A Better Way to Get Hydrogen from Water

An experimental approach to splitting water might lead to a relatively cheap and clean method for large-scale hydrogen production that doesn’t require fossil fuels. The process splits water into hydrogen and oxygen using heat and catalysts made from inexpensive materials.

Heat-driven water splitting is an alternative to electrolysis, which is expensive and requires large amounts of electricity. The new approach, developed by Caltech chemical-engineering professor Mark Davis, avoids the key problems with previous heat-driven methods of water splitting. It works at relatively low temperatures and doesn’t produce any toxic or corrosive intermediate products.  

Almost all the hydrogen used now in industrial processes, such as making gasoline, comes from reforming natural gas. If automakers start selling large numbers of hydrogen-fuel-cell vehicles, as they’ve said they plan to do eventually, the hydrogen for those is also likely to come from natural gas unless processes like the one at Caltech are commercialized.  

The basic approach in high-temperature water splitting is to heat up an oxidized metal to drive off oxygen, then add water. In Davis’s case, the starting material is manganese oxide, and the reactions are facilitated by shuttling sodium ions in and out of it. “Without the sodium, the temperatures would go up well over 1,000 °C,” Davis says. With it, the reactions work at temperatures of 850 °C or lower.

The technology is probably far from being commercialized. It still requires pretty high temperatures—a couple of hundred degrees higher, for example, than those used to drive steam turbines at coal and nuclear power plants. Producing those temperatures without fossil fuels would probably involve one of two technologies, neither of which is being used commercially right now: high-temperature nuclear reactors or high-concentration solar thermal facilities that use rings of mirrors to concentrate sunlight more intensely than occurs today in solar thermal power plants. 

The Caltech approach would also need to be tested to make sure the water-splitting cycle can run repeatedly. So far, the researchers have shown that the same materials can be reused five times, but “if you were going to have one of these things work for real, you’d need to run it for thousands of cycles,” Davis says. He says such testing is beyond the scope of his lab. “We feel good about the potential for many cycles on this one, but until you do it, you don’t know,” he says. “All we did here is prove the chemistry could work.”

The rate of hydrogen production would also need to be increased—for example, by switching to materials with a higher surface area. And Davis hopes to lower the temperatures needed still further. The goal is to use this process or a similar one to make use of waste heat at steel mills and power plants. “This is a good start, but the lower we go, the better,” he says.