwardly. Are there any other innovations in solar that you think might compete with perovskites on any timescale you choose? Like, is there anything on the horizon, any other sort of just science y stuff that you think will compete to be the future of solar? Or do you think perovskites have, like, a very clear lead here?
Joel Jean
I've been a solar researcher for over a decade, right? So I've looked across a lot, I've worked on a lot of different technologies. When I first started in grad school, my first solar cells were 2% to 3% efficient, and that's where most solar cells start, in the lab. And perovskites, when they came in 2012, the first major perovskite solar cell paper is over 10% efficient. Right. So that's just — I mean, that's obviously just a cherry picked example, but I think we talked about this, right? This is truly like in the history of solar, kind of a one of a kind advancement, something that we've never seen before.
So I won't say that there's no other technology, especially ones that are kind of inspired by this sort of material system that can surpass perovskites strictly per se, but I think that we will not see, I have never seen a better opportunity, a better technology that can actually advance this industry and advance the whole world to the lowest cost energy in the history. So, yeah, I'm excited about that.
David Roberts
Awesome well, that seems like a great place to wrap up. Thank you so much. This was incredibly helpful to demystify this whole thing. This has been sort of a fuzzy area in the corner of my eye for a long time, and this is very, very helpful and very exciting. Like, you know, as crazy as solar has been, it's just like, as you say, there's just so much upside ahead. There's so much runway left ahead, just in basic performance and cost and everything. Like, we haven't had really seen anything yet, I think.
Joel Jean
Yeah.
David Roberts
So very exciting. Thank you, Joel. Thank you for your work and for coming on.
Joel Jean
Thank you so much, Dave. Yeah, I appreciate it.
David Roberts
Thank you for listening to Volts. It takes a village to make this podcast work. Shout out, especially to my super producer, Kyle McDonald, who makes my guests and I sound smart every week. And it is all supported entirely by listeners like you. So if you value conversations like this, please consider joining our community of paid subscribers at volts.wtf. Or leaving a nice review or telling a friend about Volts, or all three. Thanks so much, and I'll see you next time.
What's the deal with perovskite solar?