MIT materials science graduate student Wesley Hong, Professor Yang Shao-Horn and colleagues reported in a paper published July 15, 2013 in The Journal of Physical Chemistry Letters that LCO nanoscale thin films had 100 times faster kinetics than bulk crystals of LCO. Collaborators at the University of Wisconsin-Madison previously proposed that the position of oxygen electronic states in an oxide could be used as a descriptor for the cathodic activity in solid oxide fuel cells, suggesting that the process could be tuned by adjusting metal-oxygen bond strength. “We would like to get a more fundamental insight into whether you can just tune the metal-oxygen bond without changing the chemistry or changing the effective charge on the cobalt ion and still change the catalytic activity,” Hong says, explaining the motivation for the study. “That way we can more directly see whether or not just changing the metal-oxygen bond strength … is relevant to catalysis.”
Working with thin films
The researchers introduced strain in LCO by growing thin films on top of Y2O3/ZrO2 (YSZ), which has a mismatched crystal lattice structure that forces a mechanical strain on the LCO, stretching the metal-oxygen bond and altering its bond strength and the cobalt spin state (spin state is the configuration of electrons around a transition metal ion).
Hong examined the thin films using electrochemical impedance spectroscopy to measure the oxygen surface exchange rate at high temperatures and compared the results for the thin film to published results for bulk single crystal and polycrystalline LCO. “When we compare those values, we see that there is up to two orders of magnitude enhancement in terms of the (oxygen) surface exchange rate. The surface exchange rate is basically a figure of merit for the catalytic activity of these oxides,” Hong says. Oxygen surface exchange — the ability to add and remove oxygen into a cathode oxide is one of the fundamental aspects of catalysis in solid oxide fuel cells. “If you have a weaker metal-oxygen bond, one might expect you could make it easier to remove oxygen from your crystal structure (and improve the exchange rate),” Hong says.
“The study is a fundamental look using a model system for understanding how to design better cathode materials. From this study, what we’ve determined is that metal-oxygen bond strength does play a crucial role in terms of tuning the catalytic activity of cathode materials inside solid oxide fuel cells, and in particular, we found that one of the ways that you can tune the metal-oxygen bond strength is by altering the spin state of the transition metal ion,” Hong says.
Measuring vibrational frequencies
Because cobalt’s electronic spin state isn’t easily observed, Hong measured it indirectly using the vibrational frequency of cobalt-oxygen bonds in the crystal lattice structure. In a compound such as LCO, bonds are formed by electrons between the atoms of cobalt, which is a metal, and oxygen. Those metal-oxygen bonds can be likened to springs. "If you perturb the spring and set it into a vibration, the frequency of that vibration is related to the stiffness of the spring. If you have a really soft spring, you're going to have very low frequency vibration, whereas if you have a very stiff spring, you're going to have a very fast vibration, so higher frequency. We can study the bond frequencies and use that to get an idea of the bond strength between the thin film system versus the polycrystalline system,” Hong explains.
“The thin films have a much lower vibrational frequency associated with them than in bulk, which is indicative of a weaker metal-oxygen bond strength,” he says.
Hong did Raman spectroscopy across a wide range of temperatures to measure the temperature dependence of the metal-oxygen vibrational frequency, and used an analytical model to correlate those measurements to changes in the spin state energetics for thin films versus bulk LCO. Experimental results show that what is known as the “breathing mode,” in which oxygen atoms surrounding a cobalt ion vibrate away from the cobalt center, decreases in frequency as temperature increases. Sharp decreases in the vibrational frequency occur at lower temperatures in thin films than in bulk LCO. Such sharp decreases are indicative of spin state transitions, suggesting less energy is necessary in thin films to access spin states that weaken the metal-oxygen bond.
“This is a report of outstanding and timely work on tuning the spin state in oxide thin films by epitaxial strain," says Professor Ho Nyung Lee, Thin Films and Nanostructures group leader with the Oak Ridge National Laboratory in Tennessee. "While many previous efforts have focused only on chemical control of d-band electron population, this paper clearly demonstrates that a small local structural modification can induce a large enhancement in the catalytic surface activity.”
Lee, who was not involved in the research, says, “The dazzling result stems from a systematic understanding of the role of local symmetry evolution by strain that critically influences how the complex d-band electrons populate. Therefore, it is clear that this work offers a new design strategy for many energy materials, including SOFC cathodes and oxygen membranes.”
The study sets a baseline for exploring spin state in other metal oxide systems. “We found that just by tuning the metal-oxygen bond strength without changing the overall chemistry of the material, we can actually create significant enhancements in the catalytic activity, so that allows us to identify a principle to use in order to identify better materials,” Hong says.