The 'blue copper' enzyme bilirubin oxidase from Myrothecium verrucaria shows significantly enhanced adsorption on a pyrolytic graphite 'edge' (PGE) electrode that has been covalently modified with naphthyl-2-carboxylate functionalities by diazonium coupling. Modified electrodes coated with bilirubin oxidase show electrocatalytic voltammograms for the direct, four-electron reduction of O(2) by bilirubin oxidase with up to four times the current density of an unmodified PGE electrode. Electrocatalytic voltammograms measured with a rapidly rotating electrode (to remove effects of O(2) diffusion limitation) have a complex shape (an almost linear dependence of current on potential below pH 6) that is similar regardless of how PGE is chemically modified. Importantly, the same waveform is observed if bilirubin oxidase is adsorbed on Au(111) or Pt(111) single-crystal electrodes (at which activity is short-lived). The electrocatalytic behavior of bilirubin oxidase, including its enhanced response on chemically-modified PGE, therefore reflects inherent properties that do not depend on the electrode material. The variation of voltammetric waveshapes and potential-dependent (O(2)) Michaelis constants with pH and analysis in terms of the dispersion model are consistent with a change in rate-determining step over the pH range 5-8: at pH 5, the high activity is limited by the rate of interfacial redox cycling of the Type 1 copper whereas at pH 8 activity is much lower and a sigmoidal shape is approached, showing that interfacial electron transfer is no longer a limiting factor. The electrocatalytic activity of bilirubin oxidase on Pt(111) appears as a prominent pre-wave to electrocatalysis by Pt surface atoms, thus substantiating in a single, direct experiment that the minimum overpotential required for O(2) reduction by the enzyme is substantially smaller than required at Pt. At pH 8, the onset of O(2) reduction lies within 0.14 V of the four-electron O(2)/2H(2)O potential.

The special properties of O2-tolerant [NiFe]-hydrogenases make it possible, in principle, to operate all-enzyme hydrogen fuel cells. These devices show unusual power characteristics, as revealed in a series of experiments in which the O2-tolerant hydrogenase (Hyd-1) from Escherichia coli is used as H2-oxidation catalyst (anode) and a bilirubin oxidase is used as O2-reduction catalyst (cathode). In a fuel cell adaptable for variable fuel and oxidant supply, three limiting conditions were examined: (1) the anode and cathode separated by a Nafion membrane and 100% H2 and 100% O2 fed to the separate compartments, (2) a membrane-free mixed feed cell with a fuel-rich (96% H2) hydrogen/oxygen mixture, and (3) a membrane-free mixed feed cell with a fuel-weak (4% H2) hydrogen/air mixture. Condition (1) exposes the effect of O2-crossover which is evident even for an O2-tolerant hydrogenase, whereas condition (2) is limited by bilirubin oxidase activity on the cathode. Condition (3) yields power only under high-load (resistance) conditions that maintain a high output voltage; a low load collapses the power (akin to a circuit breaker) because of complete inactivation of the [NiFe]-hydrogenase when subjected to O2 at high potential. Recovery of the hydrogen-poor fuel cell is not achieved simply by restoring the high load but by briefly connecting a second anode containing active hydrogenase which discharges electrons to provide a jump start. The second anode had remained active despite being in the same O2 environment because it was not electrochemically connected to an oxidizing source (the cathode), thus demonstrating that, under 4% H2, the presence of 20% O2 does not, alone, cause hydrogenase inactivation, but simultaneous connection to an oxidizing potential is also required. The investigation helps to illuminate obstacles to the application of hydrogenases in fuel-cell technology and suggests phenomena that might be relevant for biology where biological membranes are engaged in H2 oxidation under aerobic conditions.

No abstract