Issue Date: September 14, 2009
Peering Into Water Photolysis
Catalytic, light-activated oxidation of water to form H2 is key to harnessing solar power. But after four decades of research into catalysts for water photolysis, the detailed nature of such systems is still poorly understood.
“This is a key area on which theorists and spectroscopists are just beginning to focus their attention,” says Heinz Frei, deputy director of the Helios Solar Energy Research Center at Lawrence Berkeley National Laboratory. It was also the topic of a session in the Division of Physical Chemistry that Frei helped organize as part of a symposium on the physical chemistry of photon-to-fuel conversion at the ACS national meeting last month.
The goals of the theoretical work described at the meeting are to understand how water organizes near and adsorbs onto electrode surfaces, how the water-electrode interface affects electron energy levels, and what chemical steps are involved in photolyzing water. The work should provide new insights into how electrochemical systems work and help experimentalists generate new ideas to improve existing electrodes or develop new ones.
In a talk during the session, Michiel Sprik, a chemistry professor at the University of Cambridge, described molecular dynamics simulations that use density functional theory to look at the interface between the aqueous HO–/HO∂ redox couple and the surface of a titanium dioxide electrode.
Sprik and colleagues are trying to get a detailed picture of what happens to the electronic levels of the two materials when they come into contact. As water hydrates the TiO2 surface, Sprik said, the surface can take on a negative charge at high pH or a positive charge at low pH, an effect that has a “huge” impact on the electronic levels of the TiO2 surface.
So far, Sprik’s computations nicely match experimentally observed properties, in particular the acidity of the TiO2 surface. He and postdoc Jun Cheng have also successfully modeled a “compact double layer,” which forms at high pH as a layer of sodium cations from the solution assembles to counteract the negative charges on the TiO2 surface. The additional layer of ions changes the relative positions of the energy levels of the electrode and electrolyte.
Sprik’s group is still working on perfecting their computations, but with a realistic model in hand, “we’re aiming to help experimentalists understand how the redox energies of molecular reactants and the electronic states of the semiconducting electrode are influenced by this very heterogeneous environment and how they can be changed” to improve water photolysis, Sprik said.
Two other researchers, Maria Victoria Fernandez-Serra, a physics professor at the State University of New York, Stony Brook, and Xiao Shen, a graduate student at SUNY Stony Brook working for physics professor Philip B. Allen, spoke about a collaborative project focused on understanding water at the surface of the semiconductor gallium nitride, which will catalyze the oxidation of water when illuminated with ultraviolet light.
Fernandez-Serra discussed computational research to compare the behavior of water at metallic and semiconducting surfaces. When she and graduate student Adrien Poissier sandwiched layers of water between two palladium surfaces, they expected to see symmetric water-palladium interfaces. Instead, they observed that on one interface, half of the water molecules oriented perpendicular to the surface, with one H pointing toward the Pd, creating a hydrophobic environment at the interface.
At the second interface, they observed a semiordered structure with the hydrogens pointing away from the metal, making the surface environment hydrophilic. And it appears that the two different surface environments cause an electric field to develop across the system.
In contrast, when Fernandez-Serra and graduate student Jue Wang looked at the properties of water on a semiconductor surface such as GaN, they found that the water molecules dissociatively adsorb onto the surface. The majority of the resulting H+ ions stay at the surface bound to N sites, but a few diffuse away as H3O+ ions, which then disrupt the hydrogen-bonding network near the interfaces. The details of these bulk water interactions could be important for understanding the physics of photocatalysis, Fernandez-Serra said.
In his talk, Shen focused on the chemistry induced by the dissociative adsorption of H2O on GaN.
“There are a lot of difficulties in trying to model the GaN system because you can either model the surface well and not treat the chemistry and solvation very well, or you can go to the other extreme,” said James T. Muckerman, a chemist at Brookhaven National Laboratory who is also collaborating on the project, along with Mark S. Hybertsen, who leads the theory group at Brookhaven’s Center for Functional Nanomaterials. “We tried to build a model that would be faithful enough to both for us to get some insights into the problem,” Muckerman said.
The group has tracked the fate of water in the GaN catalytic cycle: Water adsorbs dissociatively onto the surface, with HO– attaching to Ga sites and H+ attaching to N (J. Phys. Chem. C 2009, 113, 3365). An electron vacancy, or “hole,” generated in the semiconductor then takes an electron from an HO–. The resulting HO• loses its proton to the solvent, leaving behind O•– at the Ga site. This is the first of four proton-coupled oxidation steps that produce a sequence of ionic surface-bound intermediates, eventually yielding O2.
Armed with the GaN piece of the puzzle, as well as other groups’ work on ZnO surfaces, the group is now studying adsorption processes on a GaN-ZnO alloy, which oxidizes water in the presence of visible light. More visible light than UV light reaches Earth’s surface, Muckerman noted, so understanding and improving a catalyst that uses visible light is important for harnessing the power of water photolysis.
Adam Willard, a postdoctoral associate in chemistry professor David Chandler’s group at the University of California, Berkeley, is also examining the dynamics of water at electrode surfaces, but from a different perspective.
Rather than describing the electronic properties of the system, Willard instead uses a classical force field that approximates adsorption interactions. “It’s almost more statistical physics than chemistry,” but the results could provide insight into redox processes and performance of real devices, he said.
In particular, Willard is examining platinum electrodes on which adsorbed water molecules orient to form hydrogen bonds parallel to the electrodes. The first thing he has observed is that the water molecules relax, or rotate to change their hydrogen-bonding network, over a timescale of hundreds of picoseconds, rather than the 3 to 5 picoseconds for bulk water.
But closer evaluation reveals that the relaxation dynamics of adsorbed water molecules are not well described by a single number. Although the average relaxation time is on the order of hundreds of picoseconds, inactive regions relax much slower, and active regions relax much faster. Furthermore, which regions are active or inactive changes over time. “If you watch a movie of this activity, the active regions appear to migrate around the electrode,” Willard said.
The behaviors differ depending on specific electrode geometries. Consequently, electrode geometry might play a critical role in water redox processes, which likely depend on the collective rearrangement of many adsorbed water molecules, Willard said.
As physical chemists continue to dig into the behavior of water and its redox processes at electrode interfaces, clearly they will have much more to uncover.
- Chemical & Engineering News
- ISSN 0009-2347
- Copyright © American Chemical Society