Reports: ND350971-ND3: Electrocatalytic Water Oxidation by Manganese Pyridinophane Complexes

Jeremy M. Smith, Indiana University

1. Electrochemical water oxidation

We previously reported that the cyclic voltammogram (CV) of (Py2NtBu2)Mn2+ (Fig. 1, R = tBu) in reveals a catalytic wave for water oxidation in basic solution (Fig. 1). Dioxygen formation was directly characterized by a number of methods, including cyclic voltammetry, gas chromatography and an O2 electrode, as well as indirectly through the pH change of unbuffered solutions. In addition, multiple experiments suggest that catalysis is homogeneous. Finally, the catalytic current varies linearly with the concentration of the Mn complex (Fig. 2, inset), consistent with a mononuclear catalyst for water oxidation.

A paper reporting these results was initially submitted in July 2013. In the course of multiple rounds of revision, we conducted additional experiments that resulted in a better estimation of the catalyst longevity. No catalysis is observed in buffered solutions, possibly because the ions required for these high pH buffers (e.g. phosphate) can chelate the cis-divacant Mn center and thereby prevent substrate binding. Since catalytic oxygen formation in unbuffered solutions decreases the pH and terminates catalytic activity, we have conducted controlled potential electrolysis experiments in which the solution pH is periodically readjusted to 12.2. This strategy allows us to better determine the catalyst longevity, and we observe turnover numbers (TON) of 19-24 with Faradaic efficiencies of 74 – 81 %. This performance compares well with the best homogeneous Mn-containing water oxidation catalysts.

Fig. 1. Cyclic voltammogram for (Py2NtBu2)Mn2+ showing a catalytic current at high potential. Experimental conditions: pH = 12.2, 0.1 M KOTf, FTO working electrode. Inset: Dependence of the catalytic current on catalyst concentration.

We have initiated a collaboration with my colleague Mu-Hyun Baik to determine the catalytic mechanism through a combination of experimental and computational methods. Surprisingly, the computed thermodynamics for electron and proton transfer involving water-derived ligands on the Mn complex are not consistent with our experimentally determined data. This discrepancy may have its origin in the irreversibility of all waves in the cyclic voltammogram of the complex. Due to this irreversibility, the measured Ep,a may be quite different from the true thermodynamic E1/2 values. Fortunately, the Baik group has recently determined a method for extracting E1/2 the scan rate dependence of irreversible waves. Initial application of this method gives E1/2 = 1.07 V vs SHE for the first oxidation wave (pH 12.2). We are in the process of applying this method for all pH values to redetermine the Pourbaix diagram. 2. Electrochemical water reduction

We previously reported that (Py3NH3)Mn2+ (Fig. 2a) shows catalytic activity for electrochemical water reduction in mildly acidic aqueous conditions. Although the overpotential is very large (ca. 1.4 V at pH 5), catalytic water reduction by Mn is very unusual, warranting further investigation. We have now completed a comprehensive study into the water reduction catalysis by this complex.

Catalysis is observed over a pH range of 4-6.5, with evidence for saturation in the proton concentration. The catalytic current is proportional to the concentration of the complex, consistent with a mononuclear catalyst (Fig. 2b). Controlled potential electrolysis in an unbuffered solution causes the pH to increase and leads to the formation of hydrogen, as confirmed by gas chromatography. Evaluating the catalyst performance is complicated by apparent adsorption of the complex on the electrode, thus making comparison with the catalyst-free water reduction difficult.

Although cyclic voltammetric measurements of the free ligand Py3NH3 reveal that it is also a proton reduction catalyst, it is apparent that the presence of Mn has a measurable, albeit small influence on the catalytic performance. Thus, while the two species have the same overpotential for electrocatalysis, the Mn complex has a greater peak current than does the free ligand under the same conditions (Fig. 2c). We speculate that the ligand is the critical element for catalysis and that the Mn may preorganize the conformation to facilitate the key bond breaking and forming events. We plan test this hypothesis by measuring catalysis in the presence of other metal ions.

(a)

(b)

(c)

Fig. 2. Electrocatalyic proton reduction (a) Proposed structure of (Py3NH3)Mn2+ in aqueous solution; (b) Linear dependence of peak catalytic current on [(Py3NH3)Mn2+]; (c) Cyclic voltammograms showing the similar electrocatalytic behavior of [(Py3NH3)Mn2+] and the Py3NH3 ligand. Unless otherwise indicated, all electrochemical measurements were made at pH 5, 100 mM KOTf electrolyte and 1 mM catalyst. 3. Career Impact and Student Impact

As mentioned above, the PI has developed new collaborations that will enhance his contribution to electrocatalysis. After a year of review and revision, our initial paper on water oxidation was recently published in Angew. Chem. Int. Ed.

A graduate student will continue to be employed on the project for part of the upcoming year. This research assistant support will allow more time to be devoted to the project.