A little mineral with a sunny disposition, perovskite is cheap and ubiquitous (Image: Gary Cook/Visuals Unlimited, Inc/Getty)
The solar cell of the future will be flexible, highly efficient and oh-so cheap – just as long as we can make it work in the rain
GOOD things come to those who wait, and Tsutomu Miyasaka had waited a long time. Knowing that a solar cell can be made using just about any pigment – coffee, chlorophyll, red wine – the Japanese physicist had spent years testing all sorts of colourful substances in the hope of finding one as efficient as it was cheap. Then, one day in April 2007, a student walked into his lab at the University of Tokyo, carrying a lump of an unremarkable mineral called perovskite.
Advertisement
It proved to be the start of something entirely remarkable. When Miyasaka reported the results from his first perovskite solar cell in 2009, it converted just 4 per cent of the sunlight’s energy to electricity. By 2012, the figure was over 10 per cent, and others were beginning to take notice. With groups around the world now on the case, efficiencies are touching 20 per cent, beating most solar cells currently on the market. “The field is moving so quickly – everyone is jumping on board,” says physicist of the University of Oxford. Is this the solar technology everyone has been waiting for?
The sun is a virtually limitless source of clean energy, beaming down enough on Earth in an hour to meet humanity’s needs for a year. Currently, photovoltaic cells supply about 140 gigawatts, or just 1 per cent of our global power demand. The most widely used silicon cells took 50 years from their invention in the 1940s to reach their maximum efficiency of about 25 per cent. The sort of slab you might install on your roof is about 15 per cent efficient in bright sunlight, and produces electricity at a cost of roughly 50 cents a watt, twice the typical cost of coal.
Younger alternatives such as copper indium gallium selenide (CIGS) and cadmium telluride cells have advantages: cadmium telluride is as cheap as silicon, while CIGS can be printed on flexible substrates. But with maximum efficiencies of about 20 per cent, neither is a market leader. The record-holding cell has long been gallium arsenide, which converts nearly 30 per cent of sunlight to electricity – .
Miyasaka’s wonder material might shake things up. Perovskite occurs in the form of calcium titanium oxide in rocks the world over, most famously in Russia’s Ural mountains, where it was discovered in the early 19th century. The name has more generally come to describe a class of compounds that share a crystal arrangement with this mineral: imagine a transparent die with fives on every side, with each dot an atom, and a single atom at its very centre. The crystal usually contains three types of ion; Miyasaka grew his first samples from ammonium, lead and either bromide or iodide.
In conventional solar cells, a semiconductor such as silicon absorbs sunlight to produce electrons and their positive counterparts, holes, and then separates them to create a flow of charged particles and hence an electrical current. Not many materials can both absorb enough light at the right wavelength and also diffuse the charges, and so in dye-sensitised cells a pigment is added to the mix that does the absorbing before transferring the charges to a semiconductor for separation.
The breakthrough came when Miyasaka, working with at the University of Oxford and others, realised that perovskite makes the semiconductor redundant – it can actually shift the electrons and holes better by itself. How exactly it does this remains something of a mystery, but one theory goes that because the material crystallises easily, the process is less likely to create defects that would otherwise stop charges flowing out freely.
“Exactly how the materials work remains something of a mystery”
Over the limit
Whatever the internal mechanisms, the material is cheap and easy to manufacture. Last year, a group including Snaith and Johnston showed that perovskite cells could be fabricated using the conventional approach for silicon cells, . An even cheaper possibility that Snaith and his spin-out company are working on is pouring a solution containing perovskite on to a substrate and letting it crystallise. Snaith, who is hoping to bring perovskite cells to the market by the end of 2016, thinks that solution-processed cells could one day generate electricity for as little as 10 cents a watt.
In April, materials scientist of the University of California, Los Angeles, reported a perovskite cell reaching a record efficiency of 19.3 per cent, although he says the US National Renewable Energy Laboratory still needs to confirm this figure. Snaith won’t yet reveal the details, but thinks that soon the materials might even beat what physicists William Shockley and Hans Queisser believed in the 1960s was a fundamental limit. They concluded that no solar cell could ever convert more than about a third of the sun’s energy into electricity. “We’re starting to wonder whether there’s something a little bit special about this system,” says Snaith.
There is still a lot more work to be done before perovskites can hit the commercial big time. Those developed so far mostly contain lead, a toxic element. In May this year Snaith’s group and another based at Northwestern University in Evanston, Illinois, showed that but the efficiency of these cells, while rising fast, is currently just a few per cent.
More problematically, when left exposed, perovskites absorb moisture and swell, losing their light-harvesting properties. “That’s a serious one,” says chemist of the University of Notre Dame in Indiana, who has recently started looking at the materials. A solar cell that cannot be left in the rain might sound ludicrous, but Snaith, while trying to tweak the materials’ composition to make them water-resistant, points out they could be sealed in glass, as many solar modules are.
Perovskite’s uses may even go beyond photovoltaics. Earlier this year, at the Swiss Federal Institute of Technology in Lausanne and his colleagues demonstrated that the materials could be made to emit light, perhaps paving the way for cheap lasers for telecoms or high-resolution displays. Such displays might even double-up as light harvesters, perhaps for advertising billboards that charge up during the day. “It’s not exactly an application to save the planet, but it might make some money,” says Snaith.
, a theorist at the University of Pennsylvania in Philadelphia who has not been involved in the cells’ development, is cautiously optimistic about perovskite’s prospects. “I think the buzz is justified,” he says. “It may turn out in the end that they are not suitable for practical power generation, because of issues with their durability or manufacturing. But they have shown an unprecedented rise in sunlight power conversion efficiency.”
Snaith is certainly confident. “I’m still a firm believer that photovoltaics is going to produce most of our power one day,” he says. “And if perovskites really take off in the next 10 years, then in principle there could be terawatts of them – and they will make the current silicon industry look minuscule by comparison.”
That is a heavy burden of expectation, says Miyasaka, who continues to work in the field he instigated. “I feel somewhat scared and responsible, but I will do my best,” he says.
This article appeared in print under the headline “Rise of shine”
