A Synthetic Nickel Electrocatalyst with a Turnover Frequency above 100 000 s−1 for H2 Production

Abstract

The break-through article “A Synthetic Nickel Electrocatalyst with a Turnover Frequency Above 100 000 s⁻¹ for H₂ Production” by Monte L. Helm et al. (Science2011, 333, 863) presents results of great interest in the field of hydrogen production, in view of a fantastic turnover frequency obtained by an elegant catalyst design, inspired by nature, and the use of a largely available and cheap metal.

One of the thrusts of many research groups worldwide is the development of effective mimics of the enzymes that are involved in the production of energy through electron-transfer processes. To this class of enzymes belong the [FeFe] hydrogenases (Figure 1)1 that catalyze the formation of hydrogen (H₂) from water with an overpotential as low as 100 mV [Eq. (1)].((1))

• ((1))

Figure 1. Model of [FeFe] hydrogenase enzyme: the nitrogen base in an azadithiolate ligand is proposed to mediate the proton transfer to or from a coordination site at the distal iron atom. Adapted from Ref [1].

A crucial role in delivering or removing proton transfer to and from the correct atom(s) of the active site of the enzyme is played by the nitrogen atom of an azadithiolate ligand.

For many years, Daniel L. Dubois and co-workers have studied the activity of mononuclear nickel(II) complexes with nitrogen atoms in the second coordination sphere, demonstrating that such nitrogen atoms function as proton relays and can remarkably accelerate the rates of intra- and intermolecular proton transfers.2–6 Among the various molecular structures designed by Dubois and co-workers there are 1,5-diaza-3,7-diphosphacyclooctane ligands with different substituents on the phenyl groups attached to either nitrogen or phosphorus atoms (1),5 as well as an 1-aza-3,6-diphosphacycloheptane ligand (2).6 The reactions of such ligands with [Ni(CH₃CN)₆](BF₄)₂ give square-planar Ni^II complexes (Scheme 1) that catalyze the production of hydrogen, using protonated dimethylformamide [(DMF)H]OTf as proton source (reaction 1), at overpotentials and turnover frequencies depending on the type and molecular structure of the ligand.

In the recent Science paper by Dubois and co-workers6 is reported that the Ni^II complex 3 (Scheme 1) catalyzes reaction in Equation (1) with turnover frequencies of 33 000 s⁻¹ in dry acetonitrile and 106 000 s⁻¹ in the presence of 1.2 M of water at a potential of −1.13 V (versus the Cp₂Fe+/Cp₂Fe couple), which corresponds to an overpotential of approximately 625 mV. Such turnover frequencies are remarkably higher than that exhibited by the natural [FeFe] enzyme (9000 s⁻¹),7 as well as those obtainable with Ni^II complexes stabilized by the ligands of type 1 (350–1850 s⁻¹).4, 5

On the basis of theoretical and kinetic studies with the diaza complex 4, a mechanism has been proposed for which the rate-determining step (rds) for H₂ production is independent of acid concentration and involves one or more steps in the elimination of H₂ from the diprotonated Ni⁰ complex 6 (Scheme 2).5 Varying the electron-withdrawing character of the phenyl substituents on the nitrogen atoms of the ligand of type 1 has allowed the authors to establish important correlations between basicity of the nitrogen ligands and turnover frequencies/potential of the Ni^II/Ni^I couple at which catalysis occurs. Indeed, catalytic currents are observed at slightly more negative potentials of the one-electron reduction of the Ni^II centre in 4.

A similar mechanism, yet with a different rds, has been proposed for the production of hydrogen catalyzed by complex 3 (Scheme 3).6 In this case, the first-order dependence on acid concentration has led the authors to conclude that a one- or two electron reduction of 3 is followed by rate-determining protonation of the reduced species. Notably, the catalysis with complex 3 occurs at a potential (−1.13 V versus the Cp₂Fe⁺/Cp₂Fe couple) at which point Ni^I is reduced to Ni⁰. It is, therefore, unclear whether the rate-determining protonation involves either a Ni^I species, as proposed for the catalytic systems generated by ligands of type 1, or a Ni⁰ species. The formation of the diprotonated Ni⁰ intermermediate 7 sounds quite reasonable, yet it seemingly contrasts with the mechanism illustrated in Scheme 2. Given for granted that the catalysis occurs at the potential of the Ni^II/Ni^I couple, the formation of intermediate 5 becomes somewhat questionable, unless one makes the hypothesis that the Ni^I complex originated by the one-electron reduction of 4 gives a species that reacts with the acid to give a monoprotonated Ni^I intermediate whose Ni^I center is rapidly reduced at the same potential of the Ni^II precursor 4.

Despite the residual ambiguity on the proposed reaction mechanisms and the poor stability of the catalytic systems under real electrolysis conditions,5, 6 the results published by Dubois and co-workers are extremely interesting; it suffices to recall the fantastic turnover frequency of greater than 100 000 s⁻¹ for hydrogen production. Moreover, Dubois’ success presents opportunities to incorporate these Ni^II electrocatalysts into electrodes, so as to allow their straightforward use in electrolyzers. Recently, this approach has been successfully achieved with the complex [Rh(OTf)(trop₂NH)(PPh₃)], consisting of a diolefin-amide (trop₂N) and a triphenylphosphine (PPh₃), as ligands coordinated by a Rh^I center.8 Once embedded in a conductive carbon black, [Rh(OTf)(trop₂NH)(PPh₃)] has been used as anode electrocatalyst in direct ethanol fuel cells, which provides power densities that are comparable to those obtained with nanostructured noble metal catalysts.8 During the catalytic cycle of ethanol oxidation to acetate, this rhodium complex, which is a powerful alcohol dehydrogenation catalyst in homogeneous phase,9 is capable of evolving, to form mimics of alcohol and aldehyde dehydrogenases as well as proton and electron transfer enzymes.