Water Splitting: Unexpected Fe(VI) Trapped by Manipulation of Reaction Kinetics

Abstract

Recently in Joule, Gray and co-workers have reported a pivotal and fully unexpected role of a very-high-valent Fe(VI) intermediate in water oxidation by NiFe oxyhydroxides. Recently in Joule, Gray and co-workers have reported a pivotal and fully unexpected role of a very-high-valent Fe(VI) intermediate in water oxidation by NiFe oxyhydroxides. Water oxidation, also denoted as water splitting or oxygen evolution reaction, is a key process in essentially all schemes for the production of non-fossil fuels. For electrocatalytic water oxidation in the alkaline regime, layered NiFe oxyhydroxides are prime candidates. They excel through high energetic efficiencies (low electrochemical overpotentials) as well as facile synthesis from none-rare-metal salts. Further and possibly knowledge-guided improvement of energetic efficiency and long-term stability is important for promoting the transition from the lab bench to large-scale application in chemical energy conversion systems, e.g., high-efficiency and low-cost water electrolysis for H2 formation. Therefore, various research groups are currently addressing this topical challenge; NiFe oxyhydroxides are clearly a major hotspot of research in chemical energy conversion. Their superior properties result from the synergy between nickel and iron ions; this is undisputed. However, the character of this synergy—specifically the mechanistic role of the Fe ions—has remained insufficiently understood and is highly disputed. Now, Harry Gray and his team at the California Institute of Technology (Caltech) present results that could promote a change of direction in the mechanistic discussion of and future research on water oxidation. Previously, the occurrence of Fe(IV) was contentiously discussed, and higher Fe oxidation states were not invoked. Recently in Joule, the Caltech team has provided strong experimental evidence for a reactive Fe(VI) intermediate, the subsequent formation of a peroxide side bound to Fe(IV), and eventually O2 formation directly coupled to the insertion of two water or hydroxide molecules.1Hunter B.M. Thompson N.B. Müller A.M. Rossman G.R. Hill M.G. Winkler J.R. Gray H.B. Trapping an iron(VI) water-splitting intermediate in nonaqueous media.Joule. 2018; 2https://doi.org/10.1016/j.joule.2018.01.008Abstract Full Text Full Text PDF Scopus (115) Google Scholar The research reported by the Caltech team has three remarkable strengths:(1)The truly unexpected involvement of a very-high-valent metal species resembling the Fe(VI)O42− ion provides a new twist to the mechanistic discussion of water oxidation—and not only for NiFe oxyhydroxides. This finding also encourages the consideration of other very-high-valent metal species in mechanistic studies, e.g., the permanganate ion, Mn(VII)O41−, as a putative reaction intermediate in manganese-containing catalyst materials.(2)Evidence for the Fe(VI) reaction intermediate comes from a broad and interesting combination of experimental approaches—too broad for detailed elaboration in a preview article. One example of an unusual experimental technique in electrocatalysis research is the verification of ferrate(IV) ions via the detection of far-red luminescence (at 1,613 nm).(3)The Fe(VI) reaction intermediate was trapped by depletion of the solvent of “substrate water.” Water depletion was realized with dry acetonitrile as the solvent. Although this is not a new approach per se, the successful application of this “trick” in water-oxidation research is remarkable. It inter alia illustrates the crucial role of reaction kinetics (and their smart manipulation) in the investigation of reaction intermediates, as highlighted below. Scheme 1 summarizes reaction-cycle intermediates of water oxidation by NiFe oxyhydroxide according to the results of Gray and co-workers.1Hunter B.M. Thompson N.B. Müller A.M. Rossman G.R. Hill M.G. Winkler J.R. Gray H.B. Trapping an iron(VI) water-splitting intermediate in nonaqueous media.Joule. 2018; 2https://doi.org/10.1016/j.joule.2018.01.008Abstract Full Text Full Text PDF Scopus (115) Google Scholar The shown reaction cycle involves two metal sites only: (1) a specific Fe ion that cycles between four oxidation states (Fe3+ to Fe6+) and (2) one further metal ion that changes its oxidation state by one unit in the course of the reaction cycle. The additional metal-ion redox pair (Ma/Ma+1 in Scheme 1) was not the focus of the study; it could be Ni2+/Ni3+, Ni3+/Ni4+, or Fe3+/4+. Scheme 1 uses S-state nomenclature, which was developed for describing the reaction cycle of photosynthetic water oxidation. The subscript indicates the number of accumulated oxidation equivalents (Si, where i = 0, 1, … 4). Applicability of this nomenclature already represents an important finding. It is applicable to the mechanistic proposal of the Caltech team only because their reaction cycle predicts the accumulation of four oxidizing equivalents before the onset of the O–O bond formation chemistry, as is also the case in oxygenic photosynthesis. Early peroxide formation, e.g., after the accumulation of only two oxidizing equivalents, is incompatible with their conclusions. Finding ways to characterize intermediates of the electrocatalytic reaction cycle is an exceedingly difficult task, but it is mandatory for mechanistic insight beyond the realm of mere hypotheses. Aside from well-designed experiments, a large amount of luck is required for the detection of a truly interesting reaction intermediate. The majority of intermediate states will often escape detection because of either kinetic or energetic reasons. For the application of catalytic potentials, typically only the intermediate state that precedes the rate-determining step of the reaction cycle (ki+1 is especially small) can easily be investigated, but even this state does not have a tag indicating its position in the sequence of reaction steps hinted at in Scheme 1. Highly reactive states are unlikely to become detectable because they are likely to be short lived (ki+1 is especially large). Moreover, interesting intermediates could be energetically high-lying states and thus undetectable because their population probability is too low. One useful approach that today is increasingly often pursued involves the application of suitable electric potentials in combination with direct tracking of the oxidation state and structural transitions of the catalyst material by in situ (in operando) spectroscopy or quasi-in-situ, freeze-quench techniques. Advanced in situ experiments involve X-ray, Raman, or infrared spectroscopy, the second and third of which were also part of the experimental toolbox used by the Caltech team. For fortunately spaced redox potential levels with Em1 < Em2 < Em3 < Em4, up to five intermediate states (S0 to S4) might become accessible. However, in real-world experiments, less fortunate Em spacing or kinetic factors typically limit the number of detectable states to two or three. And again, the detected states do not carry a tag for where to position them in the reaction sequence. For example, the highest detected metal-ion oxidation states could be Ni4+ and Fe4+, but this does not prove that a Ni4+Fe4+ site corresponds to the S4-state formed immediately before the onset of O–O bond formation. Gray and co-workers sought and found an alternative way to trap a highly interesting reaction-cycle intermediate.1Hunter B.M. Thompson N.B. Müller A.M. Rossman G.R. Hill M.G. Winkler J.R. Gray H.B. Trapping an iron(VI) water-splitting intermediate in nonaqueous media.Joule. 2018; 2https://doi.org/10.1016/j.joule.2018.01.008Abstract Full Text Full Text PDF Scopus (115) Google Scholar They blocked the O2 release step by depleting the electrolyte of the substrate of the water oxidation reaction by using a largely water-free acetonitrile electrolyte system. A cis-dioxo-iron(VI) intermediate was thereby trapped and shown by further experiments to be a reaction intermediate formed shortly before O2 release. The success of this blockage strategy was not a priori guaranteed. It relates to an important aspect of the reaction cycle: the site and state for binding of the two substrate water molecules (or hydroxides). The successful blockage implies that binding of the substrate water molecules for the next round of the reaction cycle is directly coupled to the O2-release step so that it can provide a decisive contribution to the driving force of O2 release. It moreover suggests that the Gibbs free energy of the S4′ and S4″ states in Scheme 1 is (slightly) higher than that of the S4 state so that the Fe6+ intermediate preferentially accumulates when water binding is prevented, as supported by the Caltech team by means of Raman detection of a peroxide intermediate along with the cis-dioxo-iron(VI) intermediate.1Hunter B.M. Thompson N.B. Müller A.M. Rossman G.R. Hill M.G. Winkler J.R. Gray H.B. Trapping an iron(VI) water-splitting intermediate in nonaqueous media.Joule. 2018; 2https://doi.org/10.1016/j.joule.2018.01.008Abstract Full Text Full Text PDF Scopus (115) Google Scholar Presently, it can hardly be predicted whether all the conjectures of Gray and co-workers will withstand the test of time. Their line of arguments is plausible, but some uncertainties and many major simplifications are still involved. Additional experimental and computational research is clearly required to elucidate the role of the NiFe oxyhydroxide “matrix” in facilitating the formation of a reactive Fe(VI) site. In any event, the Caltech team has reported inspiring experiments that could significantly affect future research on water oxidation. The stimulating atmosphere provided by the Cluster of Excellence UniCat (Unifying Concepts in Catalysis) is gratefully acknowledged. The UniCat cluster is financially supported by the Deutsche Forschungsgemeinschaft (EXC 314-2).