Fuel From The Sun

Water + sunlight = fuel. This equation embodies the use of solar energy to rip apart water molecules to produce hydrogen, which can be used as an energy-rich fuel for vehicles and to produce electricity. If perfected and made affordable, the technology could supply a substantial portion of future global energy demand, which is anticipated to double between now and 2050.

Key to solar water splitting is developing inexpensive catalysts to capture light efficiently and speed the process while minimizing the amount of electricity needed to drive the electrochemistry. Most catalysts so far have less than stellar efficiencies, rely on expensive and rare metals, or tend to be easily deactivated under harsh working conditions.

Two U.S. research groups have recently reported breakthrough developments that could signal a new wave of progress in producing H₂ via solar water splitting. Daniel G. Nocera and coworkers at Massachusetts Institute of Technology have made a heterogeneous cobalt phosphate water-oxidation catalyst with improved stability. And Craig L. Hill of Emory University and coworkers have created a related homogeneous cobalt catalyst supported by bulky polytungstate ligands that displays improved catalytic activity.

Both catalysts are made from Earth-abundant elements, avoid organic ligands that are prone to oxidation during electrolysis, have a built-in mechanism for self-repair to improve lifetime, and operate at neutral pH with modest electricity input.

But that’s only part of the story. The catalysts and the researchers developing them are beneficiaries of a sudden abundance of federal funding.

For most of the past 30 years, funding for solar-fuel research was flat, according to Mark T. Spitler, manager of the Solar Photochemistry program in the Department of Energy’s Office of Basic Energy Sciences. But since 2005, DOE has been flush with money for renewable energy research and is using it to create new methods of funding designed to accelerate the transition of basic research to commercial applications, Spitler says.

For example, DOE’s Advanced Research Projects Agency-Energy (ARPA-E) program has provided hundreds of millions of dollars in three rounds of funding to promising start-ups and research labs to help them move high-risk, high-payoff technologies toward fruition (C&EN, March 22, page 38).

“ARPA-E is a tremendous success story,” says chemical engineer Amir Nashat, chief executive officer of Sun Catalytix, in Cambridge, Mass. The company, which Nocera started last year to develop inexpensive solar-powered water-splitting systems to make H₂, has garnered more than $4 million in ARPA-E funds. “ARPA-E is having an incredible impact on Sun Catalytix and other small companies, enabling us to follow our dreams to turn science into technology and eventually into commercial products,” Nashat says.

Another new DOE initiative is a set of three Energy Innovation Hub grants worth $122 million each for multidisciplinary teams of dozens of principal investigators to speed up integrating the various pieces of promising technologies, Spitler notes. One of the hubs is aimed at “creating the science and technology for a solar fuels industry that currently doesn’t exist,” Spitler says.

Nocera is part of a team based at the National Renewable Energy Laboratory that is a finalist for the solar fuels hub grant, which will be awarded later this summer. Hill was part of a team based at Oak Ridge National Laboratory, but that team didn’t make the final cut.

Hill says the funding would have been nice, but he is able to pursue his work with other DOE funding and Department of Defense grants. In addition, the National Science Foundation is funding research on solar fuels, he notes.

“So much is happening right now in solar fuels and photochemistry, with grants, meetings, and scientific conferences, it’s hard to keep up,” Hill says.

Commercial technology to derive H₂ from water by electrolysis has been available for nearly a century. Because electrolysis remains expensive, industrial H₂ production continues to be primarily by steam reforming of petroleum and by coal gasification, both of which are based on limited fossil resources, comments Matthias Beller of Leibniz Institute for Catalysis at Germany’s University of Rostock, who studies iron-based H₂-generating catalysts.

“Clearly, on a mid- to long-term basis, there is an essential demand for alternative technologies to generate H₂ in a more sustainable manner if it is to be used as a transportation fuel and for producing electricity,” Beller says. “Photocatalytic water splitting offers the most straightforward production of H₂ from H₂O. In this respect, the recent work from the Nocera and Hill groups is highly interesting.”

Water splitting is a two-stage process. In an electrolysis cell, water is oxidized at the positive electrode, or anode, to form O₂, along with four hydrogen ions and four electrons. The hydrogen ions migrate to the negative electrode, or cathode, where two H⁺ ions are reduced by two electrons arriving through an external circuit to form H₂. Of the two electrode processes, both of which require a catalyst to be efficient, the water-oxidation reaction is more complex and thermodynamically demanding.

In fuel cells, which also require catalysts, the opposite reactions take place to release the energy stored in the H–H bonds: H₂ and O₂ are fed into a fuel cell, releasing electrons to make electricity and producing water.

Electricity to power water splitting can come straight from a wall outlet, indirectly from sunlight via a photovoltaic solar cell, or directly from sunlight at the anode by interfacing the water-oxidation catalyst with a semiconductor—a photocatalyst.

“Developing efficient and low-cost catalysts and electrode materials for water splitting to produce H₂ is demanding technology,” notes Can Li of China’s Dalian Institute of Chemical Physics, whose group works on inorganic semiconducting photocatalysts for water splitting (C&EN, Aug. 10, 2009, page 7). “Although H₂ produced in the reduction reaction is the fuel we need, the key to achieve high-efficiency water splitting is to solve the water oxidation problem, because it is a kinetically slower reaction. The cobalt-based oxygen-evolving catalysts developed by Nocera and Hill are breakthroughs on the way to solving this key problem of water oxidation for efficient overall water splitting.”

Scientists have pinned down iridium oxide as being the best water-splitting catalyst, but it’s also one of the most expensive. Platinum metal also works well, as do other metal oxides such as those of cobalt, ruthenium, nickel, rhodium, and manganese. But metal oxides are susceptible to corrosion in water, meaning that chemists must come up with catalyst designs that protect the metal oxides so they can be used in electrolysis cells.

In a series of papers over the past two years, Nocera and coworkers have reported the discovery and attributes of their novel, low-cost heterogeneous cobalt phosphate catalyst. The catalyst self-assembles from a dilute phosphate-buffered Co²⁺ solution at neutral pH during electrodeposition (C&EN, Aug. 4, 2008, page 7).

When a potential is applied, Co²⁺ oxidizes to Co³⁺, which precipitates with phosphate as an amorphous thin film on an indium tin oxide electrode or fluorine tin oxide electrode, Nocera explains. The phosphate serves as a ligand to cobalt oxide and as an H⁺ acceptor that shuttles the ions from the anode to the cathode, he says.

Because the catalyst is an amorphous solid, its exact structure is unknown. Nocera’s group has used various techniques, including X-ray absorption, electron paramagnetic resonance, isotope labeling, and computer simulations to gain a better understanding of the catalyst and improve its performance. The researchers believe that the active catalyst is a cobalt oxide cubane cluster with bridging oxo groups. This structure is similar to the calcium manganese oxide cluster in the oxygen-evolving complex of photosystem II, the enzyme nature uses in photosynthesis, Nocera points out.

But the key feature of Nocera’s catalyst is its ability to repair itself. Catalysts that mediate multielectron transformations operate under harsh conditions, Nocera says. In particular, oxygen-evolving catalysts with complex organic ligand systems tend to easily oxidize, he notes. “

Molecular catalysts with organic ligands appear to break down to metal oxides under the harsh conditions,” Nocera adds, “which is why we went to an all-inorganic simple system that stands up over time.”

Cobalt phosphate at the catalyst surface is continuously refreshing itself, Nocera explains. His group has shown that Co²⁺ ions readily exchange phosphate ligands during the catalytic cycle and that the Co³⁺ and Co⁴⁺ species present are substitutionally inert. Any Co²⁺ ions released from the electrode surface and oxidized to Co³⁺ during water splitting are redeposited with phosphate on the electrode surface.

In March, Emory’s Hill and coworkers reported their cobalt oxygen-evolving catalyst containing the bulky polytungstate ligands [Co₄(H₂O)₂(PW₉O₃₄)₂]^10– (C&EN, March 15, page 48). The catalyst self-assembles in boiling water from Co²⁺, phosphate, and tungstate salts. Its Co₄O₄ core is protected from oxidation by a pair of the polyoxometalate ligands.

The cobalt polytungstate is a homogeneous catalyst, so Hill’s team was able to study its X-ray crystal structure. Because the catalyst is molecular, it is easier to optimize by interchanging ligand units and interfacing with photosensitizers. And like Nocera’s catalyst, the self-assembly and exchangeable ligands enable the catalyst to constantly repair itself, Hill says.

The key feature of Hill’s catalyst is its high activity. The team showed that in a phosphate-buffered pH 8 solution with ruthenium bipyridine as a chemical oxidant, the catalyst produces O₂ at a rate of more than five catalytic cycles per second. Hill cautions that many parameters affect this turnover rate, making it hard to compare catalysts from different labs. But he believes his cobalt complex is the fastest known water-oxidation catalyst—1,000 times faster than heterogeneous catalysts and five times faster than homogeneous catalysts bearing oxidizable organic ligands.

In a working electrolysis cell, the catalyst could complete a very high number of turnovers, Hill believes. But it will still be short of what is needed. Engineers building solar water-splitting devices say a catalyst must be able to function for 1 billion turnovers to run for several years, he notes. “Stability is an enormous challenge for the oxygen-evolving catalyst,” Hill says.

“The focus on purely inorganic catalysts seems important, because such catalysts should be more stable over long periods under oxidizing conditions,” observes T. Don Tilley of the University of California, Berkeley, whose group is working on cobalt oxide nanoparticle electrodes in alkaline solution.

“The work by Nocera has helped to draw considerable attention to the use of inorganic catalysts for water splitting and the fact that such catalysts might be stable for long periods and over a range of conditions,” Tilley adds. “Hill’s work illustrates a new and different way to utilize purely inorganic catalysts with polyoxometalate ligands. The intriguing thing about Hill’s catalyst is that it offers the possibility of synthetic control over catalyst structure at the molecular level. The initial results on these catalysts are certainly encouraging.”

The biggest challenge for scientists working on solar fuels is to start putting together the components that collect sunlight, split water to make H₂, and supply the H₂ to a fuel cell. “Chemists need to start moving on from just focusing on making a water-splitting catalyst to making integrated systems—a technology,” Nocera says. “It has to be simple, though, so that it can be manufactured on a large scale.”

Hill agrees. “A lot of chemists are still interested in modeling enzyme active sites to study O₂ and H₂ production mechanisms,” Hill says. “It’s important research, but it’s time we go forward with what is working.”

To that end, Sun Catalytix is moving quickly toward commercialization and anticipates having its first products ready in four years, Nocera says. Planned water-splitting devices to make H₂ will use No­ce­ra’s cobalt phosphate water-oxidation catalyst paired with a low-cost water-reduction catalyst that Sun Catalytix is developing.

One concern is the catalysts’ current density. Commercial electrolyzers based on platinum or other costly electrodes driven by electricity from the power grid to produce H₂ are extremely efficient and operate at current densities as high as 1,000 milliamperes/cm² to keep the electrodes at a reasonable size. Nocera’s cobalt phosphate catalyst was originally reported to operate at a current density of 1 mA/cm², generating a very small amount of O₂. That value is well below the 40 to 50 mA/cm² that a solar-powered electrolyzer would need to generate H₂ in full sunlight.

Nocera’s more recent electrodes are operating at about 100 mA/ cm², he says, which should be sufficient for what he has in mind. Nocera envisions developing low-cost “personalized” energy systems in which rooftop solar panels supply energy to a home, with excess energy going to split water in an electrolyzer to make H₂. The H₂ can be temporarily stored in a tank, Nocera says, and then used to run an internal combustion engine or a fuel cell when the sun isn’t shining. “Because we are designing for solar photovoltaic input, we don’t need higher current density,” Nocera says.

Personal-sized electrolyzers are already available for home use, but Nocera wants to create one that runs directly off sunlight and would be affordable to the 20% of the world’s population that has no regular access to electricity.

Hill is likewise aiming to develop a commercial system. His team’s catalysts are being patented, Hill says, and he is in discussion with potential partners whom he would put in charge of starting a company. “It’s still too early to think about what type of commercial applications might be possible,” he says.

Hill points out that his homogeneous catalysts would likely have to be tethered to the electrode surface in a functioning device. One approach might be to covalently attach the catalyst to light-absorbing doped titanium dioxide nanoparticles that can be deposited on the anode surface.

Both Sun Catalytix and Hill’s proposed new company are coming into existence at an interesting time. Some experts in the field are still wondering what the advantages might be to making a better water-splitting catalyst and how much better they can be over current commercial electrolyzers using precious-metal catalysts. Some scientists contacted by C&EN think Nocera’s idea of personalized energy is pure folly. They believe it is more realistic to produce H₂ at large-scale facilities and then generate the electricity and feed it into a power grid.

Both models for H₂ generation may be viable one day, but many technical hurdles remain before integrated water-splitting technology will see wide-scale use. For the scientists involved, the knowledge that more energy strikes Earth in one hour from the sun than is produced from fossil fuels globally in one year is incentive enough to continue driving the research forward.