Designing Artificial Enzymes: Supported [FeFe]-Hydrogenase Mimics with Enhanced Catalytic Hydrogen Production and Oxygen Stability in Aqueous Media

Hydrogen is an attractive clean fuel for the storage and transport of energy generated by clean energy sources. It is energy dense, storable, and its use as a fuel produces no carbonaceous by-products whether used in a conventional means as a combustible fuel, or preferably, used in a fuel cell to directly generate electricity from the oxidation of molecular hydrogen. Currently, the vast majority of hydrogen is produced through steam reforming of natural gas. This process is useful for the production of hydrogen as a chemical feedstock, but requires very high temperatures (>700 °C) and utilizes natural gas as a feedstock, which is obtained via fracking and other extractive methods which are detrimental to the environment and human health. Further, stream reforming produces approximately ten tons of carbonaceous by-products (primarily CO2 and CO) for every ton of hydrogen produced. The hydrogen produced in this process also generally contains significant amounts of carbon monoxide, a contaminant known to poison the catalysts used in conventional hydrogen fuel cells. The other method used for hydrogen production is the electrolysis of water, in which protons from the water molecules are reduced with electrons supplied by an electrode to generate hydrogen gas which is of high purity – but it also comes at high cost. The electrolyzers used in this process depend on expensive nano-structured platinum catalysts for the cathodic reaction (that is, the reduction of protons to molecular hydrogen) which operates at essentially diffusion limited rates in acidic media.[1] Platinum is both expensive and rare, limiting its application and preventing the use of platinum cathodes in large scale production of hydrogen as a fuel. Accordingly, a large body of research has emerged around developing new catalysts using inexpensive, earth abundant materials to enable large scale production of hydrogen for the storage and transport of clean energy. Hydrogenase metalloenzymes are a class of naturally occurring enzymes which are capable of producing or metabolizing hydrogen gas in anaerobic bacteria as a metabolic process. There are several classifications of these metalloenzymes, the most active of which are the [FeFe]-hydrogenases, which has been found to catalyze the hydrogen evolution reaction at rates on the order to 104 molecules of hydrogen per second, with very low overpotential requirements.[2,3] Accordingly, mimicking this activity has been an active thrust of research for over 25 year, thanks in part to the synthetic accessibility of structurally similar [2Fe-2S] butterfly organometallic complexes, first synthesized over 50 years ago.[4,5] To date, hundreds of structural analogs of the active site of the [FeFe]-hydrogenases have been synthesized, but none have managed to replicate the performance of the enzyme.[6,7] More recently, some research groups have begun to focus on mimicking not only the active site of the enzyme, but also the macromolecular environment around the active site – in part because of a growing body of work which indicates the protein architecture around the active site plays several key roles in its stability and activity.[8–11] In our work, we were inspired to use polymer supports to improve the stability and activity of small molecule [2Fe-2S] complexes. Our first attempt to incorporate [2Fe-2S] complexes in a polymer matrix centered around the synthesis of oligothiophene-[2Fe-2S] complexes which we hoped could be electropolymerized into electrocatalytic films on the surface of electrodes. While these systems were not amenable to such an electropolymerization, we did discover an intriguing and extensive level of electronic communication between the oligothiophene ligands and the diiron system, which enabled their use as photo-catalysts for hydrogen evolution, without the need for an expensive external photosensitizer (such as an iridium or ruthenium complex, or cadmium-chalcogenide based quantum dots which are undesirable due to the use of highly toxic cadmium) which virtually all other photocatalytic [2Fe-2S] systems require. Undeterred, we developed a new system based in modern polymer chemistry – an initiator for a controlled radical polymerization was functionalized with a [2Fe-2S] moiety and successfully incorporated into several methacrylic metallopolymers and metallo(co)-polymers using atom transfer radical polymerization (ATRP). Extensive electrocatalytic studies on these polymers revealed that an amine rich, water soluble metallopolymer polydimethylaminoethylmethylmethacylate-graft-[2Fe-2S] (PDMAEMA-g-[2Fe-2S]) is able to catalyze the reduction of protons from neutral water with a low overpotential requirement (0.33 V to reach 0.1 mA cm-2 current density) extremely high turn over frequency (in excess of 200,000 molecules of hydrogen produced per second), is stable to operating voltages for up to six days with each molecule of catalyst producing approximately 40,000 molecules of hydrogen before becoming deactivated, and exhibits complete aerobic stability under optimized operating conditions – a feat which in unprecedented by either the enzyme or any other [2Fe-2S] mimic.[12–16] These catalytic figures of merit put it in a league of its own, as it has in many ways (rate, aerobic stability) surpassed the activity of the enzyme which inspired the work. To elucidate the reason for such astounding activity, we synthesized a second water soluble metallopolymer poly(oligoethyleneglycolmethylmethacylate-graft-[2Fe-2S] (POEGMA-g-[2Fe-2S]) which lack the amine functional groups of PDMAEMA. We found that while it is an active electrocatalyst, it is outperformed in every way by the PDMAEMA-g-[2Fe-2S] system. Random metallo(co)-polymers composed of approximately 50/50 and 70/30 ratios of DMAEMA/OEGMA monomers exhibited intermediate electrcatalytic activity in terms of overpotential, electrocatalytic current density (and therefore rate) and oxygen stability, with the POEGMA-g-[2Fe-2S] homopolymer losing all activity under ambient aerobic conditions. Finally, in a side project preformed in collaboration with the Heien lab, low order oligoethylenedioxythiophene (OEDOT) chains were found to form in the presence of Nafion upon slow evaporation of dilute solution of the two Nafion and EDOT in acetonitrile. Thorough characterization and a mechanistic investigation revealed that a cationic oligomerization mechanism is likely the cause for the formation of the OEDOT chains, and once they grown to a certain size, the formation of polaron on the OEDOT chain led to a strong electrostatic interaction between OEDOT and negatively charge Nafion, in which Nafion entangles the OEDOT chain and forms a stable colloidal dispersion in acetonitrile. These colloidal polymer particles were found to be sensitive probes for the water content of acetonitrile solutions, as they undergo an irreversible conformational change upon encountering water in solution, resulting in a bathochromic shift. We found dispersions of these particles are able to accurately and rapidly detect water concentrations between 125 and 2500 ppm using a simple UV-Vis measurement, at a cost of less than $2 USD per measurement.Release after 07/31/201