If we’re ever going to reach net carbon zero, we need to diversify where we get our power from. Wind is abundant in the UK, but NIMBYs object to their placement; hydroelectrics require dams to be built, which can destroy natural aquatic habitats. Most sources of ‘green energy’ have their problems.
Solar panels, however, are becoming cheaper than ever. Advances in the technology that deposits their power-generating layer, known as the ‘photovoltaic’, make them easier to produce than ever before. But why are we still using silicon solar panels, when there are so many better options out there?
I interviewed David Scanlon, Professor of Computational Materials Design at UCL Chemistry with a joint appointment at Diamond Light Source, about why we still use silicon in our solar panels and what new materials are in the pipeline.
How do solar panels work?
Light is, by its nature, energy “..which is why if you leave something out in the sun for a long time it’ll get hot, and sometimes in very hot countries you can cook an egg on the bonnet of a car”, David told me.
That’s the core of how solar panels function – being able to turn the sun’s energy into electricity. “If you shine light on a material that we call a solar absorber, it’ll be able to give enough energy for an electron to hop from what we call the occupied states to the unoccupied states, and this can drive a voltage.”
David here is describing the basics of a quantum mechanical look at materials, using something we call ‘band theory’. In essence, there are so many possible energy levels the electrons in the solid can sit in that they form an ‘ocean’ of occupied states. If the electrons get a bit more energy, they can ‘hop’ to a level that’s less crowded where they can ‘flow’ as electricity.
In the modern-day, silicon solar panels make up the vast majority of the market– but that’s not necessarily because they’re the best choice; we just know a lot about how to make them.
“We have the ability, built up over 50, 60, 70 years, to controllably grow defect-free (well, relatively defect-free) silicon. We can control the thickness and we can control how we grow it into different architectures, so it’s a really really well-studied material”, David said.
“As a solar absorber, it’s inherently not actually efficient. It has something called an indirect bandgap, so it doesn’t actually absorb light well. This means you have thicker layers, which drives up the cost. The only reason we can afford to have silicon solar cells is because its price has been going down, since we know how to do it now; that doesn’t mean we should stick with it, because it isn’t the most efficient thing in the world.”
David does admit that silicon solar cells may not dominate the market in the future, anyway. “Unfortunately, the way that we make these ultra-pure silicon wafers, we use a particular type of sand, and the particular sand we use is actually running short. So we will probably have to have a rethink, and figure out a way to transform other types of sand into the silicon that we want.”
It took silicon solar cells 70 years to get to the position they’re in in the market now; we cannot expect the new materials we’re working on to reach the market in 1 or 2 years… People forget that, because we’re in a world where technology is changing every day, you know. You have a new iPhone every six months… In reality, everything has been built on and built on since, you know, the dawn of the technological revolution, so we have to be more patient.
David Scanlon, Professor of Computational Materials Design at UCL.
Better than Silicon
There are other options out there. MAPI, also known as methylammonium lead iodide, seems like a wonder material which is far more efficient than silicon, with elements common on this planet, in a form that’s easy to turn into a pure thin film. So why aren’t we using it in all of our solar panels today?
David claims this is to do with its stability. “If you shine light on [MAPI], it can actually degrade- which is not what you want for a solar cell. Lead is toxic so there is a worry: if you put these into mass production will you end up with toxic lead leeching out into our water supplies? And then you’ve got a world of pain that we don’t really want.”
MAPI’s rapid development is just as important as its efficiency. “The real interest in these materials is that, if you plot the efficiency gains in this material over the last 70 years, they go up maybe a quarter of a percent a year… but these MAPI based systems went from 3% efficiency to (now) 25.2% efficiency in about 7 years.
“This has also fed into our expectation that things will happen fast. So now we’re trying to move away from these lead-based systems if we can, but if you get a 3% efficient solar cell on your first go everyone claims ‘it’s no good, throw it away’. But even MAPI had to go from 3% to 25%; we have to work on all of these new materials. MAPI is such an anomaly that everyone thinks everything will just shoot up in efficiency really fast like MAPI did, but it’s just not going to happen. It’s an unusual material where everything worked first-time.”
“[MAPI] has everything you’d want from it, except it’s not stable. And if you can’t put a material in a solar cell for at least 10 years, then no one should buy it. You don’t want to be replacing your solar panels every 6 months because the material has fallen apart because you’ve been shining light on it, or it got a bit wet and it fell apart.”
Crystals for Dummies
Crystals are more than just diamonds at the museum; everything that’s solid has a crystal structure. In solids, atoms are arranged close together. Usually, they stay in an ordered form that we call a ‘crystal’, in a structure that repeats thousands upon thousands of times. In table salt, sodium and chlorine have exchanged a negatively-charged electron, so they’re charged- sodium with +1 and chlorine with -1. They’re attracted to each other, holding the crystal together in this periodic fashion.
MAPI takes on a perovskite structure, which is slightly more complicated (above, right). Iodine (in pink) surrounds lead (blue, in the centre of the grey tetrahedra), and a methylammonium ion sits in the centre of the unit. That then repeats thousands of times to form a crystal.
Simulation Science
MAPI is probably not going to happen commercially unless there’s a big breakthrough; such is life. We need something else. This is where David’s research group comes in.
When you think of scientists, you usually think of people in white lab coats mixing chemicals together. In reality, that’s part of the process, but only once a lot of work has already been done.
One of David’s favourite examples of his group’s work is their analysis of double perovskites in 2016. These materials were developed to try and fix MAPI’s stability and toxicity problems by replacing lead, a 2+ ion, with a mixture of 1+ and 3+ ions. In theory, those charges would balance out to the same overall (if we double the size of the crystal we’re looking at- making it a so-called ‘double’ perovskite), leading to the same structure with just as impressive efficiency, and no lead.
Researchers “…were all jumping on the bandwagon that these ‘double perovskites’ would be the next big thing. We showed that these double perovskites would have indirect band gaps, and would not be useful for photovoltaic applications,” David told me.
The group did this by simulating the materials.
“You can tell a computer ‘I want this combination of three elements, in this stoichiometry [ratio]. Find me the lowest energy structure that could form.” You can calculate the energy of that structure and compare that to the energy of every possible other competing phase… and if your material is stable, and is dynamically stable, then you know it can be made.”
That’s where the scientists in lab coats come in. “You go to your experimental chemist… and if they trust you, they’ll put somebody in the lab to try and make it.”
A lot of David’s work rests on interrupting research before resources are wasted in the lab on materials that just won’t work- and this is exactly what happened with the double perovskites.
“I think it’s a nice example of using calculations to do rational analysis on systems, and maybe bursting the hype train a bit!”
Thing is, it’s not always doom and gloom; sometimes materials do the job when they hit the lab. “If [a material] works, then that’s probably the best feeling you could get, for what I do anyway. And sometimes it doesn’t work, sometimes you find a mistake in your calculations, and that’s really frustrating- but that’s research, we all make mistakes. But if it works, it’s pretty cool.”
Thanks to David Scanlon for talking to me about his research group and photovoltaics.
I hope everyone is keeping safe. I have a lot more free time now- so hopefully I can use some of this newfound time to write articles about what I’m being examined on. It’s a double whammy!
Ollie Wardle 