A porphyrin-doped polymer catalyzes selective, light-assisted water oxidation in seawater
In a classic experiment, Naruta and co-workers demonstrated in 1994 that the dimanganese complexes 1 (Scheme 1) facilitate water oxidation catalysis yielding dioxygen (O2) at potentials above 1.2 V vs. Ag/AgCl. The corresponding, unconnected Mn-porphyrin monomers were, however, catalytically inactive. Subsequent work suggested that OO bond formation leading to O2 generation by 1 involved a concerted interaction between two short-lived, high-valent MnV=O intermediates at each of the porphyrins, presumably during conformational flexing of the dimer. The transience and brief lifetime of these intermediates likely rendered the free monomers inactive. Catalytic actions like those of 1, which encompass a synchronized, cooperative interplay between two or more catalytic groups, are of significant fundamental and practical interest. For example, enzymes are believed to employ synchronous protein motions to cooperatively harness reactive intermediates that are often too short-lived to be utilized in other classes of catalyst.[3–5] This may explain how they can catalyze some reactions that cannot be catalyzed outside of biology. The question that arises is: how can one design simple, practical abiological molecular catalysts to synchronously harness very short-lived reactive intermediates? In a previous report we described an approach to this problem that involved drastically concentrating the corresponding monomeric catalytic groups within a limited volume. This may conceivably cause some small but statistically significant proportion of the monomers to be adventitiously ideally placed to facilitate cooperative catalysis. If the monomer-bound reactive intermediates are too short-lived to be sequestered and exploited in any other way, then only the product deriving from cooperative catalysis should be obtained. Here we report the application of this “statistical proximity” approach to water oxidation catalyzed by Mn-porphyrins. We show that concentration of the sulfonated, monomeric Mn-porphyrin 2 (Scheme 1), which is normally catalytically inactive, within a thin layer of poly(terthiophene) (PTTh) yields a remarkable light-assisted catalyst with a low overpotential for water oxidation at pH 7. The catalyst selectively oxidizes water before chloride in seawater. Mn-porphyrin monomer 2 was uniformly incorporated as an anionic counter-ion into a thin PTTh film during the electrochemical polymerization of TTh monomer in ethanol/ dichloromethane (1:1 by volume) containing 2 (see Supporting Information). PTTh-2 was deposited as a composite film onto indium tin oxide (ITO) glass or flexible ITO-coated poly(ethylene terephthalate) (PET) sheet. Figure S1 (Supporting Information) shows the flexible electrode obtained when PTTh-2 was coated on ITO-PET. UV/vis measurements confirmed the incorporation of 2 in the coating. Energydispersive X-ray mapping indicated that 2 was uniformly dispersed in the coating (Figure S3). Elemental analysis indicated a high density of 2 within the PTTh-2, with the mole ratio of 2 (identified by Mn+S):terthiophene (identified by S) being ca. 1:3. The PTTh-2 films were then studied as putative working electrodes in photocatalytic oxygen generation from water. Cyclic voltammograms (CVs) of the PTTh-2/ITO glass electrode were taken with and without illumination using SoLux daylight MR16 halogen light bulbs (12 V, 50W, 248) in an aqueous 0.1m Na2SO4 electrolyte. Figure 1 depicts the data that was obtained. As can be seen, the CVs with and without light are significantly different. Substantially larger currents were observed positive of 0.68 V with illumination than without illumination. This is a region in which we have previously observed water oxidation in comparable systems. Moreover, the reduction peakAin Figure 1 was also observed only under illumination. In our experience, peaks of this type are often characteristic of adsorbed dioxygen. To study the peak at A, we conditioned the PTTh-2 coating in aqueous 0.1m Na2SO4, by maintaining it at 0.8 V (I), 0.9 V (II), or 1.0 V (III) for 1 h, and then immediately thereafter, performing a linear sweep voltammogram (LSV). The resulting LSV data are shown as the inset in Figure 1. As can be seen, under these conditions the broad peak at A resolves into two separate peaks: a large peak A’ and a small Scheme 1. Diporphyrin 1 and monomer porphyrin 2. Ar=4-tBuC6H4, 2,4,6-Me3C6H2, or C6F5.