Why solar panels can't be more than 33.7% efficient

There is a strict theoretical limit on how good solar can ever be, but I think this has actually been a blessing in disguise, and it's allowed solar to not only be the most important technology we have for beating climate change, but also literally the cheapest energy that we have. That's why I've partnered with the Charity Sunday, who are promoting awareness about how awesome solar really is. More about them later. But first let's understand exactly how solar works 'cause it's surprisingly straightforward and it's also quite obvious what the limitation is. Fundamentally, all solar energy does is it tries to take light coming from the sun and get these electrons to absorb that light and use it to flow around this circuit. Because flowing electrons are what make electricity, and you can use the energy coming from the sun to these electrons to do things like boil your kettle, make yourself a cup of tea. So fundamentally, that's all that solar is trying to do. So the way that solar panels try to do this is by using a semiconductor. Semiconductors though have really interesting physics because the electrons inside of a semiconductor typically can't actually flow, and the reason for that is because of the energy of the electrons in here. Inside of a semiconductor, electrons can have a whole range of energies called the valence band. So they could have energy up here or down here, or anywhere in between. But the valence band is very constricting because if an electron wanted to flow, it would get in the way of its neighbor. And so electrons don't generally flow in a semiconductor, but they can. If they're given enough energy to jump away from the valence band into the conduction band inside of the conduction band, the electron is very free to flow around, and so the electron would be able to flow around this circuit, give its excess energy from the conduction band to this kettle, and then go back to where it started. So solar energy boils down to being able to move electrons from the valence band of energies to the conduction band of energies. So as long as you can figure out a way to give an electron enough energy to bridge this quantum leap. Then you'll be able to make a working solar panel. But as you can see, sunlight actually comes in many different colors, and that turns out to be a bit of a problem for solar panels. Think of it like this. If all of these different colored lights were shining on these electrons in the valence band, which ones would actually be able to get enough energy to escape to the conduction band? Not all of them. These electrons need this exact amount of energy or more to be able to get into the conduction band, and some of these colors of light won't be able to provide it, and that's because of the quantum mechanics of light. You see, red light and violet light actually have very different amounts of energy per photon. The electron with the red light might only get this much energy, whereas the photon with the violet light might end up getting. That much energy, and this is not an allowed state, an electron can't have an energy in between the valence band and the conduction band, and so this actually just won't happen. So what will actually happen is this red light will just go right through and be used up for something else, whereas this violet light actually did get absorbed usefully by this electron that's now able to complete this circuit. If the band gap is really big, then that means that most of these colors of light won't have enough energy in them to get an electron across. But if the band gap is really small. Then all of these will have enough energy to get electrons across, and so all of those different colors of light can be absorbed. So it all comes down to the size of the band gap, and that is determined by the material that this semiconductor is made out of. Like every type of semiconductor has a unique band cap. So you might think the obvious thing to do is pick a semiconductor with a really tiny band cap, but. There's actually a big trade off. Let's think about this violet light again. So this electron that gets that light will jump all the way to this high energy. Remember, it only needed this much energy to get into the conduction band, but it actually had more energy and so it's ended up at a very high spot inside of the conduction band. Now, that could be useful if we could harness. All of this electron energy, but we actually can't, basically inside of this circuit, you can only harness the

band gap energy and not really any of the extra energy inside of the conduction band. So when this electron goes around the circuit, it will give this much energy to the kettle, but the rest of the energy actually just goes to heating up the wire and the solar panel. And then it ends up back where it started. So in fact, a lot of the violet light was wasted. And the problem gets even worse, the smaller the band gap is because if really small. More of the electrons are going to make it across, but the amount of energy that they're going to be able to usefully use is very small, and the wasted energy per electron is much higher now. So it's a balance. You don't want the band gap to be too small or else each of these colors is only gonna give you a little bit of energy, but you don't want the band gap to be too big because then not many of these colors are gonna be useful. So then there must be an optimal band gap. In other words, an optimal semiconductor material, one that allows us to absorb as many colors as possible, especially the colors that the sun mostly shines in, but it still lets us minimize the amount of loss per color. You can actually do that calculation, and it's called the shockly quasar limit, and it finds that the optimal band gap energy is 1. 34 electron bolts. That probably doesn't mean too much to you, but the important thing is if you actually found a material with that exact band gap, then the theoretical maximum amount of energy that you can get from a sun from a solar panel is 33. 7% efficiency. So the way forward is obvious, right? You've gotta find the right semiconductor material. Yeah. What about silicon? Silicon is an extremely common material for a semiconductor because it's so abundant on earth. In fact, sand is made out of the same silicon crystal structure as this solar cell, so there is at least as much silicon as there are grains of sand on earth. So yeah, that would be kind of ideal. But unfortunately the band gap of silicon is, is kind of close. It's 1. 12 ev. In theory, the highest efficiency that you could get by using silicon is actually 32%. So, you know, kind of close to the limit. But in practice, silicon tends to actually have much lower efficiency than that. Contrast that with gallium arsenide. So Gallium arsenide has a band gap energy of 1. 42 ev. Which leads to an efficiency of about 32. 8%. So yeah, extremely close to the limit. In fact, way back in the 1990s, they were using gallium arsenide solar panels on spacecrafts, and they were reaching efficiencies of like 30%. So yes, very, very close to the theoretical limit. So obviously the way that the industry should have gone is they should have bet hard on gallium arsenide since it almost saturates the shockly quasar limit. But they didn't. They went with silicon, and here's why. That was a stroke of genius. You see, the engineers could have relentlessly focused on efficiency, and if you're gonna do that, you wouldn't pick silicon. Even now, all these years later, it's still only about 25% efficient, but there aren't that many gains to be made by focusing on efficiency. The shockly, Quas are limit means that a really hard stop on how far you can go just doing that. And so instead, engineers decided to focus their energies on something else. And that metric that they decided to look at was cost per unit of energy, and that has made all the difference. Actually, I wanna tell you a funny story about this first. So coincidentally, my dad called me this weekend very excited to tell me that he was updating his solar panels. You see, they had gotten solar panels all the way back in 2011 because in Australia rooftop solar is extremely popular for the obvious reasons. Back in 2011 when they got their solar panels, they only got a fairly small system, which was standard at the time, so 1. 5 kilowatts, but now they're upgrading to 6. 5 kilowatts because that's the new standard. And yet, even though. They're getting about four times more power out of the solar cells. Now they're actually paying less than they did originally. So when you factor in both of those, you find that solar has actually become six times cheaper per

unit of energy for them. And that little anecdote actually matches up pretty well with the overall trend in the solar industry. By the way, my parents also got a battery and my dad was very excited to tell me that they basically wouldn't be paying electricity bills at all anymore. So even though we're talking about price, which you might think doesn't really matter, this price drop has allowed them to make their household electricity consumption. Net zero. That's why this chart is the most exciting story in climate tech right now. It shows how the price per unit of energy has dropped dramatically. For solar panels, for example, the prices has dropped by a factor of 45 since 1990. And even if we compare to a much more recent time, like 2011 when my parents got their solar panels and there was a big push for people to get rooftop solar. In Australia, we find that the price has dropped by a factor of eight, and solar isn't just confined to rooftops anymore. Solar Farms account for about 60% of the market share for solar, and those are even more profitable and it's being installed everywhere because it's just good business. This is why I think solar energy is the most exciting type of energy that we have now, and if you wanna know more about it, then September 21st is Sunday, and there are loads of local events all over the US and around the world, like tours of solar farms. So to find an event near you, check out the link in the description. But let's get back to this chart and why it is that solar panels and in particular silicon solar panels have dropped in price this dramatically. A big part of the answer comes down to the fact that these can be mass produced. There are massive factories, especially in China that make many, many of these at a time. And so they've learned from doing it again and again, exactly how to do it as cheaply and efficiently as possible. And also the demand for solar panels in China especially has been very strong. And so the more that you make a solar panel. learn about how to make it better, and so the more you sell, the cheaper it gets. That is what's so amazing about solar. On the other hand, compare that to something like gas or coal. When you make a gas plant, you have to make something that's fairly bespoke for the area. And so there's a lot of sort of individual design work, and it's a big mega project. And so there isn't as much learning by doing because you're doing it much less often. And that's part of the reason why the price of those things hasn't really dropped over time too much for gas or coal. You need to be mining materials. For every what of power that you use. Contrast that with solar, where yes, you do at some point have to mine for the silicon and the other trace materials in here, but it's only once in the 30 year lifespan of this solar panel instead of every single time that you use that form of energy. And so there's a natural flaw for how cheap coal or gas can ever get, and that flaw is determined by the price of just extracting the materials themselves. Not only that, something like coal has waste all along its process that all pollute the earth and cost money. Whereas with solar, the sun's energy is free and we have been able to make these things cheaper and cheaper because of learning by doing. You might say though, yeah, it's all well and good that solar energy is cheap, but the sun. Isn't always shining, but while researching this video, I found out that there are actually great solutions to this problem. Even the problem of the sun shining much more in summer than in winter, and the solution involves just building more solar panels, which we can do since they're very cheap. And also using batteries. Batteries have also, similarly to Solar gone on their own cost curve. That has meant that batteries are becoming more and more affordable. If you want more information about how solar plus batteries can solve the seasonality problem, there's a really great article from Brian Potter that I will link. There's one last thing I wanna get off my chest. So before researching this video, I would casually watch videos about solar energy and, you know, people would make stuff about like pite solar cells or like multi-layer solar cells, all these exotic designs. And I figured that, you know, that was the frontier of solar and that's where all the important work was happening. In fact, if you layer different types of solar cells on top of each other, the new limit is actually 60. 8. 7% efficiency. So you know that's way higher than the single junction solar cell limit. So you might think that the most exciting thing happening in solar energy is all of that crazy, exotic stuff.

But I'd like to put in a word for the humble silicon solar cell. The design has barely changed at all since the original one from the fifties, from Bell Labs, and yet. Over the intervening decades, legions of engineers and manufacturers have taught each other how to make it more and more cheaply. And so yes, there are exotic solar cells that have achieved like 50% efficiency in labs, but it is a completely different thing to be able to make a solar cell in a lab setting. Then being able to scale it out to the masses and solve climate change. And I think that that's what Silicon is going to be able to do.