59th Annual Report on Research 2014

In work with Hollandite-type materials, we followed our previous report of K_1-xIr₄O₈ to prepare the series K_1-xRu₄O₈. As with the iridate, all samples were metallic with a weak magnetic response. After the two-step synthesis of K_0.5Ru₄O₈, which is non-trivial due to the volatility of potassium and iridium, we were able to systematically insert barium using a barium-benzophenone radical anion adduct prepared in-situ in THF, as demonstrated by the systematic change in lattice parameters (Figure 1 inset). On room temperature removal of potassium from KRu₄O₈ to form K_0.5Ru₄O₈, the a lattice parameter decreases and c increases; however, once barium is inserted to form K_0.5Ba_0.5Ru₄O₈, the lattice parameters return to almost the same values as found for KRu₄O₈. Initially this was surprising due to the differing electron counts, but can be explained as arising from the fact that potassium and barium have virtually identical ionic radii. While the physical properties of K_1‑xBa_yRu₄O₈ only vary slightly with electron count, these results demonstrate that it is possible to insert barium using organometallic chemistry. More generally, they offer an exciting possibility: by having the ability to prepare two isostructural compounds with the same lattice parameters but differing electron counts, it should be possible to disentangle the effects of electronic changes and structural changes on the physical properties; this is not normally achievable due to concomitant change of both in a typical solid solution.
Figure 2: Reaction progress of deintercalation of Na_2-xIrO₃ for two different initial conditions. There is a pronounced 'kink' in the progress, with or without stirring that changes with initial conditions.

In our explorations of the reaction mechanisms of low temperature soft chemistry, we carried out spectrophotometric monitoring of the kinetics of deintercalation of sodium from Na₂IrO₃ using iodine in acetonitrile (the system was chosen because of convenient timescales for the measurements). Representative data showing the reaction progress (i.e. x in Na_2-xIrO₃) as a function of time for two different initial values of the ratio [I₂]₀/[Na₂IrO₃]₀ are shown in Figure 2. There are two surprising features in the data. First is a pronounced 'kink' in the reaction progress, which occurs at the same value of x independent of the rate of solution stirring. Such a kink must arise due to a change in the rate limiting step of the reaction. Second, the initial rate of the reaction increases dramatically upon mixing, indicating that the pre-'kink' rate limiting step at least partly involves diffusion in the liquid solution, instead of being solely controlled by the rate of ion diffusion in the solid.
Figure 3: There is a near-perfect linear relationship between the initial concentration of iodide (relative to solid), [I₂]₀/[Na₂IrO₃]₀, and the value of x in Na_2-xIrO₃ at which the 'kink' occurs in the kinetics data.

A more detailed view of the 'kink' is obtained by plotting the value of x at which the 'kink' occurs as a function of the ratio [I₂]₀/[Na₂IrO₃]₀, Figure 3. There is a near-perfect linear relationship between the two parameters, with a slope and intercept indicating that the 'kink' occurs when the reaction reaches ~60% of what it would be at completion (each I₂ can remove two Na ions). We are presently completing experiments to understand the origin of this unusual behavior: does it arise due to triiodide formation? Or does it have to do with a switch to the rate limiting step in the solid state? Preliminary experiments using Br₂ in place of I₂ suggest triiodide is not important. But if it is instead the latter explanation, why does it vary with initial iodine content?

Support from the PRF has enabled our young group to generate initial results worthy of publication.  As previously reported, this research made grant applications more competitive, with success in obtaining public and private funds, including an NSF-CAREER award and a Packard fellowship for science and engineering. This work was a significant contributor to the awarding of the Exxon-Mobil Solid State Faculty Fellowship to the PI. The methods developed have been included in multiple additional proposals, most recently the successful renewal of the Department of Energy funded Institute for Quantum Matter at Johns Hopkins University. In addition, students have been exposed to synthetic solid state chemistry techniques and have gained familiarity with diffraction methods and electrical and magnetic property measurements, all of which are important skill sets for their future careers.  In short, this funding has allowed us to demonstrate the viability of soft chemistry methods for preparing metastable solid state materials. Three publications are expected directly from this work (one published, one under review, one in preparation), with indirect contributions to two more (both published).