本文作者：小红猪小分队

Source：http://www.newscientist.com/article/mg21729011.900-midnight-sun-how-to-get-24hour-solar-power.html?full=true

Rust may be the scourge of electronics but it could help solar power run all night

MOST engineers would have been horrified to find even a little bit of rust on their electrodes. But Kenneth Hardee and Allen Bard had made theirs entirely of the stuff. In their pursuit of cheap solar power, the pair had been trying to coax a current out of the cheapest material they could find. And they succeeded: exposed to visible light, it produced a small but decent current.

That happened in 1975, just as silicon was becoming the next big thing. Silicon’s greater efficiency made it the mainstay of photovoltaic solar cells, and it has stayed at the top of the market ever since. Rust simply didn’t have the electrical properties to compete. The small breakthrough at the University of Texas at Austin fell by the wayside and the only time anyone thought about rust, if they thought of it at all, was when they wanted to get rid of it.

But over the past few years, the spotlight has been swinging back to a substance that, contrary to popular opinion, may actually be something of a wonder material. Though iron oxide can’t compete with silicon’s efficiency at turning the sun’s energy into electricity, it can do something silicon cannot: help store the sun’s energy. Humble flakes of rust might be the way around one of the most intractable problems plaguing solar power – the night.

Solar research has focused almost exclusively on efficiency. Every day, the sun bathes our planet in more energy than we can hope to consume in a year. But harvesting it is not easy. Even the best available technology – the billion dollar solar panels made from expensive combinations of rare earth elements used by the International Space Station – can only convert 46 per cent of the sun’s energy to electricity, and that’s under ideal conditions. The usual figure is much lower. Back on Earth, cheaper, silicon-based photovoltaics mop up about 15 to 20 per cent.

Ways to store excess energy so it can be used when the sun is not shining are urgently needed. In part because it can only be used at the moment it is generated, this apparently limitless resource contributes the least of all renewables while remaining about 20 times more expensive than power from fossil fuels (see diagram).

Batteries are the most obvious solution, but their low energy density combined with the high cost of a system capable of powering an entire home – which would need replacing every few years – makes them an option only for the rich. A far better way to store solar energy is by using it to make hydrogen. The element’s chemical bonds pack a huge punch, storing as much as 170 times the energy per kilogram as standard lithium ion batteries. Hydrogen is also versatile: once you’ve got it, you can use it however you want. Put it in a fuel cell and you can generate electricity on demand by recombining it with oxygen; combine it with carbon monoxide and it becomes methanol biofuel; stored properly, it can even be burned like any other gas fuel.

The simplest way to turn power from photovoltaic cells into hydrogen is by using that power to run an electrolyser. This splits water – H₂O – into hydrogen and oxygen. Simple, perhaps, but also inefficient. Of the miserly 15 per cent of solar radiation that standard photovoltaic cells are able to capture, another 30 per cent is lost in the conversion. By the time you are done, you’re better off with the rechargeable battery.

Power from water

The better option would be to find some cheap electricity-conducting material that can bypass photovoltaic cells altogether and simply use the sun’s photons to electrolyse water and make hydrogen.

For a material to electrolyse water directly, it must release electrons of the right energy when hit by a photon. When these electrons are excited enough to vacate the material, they leave behind gaps known as holes. To fill these holes, a water molecule donates one of its own electrons. In this way, electrons and holes work together to oxidise water, and turn it into hydrogen and oxygen.

Silicon is not the right tool for the job; its electrons don’t have the right energy. All materials need different, precise amounts of energy to make their electrons jump free of their atoms. Silicon atoms need only 1.11 electronvolts to loosen an electron – but splitting water requires electrons with at least 1.23 eV.

Materials that do hit the spot can be created from exotic compounds. By combining zinc selenide and cadmium sulphide crystals with a platinum catalyst, for instance, engineers at Bowling Green State University, Ohio, were able to liberate the proper electrons. But the complicated process and rare materials add up to a device that works in the lab but could be too costly to commercialise.

And so researchers have begun to revisit rust. Iron oxide’s perfect energy hit – 2.1 eV – isn’t even the most compelling reason to pursue this oft-maligned substance. It is also non-toxic and – literally – dirt cheap. What’s more, it is abundant to the point of ubiquity. Rare earth elements aren’t just expensive and toxic; obtaining them can also be politically unpredictable. Researchers talk about their availability in the same way that some people forecast peak oil. When China recently clamped down on neodymium exports, for example, entire industries suffered, from car motors to magnets. That is not a problem for iron oxide. “Nobody controls rust,” says Klaus Hellgardt, who works on iron oxide hydrogen generation at Imperial College London.

Rust resistant

Of great importance, too, is its stability. Many materials buckle under the corrosive effects of water splitting, but iron oxide can last up to a year – some think longer – because, as Hellgardt says, “it’s not like it will rust”.

Although its ability to convert solar power to hydrogen is not the most efficient in the world – recent research indicates a theoretical limit of 16.8 per cent – its sheer abundance means quantity can make up for quality.

But the Cinderella material is still missing a shoe. “It hasn’t performed well to date,” says Nate Lewis of the California Institute of Technology in Pasadena. “That doesn’t mean we couldn’t make it perform well.”

Just because rust has the right physical properties to electrolyse water doesn’t mean it can do so unaided. So the majority of the rust research of the past decade has revolved around coaxing its electrons to the water.

The first problem to solve was the one that had stymied Hardee and Bard in 1975. Iron oxide doesn’t conduct electricity very well, meaning that by itself it can’t send enough electrons out to the edge, where they would be more useful. It needs a kick. One way to do that is to obtain additional solar power from a device called a tandem cell. In 1991, Michael Grätzel, an engineer at the Swiss Federal Institute of Technology in Lausanne (EPFL), used a thin layer of titanium dioxide, which had been dyed to boost its photon absorption, to create a dye-sensitised solar cell, a simple and cheap way to generate an electric current without silicon. By feeding the resulting current into the rust layer below, they were able to push the correct electrons out to electrolyse the water (Nature, vol 353, p 737).

Grätzel’s device managed an unprecedented efficiency of 4 per cent. However, that required two extra tandem cells. The extra energy was necessary to kick the electrons to a higher energy level. Without these, the rust would suck the electrons back into its crystalline matrix, reabsorbing them before they could get clear.

The only solution would be to make the rust layer thin enough to allow the electrons to escape – on the order of tens of nanometres. In 1975, and even in the early 1990s, this would have been impossible. By the turn of the 21st century, however, nanotechnology had advanced sufficiently for it to be possible to manipulate a material’s physical structure – and yield some surprisingly elegant solutions.

Jordan Katz, at Denison University in Ohio, has created a thin coating comprised of rust rods a few nanometres wide. That narrow width gives the device a very high surface area, while allowing water to seep into the nanosized crevices between the rods. This lets electrons and holes escape the material and meet the surrounding water. But Katz says he is far from finding a material with marketable efficiency.

Researchers at EPFL found a way to do it. To aid the electrons’ escape, Kevin Sivula creates nanorust using “cloud” deposition, which involves spraying a mist of iron solution onto a surface. This deposition method causes the iron oxide to grow into forests of microscopic cauliflower-shaped “trees”, creating the kind of fractal surface area that allows the electrons to make their getaway, but that can also be mass-produced.

Last year Sivula’s group created a working device using nothing more expensive than glass. At 3.6 per cent, its efficiency rivalled that of the Grätzel device, but without help from extra tandem cells (Nature Photonics, vol 6, p 824). And Sivula says he can push that to 10 per cent within a couple of years.

However, his goal might be stymied by a problem that arises, paradoxically, when the rust layer is very thin. A fundamental tension for any electrolysing material is that you need it to be both as thick and as thin as possible. Thinner is better if you want your electrons to have any shot at escape. But to absorb as many photons as possible, the rust layer needs to be thick. A 20-nanometre layer absorbs only 18 per cent of the total absorbable photons. Boost the thickness to 1 micrometre and you catch nearly all of them – but then they get stuck.

To resolve this, Avner Rothschild and his team at Technion University in Haifa, Israel, turned to quantum physics. Their device traps incoming light in 30-nanometre rust films. When the photons enter the device, they are forced into a chamber of facing, V-shaped mirrors, which bounce them back and forth until they are absorbed. What’s more, the interference between the ensuing forward- and backward-propagating light waves further boosts the absorption, especially close to the film’s surface. The electrons and holes can easily reach the surface before recombination is possible. Thanks to this tweak, the device is able to absorb 71 per cent of the incoming photons, but it is so thin that electrons can escape, leading to a theoretical efficiency of 4.9 per cent (Nature Materials, in press).

That’s impressive by iron oxide’s low standards, but not exactly the stuff of commercial products – or is it?

Here at last is the real genius of rust, and why it may eventually eclipse silicon despite the most feeble efficiencies. Even if it never reaches its 16 per cent maximum, says Sivula, it’s so cheap that you can make vast swaths of it, which is exactly what he and the other rust researchers are planning. “What ultimately matters is not efficiency, but cost per watt,” says Katz. Even 10 per cent efficiency “at the right price”, he says, would beat a 50-per-cent efficient photovoltaic cell, because it would make rust worth spraying on every surface.

And this is exactly the goal. Sivula thinks you could coat his iron “cauliflower” mixture onto something akin to wallpaper, printing out sheets of solar cells, generating hydrogen everywhere and anywhere. Lonely desert outposts would be a perfect home, and the process could use filtered waste water.

The hydrogen problem

To be sure, a few more issues must be resolved before this dream can be realised. Once water splits, for example, “you’ve effectively created a bomb”, Hellgardt says, because oxygen and hydrogen can react explosively. A more benign yet equally poor outcome is that your hydrogen and oxygen just recombine to form water that is slightly hotter than it was before.

Separating the two gases is straightforward. In Sivula’s cell, for example, a membrane attracts oxygen and hydrogen differentially, letting them bubble up separately. Hellgardt has a different idea: if you don’t plan to use the oxygen, why generate it in the first place? His design uses low-grade waste water to “eat” the oxygen. Instead of becoming a gas, it reacts with the organic compounds in the water, leaving the hydrogen to bubble safely off to the storage tank.

And there’s the final rub for rust solar: while it can store solar energy by making hydrogen, storage presents its own problems. The gas is notoriously difficult to keep confined without relying on expensive, tough materials that won’t corrode – or explode. Indeed, this problem has tripped up the entire promise of the oft-hailed hydrogen economy.

Researchers have been working on a constellation of solutions to the problem. Alongside steady improvements in fuel cells, a number of new approaches are under way. For example, researchers at the University of New South Wales in Australia recently used nanoscale sodium borohydride for storage. Normally, the salt must be heated to 550 °C to release hydrogen stored in its bonds, but at nanoscales, it was coaxed to do so at 50 °C. That is a promising development for portable hydrogen at many scales.

Promising, but maybe not necessary. Simple canisters of hydrogen, stored on site and burned like camping fuel, would also do the trick. This is what Brian Holcroft, director of Stored Solar, sees as the immediate niche in places like Kenya, where sunlight is plentiful and energy infrastructure lacking. He has collaborated with EPFL to use the tandem cell-and-iron oxide setup for his company, which makes off-grid energy solutions. He is keen to get these devices onto roofs in the developed world as well, where their owners could get hydrogen fuel and electricity without a grid.

And maybe they wouldn’t need the tandem cell. Insights from decades spent coaxing rust electrons to split water may move Hardee and Bard’s original dream from the past into the future: a rust photovoltaic device, albeit an inefficient one, coupled with a storage device.

“If you don’t care about efficiency at all, a rust cell could function either to make fuel or electricity, or both at the same time,” says Katz. “It could make electricity during the day during peak electrical demand, but produce fuel instead when demand is not as high.” Given the economic realities of solar energy, the tiny current Hardee and Bard tapped into in 1975 may yet become a renewable energy source that could cover the planet. Maybe it’s time to enter the Rust Age.

Naomi Lubick is a writer based in Sweden