Climate change due to carbon dioxide emissions is one of the biggest challenges humanity faces. A critical technology for replacing fossil fuels is artificial photosynthesis, which takes a cue from its natural counterpart to convert solar energy into chemical fuels, but uses material surfaces rather than biology to do the work. Here at Berkeley, chemistry professor and Lawrence Berkeley National Lab researcher Tanja Cuk, in collaboration with colleagues at LBNL’s Molecular Foundry, is revealing the previously hidden initial catalytic steps of solar-to-fuel conversion on material surfaces.
At a very broad level, artificial photosynthesis can be divided into two parts. The first part is light-driven water oxidation, where charge created by energetic photons breaks bonds between hydrogen and oxygen in water molecules (H2O), releasing electrons and allowing oxygen atoms to combine to form O2, oxygen gas. The second part, CO2 reduction, uses the liberated electrons to convert carbon dioxide into more energy-dense molecules like methanol.
Cuk is focusing on the initial steps of the water oxidation reaction. The more she and other researchers can discover about these fundamental physical processes at the microscopic level, the more quickly they can eventually be scaled up for renewable fuel production.
Cuk says, “While several theories have been proposed about how the microscopic steps of water oxidation might take place, experimentally defining them has been a great challenge. We set out to investigate the earliest step of catalysis on surfaces, or how charge within the material can induce the transformation of a molecular species at the surface — a catalytic intermediate — that can form a new bond.”
Cuk investigates a semiconductor photovoltaic (PV) and catalytic system that is tractable for both advanced theoretical and experimental techniques, one that lets the insights from both build upon each other. Her experiments must measure data with an extreme sensitivity on very short time scales — less than a trillionth of a second — using short pulses of laser light. As simple as the measured material is, the atomic scale details of the associated laser-excited wet interface are so complex that LBNL’s supercomputers are required to help interpret the results theoretically.
At the atomic level, the charge created by photons in the PV material is localized in trillionths of a second into a catalytic intermediate at the surface. As diagrammed in Figure 1, in the experiment a strontium titanate crystal (black) is excited by UV laser light (yellow wedge) representing solar energy input. The result is charge separation — negatively charged electrons and positively charged holes. This is not unlike how sunlight striking a solar panel creates electricity.
Fig. 1: A strontium titanate crystal (black) is excited by UV laser light (yellow wedge). The laser light is produced in extremely short bursts lasting about 10 trillionths of a second. After the excitation pulse the surface is then probed by infrared light creating evanescent wave (small red wedge) which highlights surface vibrations.
The difference is that unlike sunlight, the UV laser light is produced in extremely short bursts lasting about 100 femtoseconds (10-13 sec, or one tenth of a trillionth of a second.) These short pulses allow the initial steps of the reaction to be tracked precisely from their earliest moments. At a predetermined time interval after the excitation pulse, the surface is then probed by infrared light, which is sensitive to molecular vibrations. The geometry of the infrared probe creates what is called an evanescent wave (small red wedge), which highlights surface rather than bulk vibrations.
The results of the experiments and theoretical analysis are described using the illustration of light-driven water oxidation shown in Figure 2. From left to right, the first image shows a side view of the strontium titanate crystal, with water absorbed on the surface of titanium atoms (blue) in the uppermost layer, bound in a lattice with oxygen (red) and strontium (green) atoms. After a blast of the UV light, positively charged holes migrate to the surface, where they are captured by the titanium hydroxide (Ti-OH) on the crystal/water interface.
Fig. 2: From left to right, the first image shows a side view of the strontium titanate crystal, with water absorbed on the surface of titanium atoms (blue) in the uppermost layer, bound in a lattice with oxygen (red) and strontium (green) atoms. After a blast of the UV light, positively charged holes migrate to the surface, where they are captured by the titanium hydroxide (Ti-OH) on the crystal/water interface.
In the second image from left, a very reactive titanium-oxyl radical (Ti-O•) is created. The formation of the titanium-oxyl radical can be tracked using infrared (IR) spectroscopy, which is sensitive to an associated vibration of the Ti-O bond just below it in the lattice (dark blue arrow). In the third image two of these intermediates split the water molecule, and in the final right-hand image the singly bonded O-O is formed.
In the second image from left, a very reactive titanium-oxyl radical (Ti-O•) is formed with one electron missing from the oxygen’s outer orbital. Cuk and co-workers found that the formation of the titanium-oxyl radical can be tracked using infrared (IR) spectroscopy, which is sensitive to an associated vibration of the Ti-O bond just below it in the lattice (dark blue arrow). The vibrational signature or “ping” of this buried vibration is unique to the titanium-oxyl radical, and so, by “listening” for it using IR spectroscopy, Cuk could tell when the radical is formed.
Identifying the buried vibration as the signature of the oxyl radical had not been previously suggested or predicted. It was discovered when Cuk enlisted the support of LBNL’s Molecular Foundry (foundry.lbl.gov) to develop some theoretical insight.
Foundry researchers Das Pemmaraju and David Prendergast helped interpret Cuk’s experiments using theoretical models of the photoexcited states at the crystal-electrolyte interface. These complex models required calculations on supercomputers at LBNL’s National Energy Research Scientific Computing Center (www.nersc.gov). The theoretical results provided an atomic scale insight as to the origin of the measured signature in Cuk’s experiment — a Ti-O vibration located just below the photoexcited oxyl radical.
A surprising discovery was made in identifying the oxyl radical. Namely, Cuk and co-workers found that its associated buried vibration weakly resonates with the rocking motions of the water in the electrolyte, as cartooned in Figure 3. These rocking motions make and break the complex hydrogen bonds in the water. Since catalytic intermediates should disturb reactant molecules in order to form new bonds, this resonance had been anticipated, but these experiments are the first to detect it.
Fig 3: The buried vibration described in Figure 2 weakly resonates with the rocking motions of the water in the electrolyte, as cartooned in this illustration. These rocking motions make and break the complex hydrogen bonds in the water. (Image credit: Nature Chemistry)
The coupling is shown through an analysis of the asymmetric spectral lineshape (Figure 4) known as a “Fano resonance.” The hope is that experimentalists can now use this coupling to sense how water motions are involved in creating the first bond in the water-oxidation cycle. The next stages of the cycle are suggested in the third image of Figure 2, where two of these intermediates split the water molecule, and in the final right-hand image, where the singly bonded O-O is formed. After a few more steps, it will produce O2.This collaboration between experiment and theory lays the basis for future work on detecting surface-bound intermediates by their unique, buried vibrations. Explains Cuk, “By finding a way to put the lens on how charge in the material changes surface bonds, new spectral signatures unique to catalysis — such as a catalytic intermediate’s buried vibration and its resonance with water motions — have been found.”
Fig. 4: The coupling described in Figure 3 is shown through an analysis of the asymmetric spectral lineshape known as a "Fano resonance." This is a type of resonant scattering phenomenon where interference between a background and a resonant scattering process produces the asymmetric line-shape.
Such a microscopic picture of water-splitting reactions arms researchers with new insights into designing artificial photosynthetic systems for improved efficiency and scalability. This in turn can help create solar fuels that have the potential to replace fossil fuels and help reduce the rising level of carbon dioxide in the atmosphere.
The two graduate students that constructed the experimental setup and collected the data are David Herlihy (first author) and Xihan Chen. The postdoctoral fellow that also collected data and helped execute the project is Matthias Waegele (now at Boston College). Das Pemmaraju and David Prendergast of LBNL’s Molecular Foundry created theoretical models of the photoexcited states and performed the necessary supercomputer calculations at NERSC.
Funding was provided by the Air Force Office of Scientific Research and Department of Energy (DOE) Office of Basic Energy Sciences.
For more information see these articles in Nature Chemistry: