Standard Article

You have free access to this content

Hydrogenases and Alternative Energy Strategies

Part 3. Biocatalysis

  1. Olaf Rüdiger1,
  2. António L. De Lacey1,
  3. Victor M. Fernández1,
  4. Richard Cammack2

Published Online: 15 MAR 2010

DOI: 10.1002/9783527628698.hgc032

Handbook of Green Chemistry

Handbook of Green Chemistry

How to Cite

Rüdiger, O., De Lacey, A. L., Fernández, V. M. and Cammack, R. 2010. Hydrogenases and Alternative Energy Strategies. Handbook of Green Chemistry. 3:8:213–242.

Author Information

  1. 1

    Instituto de Catálisis, CSIC, Madrid, Spain

  2. 2

    King's College London, Department of Biochemistry, London, SE1 9NH, UK

Publication History

  1. Published Online: 15 MAR 2010

Chemistry Terms

Choose one or more boxes to highlight terms.

Abstract

  1. Top of page
  2. Introduction: The Future Hydrogen Economy
  3. Chemistry of Hydrogenase Catalytic Sites
  4. Experimental Approaches
  5. Catalytic Mechanisms of Hydrogenases
  6. Progress So Far with Biological Hydrogen Production Systems
  7. Conclusion and Future Directions
  8. References

This chapter considers the hydrogenases, the enzymes that produce and consume molecular hydrogen; how they work and how they might be exploited in a future hydrogen-based economy. Hydrogenases are enzymes containing iron and in some cases nickel, which efficiently couple the proton-hydrogen couple to the reduction or oxidation of electron carriers. They have many metabolic functions in the microbial world. Much has been learned from recent research about their active sites and mechanism of action. There are three known types of hydrogenase chemistry: that based on a dinuclear nickel-iron site (NiFe-and NiFeSe-hydrogenases); on more complex iron-containing clusters (FeFe-hydrogenases); and an enzyme containing a mononuclear iron site, which catalyzes a specific direct hydrogenation of a cofactor (Hmd). Variations on the NiFe-hydrogenases are involved in the sensing of H2 in cell regulation. Where necessary, hydrogenases have structural features that avoid the inhibitory effects of oxygen, carbon monoxide and other pollutant gases which poison other catalysts such as platinum. The structures of the different types of hydrogenases have been determined and their catalytic mechanisms investigated by spectroscopy and electrochemistry. Genetic and biochemical studies are elucidating the complex metabolic pathways by whereby the different types of hydrogenase catalytic centers are assembled. Three possible energy strategies are suggested by hydrogenase research: (1) hydrogen biotechnology, using microorganisms to produce hydrogen by fermentation of organic wastes, with or without the assistance of photosynthesis; (2) biomimicry, using insights from the structures and mechanisms of hydrogenases to synthesize improved catalysts as a substitute for the limited resources of precious metals; and (3) harnessing hydrogenases on carbon electrodes in electrolysis cells for H2 production or fuel cells to produce electricity from hydrogen at low concentrations and with low overpotentials and at rates comparable to those with platinum.

8.1 Introduction: The Future Hydrogen Economy

  1. Top of page
  2. Introduction: The Future Hydrogen Economy
  3. Chemistry of Hydrogenase Catalytic Sites
  4. Experimental Approaches
  5. Catalytic Mechanisms of Hydrogenases
  6. Progress So Far with Biological Hydrogen Production Systems
  7. Conclusion and Future Directions
  8. References

With the increasing uncertainty about the availability of fossil fuels and concerns about CO2 emissions and their effects on climate, a sustainable energy source is essential. Hydrogen is widely considered to be a viable convertible form of energy for the future. It is readily transported through pipelines and can be stored, with some energy loss, in the form of compressed or liquefied gas. A great advantage is the high conversion efficiency into electricity in fuel cells, in comparison with the inevitable thermodynamic losses in heat engines. At present, only limited amounts of hydrogen are produced commercially, mainly by steam conversion of fossil fuels (1, 2). However in a future sustainable economy, hydrogen could be produced by electrolysis of water (Figure 8.1). A major limitation to be overcome is the cost of the catalysts. New catalysts are being developed for the consumption and production of hydrogen (3) but those used in fuel cells currently rely on precious metals such as platinum or their alloys (4). The known reserves of the noble metals in the lithosphere represent a limitation to their continued use (5). Moreover these catalysts have problems of side-reactions with O2 and sensitivity to gases such as carbon dioxide, carbon monoxide and sulfide (6, 7).

original image

Figure 8.1. Outline of renewable energy resources involving hydrogen. Shaded boxes indicate processes that currently require precious metals such as platinum and could be substituted by hydrogenases.

In the biosphere, hydrogen has been an energy carrier for billions of years. It is efficiently consumed and produced by microorganisms, in electron-transfer reactions catalyzed by the hydrogenases. These enzymes harbor iron- and nickel-containing clusters with the special property of reacting with or producing H2, according to the following simplified general reaction:

mathml alt image(8.1)

where e represents reducing equivalents, usually from a suitable electron donor protein. They are highly efficient catalysts. They catalyze the oxidation and production of hydrogen at rates comparable to those with metals such as platinum, and, moreover, with a low overpotential (8, 9). Where necessary, hydrogenases can be resistant to inhibition by oxygen and carbon monoxide (10).

The hydrogenases, combined with photosynthesis, offer the tantalizing prospect of carbon-neutral H2 production from renewable resources and conversion to electricity (11). It has been known for many years that some green algae can produce hydrogen and oxygen by photolysis of water using only sunlight (12), the ultimate clean energy technology. This has spurred the study of the chemistry, mechanism and applications of hydrogenases. Over the past 40 years, a great deal has been learned about their structure and chemistry, how they are made and how they are used in biological processes. Several different lines of approach have been made for the application of hydrogenases for hydrogen energy:

  • The harnessing of hydrogen-producing microorganisms to produce hydrogen. This is somewhat equivalent to the domestication, selective breeding for human exploitation, a process which for plants and animals took place over thousands of years. There are two major approaches: by anaerobic fermentation of organic wastes and as a byproduct of oxygenic or anaerobic photosynthesis.

  • The design of biologically inspired catalysts or biomimicry, based on the chemical structure and mechanism of hydrogenases.

  • The application of hydrogenases as catalysts in the reversible interconversion of H2 and electricity. There has been recent success in connecting or ‘wiring’ hydrogenases to carbon electrodes as a possible substitute for precious metal electrodes. This application will be a major topic of this chapter.

8.1.1 Biological Hydrogen Energy Metabolism The outline scheme of hydrogen metabolism in the biosphere is shown in Figure 8.2. Eukaryotes that can produce H2 include certain algae, protozoa and fungi, but not higher animals or plants. Unlike the hydrogen energy technologies outlined in Figure 8.1, in biological systems hydrogen gas is not allowed to accumulate, since it is usually assimilated, either by the organism that produces it or by other microorganisms, a process known as interspecies hydrogen transfer (13). H2-producing microorganisms are not restricted to normal temperatures or pH; nor do they need pure water. Hyperthermophiles can grow at temperatures up to 100 °C and their hydrogenases operate optimally at these temperatures. Species can be found that grow in strongly saline environments and over a pH range from strongly acidic to alkaline (the internal pH of the cell remaining near 7, however). They can also be tolerant to heavy metals.

original image

Figure 8.2. Hydrogen metabolism in the biosphere. CH2O represents fermentable substrates from biomass, such as carbohydrates.

Hydrogen is produced in anaerobic fermentations, in organisms such as Clostridium species and Escherichia coli (14). When there is an excess of organic reductant, dissolved H2 may be released as a metabolic byproduct.

In biological systems, H2 is produced and used locally; the transport distance is often no further than the micrometre distance between neighboring cells. In biological systems H2 occurs not as a gas, but in solution and at atmospheric pressure the maximum concentration is only 2 mM. If reducing equivalents are stored, it is in the form of more stable metabolites such as polysaccharides or lipids, which are readily produced and interconverted by enzymic systems. Poly-3-hydroxybutyrate, another storage compound, is interesting as a potentially valuable, biodegradable material (‘bio-plastic’), which is produced as an alternative to H2 (15).

Nitrogenases are another source of biogenic hydrogen. They are microbial enzymes used in nitrogen fixation, that catalyze the six-electron reduction of nitrogen to ammonia, using adenosine triphosphate, according to the reaction

mathml alt image(8.2)

where e represents an electron donated by either ferredoxin or flavodoxin. The production of at least one H2 appears to be an obligatory part of the reaction, but nitrogenases also produce further H2 depending on the conditions, even when N2 is excluded.

In contrast to nitrogenase, hydrogenases operate almost at thermodynamic equilibrium, requiring only a sufficiently low redox potential to convert protons to H2. In hydrogen-oxidizing bacteria such as Ralstonia eutropha they are able to oxidize H2, even at low concentrations, in air.

H2 produced by anaerobic fermentation is reoxidized by autotrophic organisms. The oxidants are compounds such as O2, nitrate, sulfate and CO2, which are reduced to H2O, nitrogen oxides or N2, H2S or CH4, respectively (Figure 8.2). In each case a hydrogenase is required to react with H2, according to Equation 8.2, producing H+ and e. The reducing equivalents, e, are oxidized in membrane-bound protein complexes, generating transmembrane gradients, which are then dissipated by the F0/F1 ATPase to generate ATP (16).

8.2 Chemistry of Hydrogenase Catalytic Sites

  1. Top of page
  2. Introduction: The Future Hydrogen Economy
  3. Chemistry of Hydrogenase Catalytic Sites
  4. Experimental Approaches
  5. Catalytic Mechanisms of Hydrogenases
  6. Progress So Far with Biological Hydrogen Production Systems
  7. Conclusion and Future Directions
  8. References

H2 is inherently an extremely stable molecule (17). Despite the low redox potentials that are generated in the metabolic pathways of photosynthetic and fermentative bacteria, only hydrogenases, nitrogenase and a few enzymes that have some characteristics in common with hydrogenases (notably, the nickel-iron-sulfur protein carbon monoxide dehydrogenase and the complex iron-sulfur protein pyruvate:ferredoxin reductase (18)) have been reported to reduce protons to H2. Of these, the hydrogenases are by far the most efficient (8). The biological role, properties and applications of hydrogenases have been described in an edited volume (19) and recently in a thematic issue of Chemical Reviews on hydrogen (17, 20-24).

As can be surmised from Figure 8.2, hydrogenases in different organisms are able to transfer electrons to and from many different substrates, depending on the conditions of growth. An organism may contain several different hydrogenases, which are expressed under different conditions and which have different specificities for electron donors and acceptors. The reactions with acceptors and donors are all thermodynamically reversible reactions, but some hydrogenases catalyze the reaction much more rapidly in one direction than others, reflecting their cellular function. Hence there are uptake, production and bidirectional hydrogenases. Many hydrogenases have a modular construction, with domains that resemble electron-transfer proteins from other redox enzymes (25, 26), resulting in electron transfer chains for different electron donors and acceptors (Figure 8.3).

original image

Figure 8.3. Hydrogenase general working scheme, where M is Ni for the [NiFe] hydrogenase or Fe for the [FeFe] hydrogenase.

Hydrogenases have evolved for efficiency in hydrogen production or consumption and a number of common features have been selected. There are three types, based on the organometallic complexes in hydrogenases. All of them contain iron with CO ligands, almost all contain iron–sulfur clusters and some of the most efficient enzymes also contain nickel. The type of active site that is employed in any biological environment may be partly a consequence of the availability of nickel as a trace metal in the environment and partly a consequence of evolution. Nickel–iron-containing hydrogenases are found in the Bacteria and Archaea, whereas iron-containing hydrogenases are found in Bacteria and Eukaryotes such as green algae, fungi and protozoa.

8.2.1 NiFe The active sites of NiFe hydrogenases comprise a dinuclear center, with an Ni ion with cysteine sulfur coordination, bridged to a low-spin Fe(II) which has CO and CN ligands (Scheme 8.1a). This structure was revealed by X-ray crystallography of the hydrogenase of D. gigas (27, 28) and is representative of many nickel-containing hydrogenases. The dinuclear center is stable in different redox states, some of which are paramagnetic (Scheme 8.2). The oxidized states of the hydrogenase contain an interesting bridging ligand, which in the first crystallographic structure was somewhat disordered (27), but which is now believed to represent an oxygen-derived species (29). Current mechanisms favor the view that this bridging position is the location for a hydride species in the catalytic cycle (23, 30).

original image

Scheme 8.1. Schematic representation of the different hydrogenase active sites. (a) NiFe-hydrogenase; the active site is coordinated to the enzyme through two bridging and two terminal cysteines. The X ligand can be O for the oxidized species, CO for the inhibited forms or H in the catalytic cycle. (b) NiFeSe-hydrogenase has selenocysteine instead of the equivalent cysteine to Cys530 of D. gigas NiFe-hydrogenase. (c) H cluster from the FeFe-hydrogenases. X is still unknown, but N or C is suspected. (d) Predicted structure of the active site iron of the iron-sulfur cluster-free hydrogenase (Hmd).

original image

Scheme 8.2. Scheme of reactivity and redox states of the active sites of standard NiFe-hydrogenases. The paramagnetic EPR-active states are marked with an asterisk. The formal redox potentials, Ea, correspond to those measured by FTIR-spectroelectrochemistry of D. gigas hydrogenase (169). The catalytic cycle of the enzyme is proposed to involve cycling between the states Ni-SI, Ni-C and NiR, then back to Ni-SI with the release of H2. States Ni-A and Ni-B, labeled with an asterisk, are paramagnetic, oxidized forms which become active on reduction (see text). Enzyme in the Ni-C state is light sensitive at cryogenic temperatures, becoming converted to another state, Ni-L. Other states are produced by reversible inhibition with CO. For more details, see (22).

Most NiFe-hydrogenases lose activity in the presence of O2. In the case of D. gigas hydrogenase and similar hydrogenases, two paramagnetic oxidized forms are produced, which contain nickel in the formal oxidation state Ni(III). These were named the Ni-A nor Ni-B states, after their EPR signals. Both are inactive towards H2, but they can be reduced to EPR-silent states, Ni-SU and Ni-SII, respectively, which are still unreactive with H2, but which can then recover their activity. The Ni-A state is also called the unready state, since it required prolonged reduction to become activated, whereas Ni-B is a ready state, which can be activated rapidly by reducing agents. On reduction a third paramagnetic state, Ni-C, is formed, which is considered to be an intermediate in the reaction cycle. A likely assignment of these states is as follows: Ni-A and Ni-SU are µ-peroxo, Ni-B and Ni-SII are µ-hydroxo and Ni-C and Ni-RI are µ-hydrido species. The paramagnetic states, indicated with an asterisk, are Ni(III), whereas the EPR-silent states are Ni(II). The bridging position in the Ni-SII state can bind H2, CO or other gaseous ligands.

The oxygen-insensitive soluble hydrogenases, such as the NAD-reducing enzyme from R. eutropha, are more complex proteins. Although their sequences indicate the presence of typical NiFe centers, their NiFe center does not normally show nickel EPR signals in any state. This appears to reflect additional features to avoid inhibition by O2, as will be discussed later.

Some other proteins with hydrogenase-like NiFe active sites function in the sensing of hydrogen concentrations and regulation of hydrogenase expression. These ‘sensor hydrogenases’ have very limited hydrogenase activity with electron acceptors.

8.2.2 NiFeSe A subclass of nickel–iron hydrogenases, NiFeSe contains selenium in the form of a selenocysteine residue replacing one of the terminal cysteine ligands (Scheme 8.1b) (31-33). In the hydrogenase from Desulfomicrobium baculatum the [3Fe–4S] cluster found in NiFe hydrogenases is replaced by [4Fe–4S] and a magnesium ion in the large subunit is replaced by Fe(II) (33). These hydrogenases have high catalytic activity. They do not form the Ni-A (peroxo) state, although they do have an Ni–C oxidation state.

8.2.3 FeFe FeFe-hydrogenases contain iron, but not nickel. They tend to be used in fermentative metabolism in which H2 is produced (34, 35) and to be more sensitive to O2 than the NiFe-hydrogenases. The active site responsible for the activation of H2 in FeFe-hydrogenases, known as the H cluster, consists of a dinuclear iron cluster bonded to the protein backbone through a cysteine sulfur, which bridges to a [4Fe–4S] cluster (Scheme 8.1c). The two iron atoms have CO and CN ligands and are bridged by a unique dithiolate linker. This ligand has never been isolated, but appears from the crystal structure to be either 2-azapropane-1,3-dithiol or propane-1,3-dithiol (33, 36); the former is an attractive possibility because the nitrogen could act as a base to accept protons during the reaction. The protein, in addition to stabilizing the H cluster structure, maintains the center in an ‘entatic state’ which facilitates the production of H2. As with the NiFe-hydrogenases, the protein facilitates the transfer of electrons, H+ and H2 and the dimeric FeFe site is stable in several different oxidation states, some of which are paramagnetic (Scheme 8.3).

original image

Scheme 8.3. Scheme of reactivity and different redox states of the active site of FeFe-hydrogenases. The EPR-active states are marked with an asterisk. Formal redox potentials (at pH 8.0) measured by FTIR spectroelectrochemistry of D. desulfuricans FeFe-hydrogenase (170).

8.2.4 Fe (non-Fe–S) Hydrogenase (Hmd) An interesting, and so far unique, hydrogenase is found in methanogenic Archaea, which catalyzes a direct hydrogenation of the substrate using H2. It does not contain an intermediate electron-transfer pathway of iron–sulfur clusters. This highly evolved mechanism uses an active site that contains iron with carbonyl ligands, but is completely different from the hydrogenases discussed above. In the earlier literature it is described as the ‘metal-free’ hydrogenase, because the iron present in active preparations was very low. It was not appreciated at first that the activity was due to a small proportion of the active enzyme, which has a very high specific activity; moreover, the enzyme is extremely light sensitive. It is now known as Hmd (H2-forming methylenetetrahydromethanopterin dehydrogenase) or the iron–sulfur cluster-free hydrogenase. The crystal structures of the Hmd apoproteins from Methanocaldococcus jannaschii and Methanopyrus kandleri have been reported (37). Recently the structure of the active protein from M. jannaschi has been published (171), which confirms the expected features of the catalytic cycle. As for the catalytic center, Mössbauer spectra show that the iron is low-spin Fe(II) (38) and FTIR spectra indicate the presence of two carbonyl ligands per iron (39). Upon photolysis, the enzyme releases iron, CO and a unique guanine nucleotide containing a pyridone (40, 41). The N/O ligands illustrated in Scheme 8.1d were indicated by X-ray absorption spectroscopy (41) and may originate from the pyridone (42). A possible open site for H2 binding is modeled at the position trans to a CO ligand, considering that the ligand would be expected to increase the acidity of bound H2.

8.2.5 Biosynthesis of the Active Sites The assembly of the active sites of the hydrogenases requires complex sequences of reactions, facilitated by specific proteins known variously as chaperones, metallochaperones and scaffold proteins. The genes for these assembly proteins are often found in an operon together with the structural genes for the hydrogenase proteins (20, 43). In E. coli, the synthesis of the complete NiFe center involves a minimum of seven maturation enzymes plus carbamoyl phosphate, GTP and ATP (44). Carbamoyl phosphate is now known to be the precursor of the cyanide group, but the origin of the carbonyl is still unknown (45). Starting with the hydrogenase apo-protein, the iron atom of the dinuclear cluster is added first by an unknown mechanism. Specific enzymes are responsible for transport of nickel into the cell and its insertion into the partially formed hydrogenase metallocluster (46).

The biosynthetic pathway for the assembly of the H clusters in the FeFe-hydrogenases is not completely understood, but two novel radical S-adenosylmethionine-dependent proteins have been found to be required for the assembly of an active Fe-hydrogenase (47). By analogy with the biosynthesis of biotin or lipoamide, these enzymes may be responsible for inserting the sulfur in the azapropane-1,3-dithiol ligand.

8.3 Experimental Approaches

  1. Top of page
  2. Introduction: The Future Hydrogen Economy
  3. Chemistry of Hydrogenase Catalytic Sites
  4. Experimental Approaches
  5. Catalytic Mechanisms of Hydrogenases
  6. Progress So Far with Biological Hydrogen Production Systems
  7. Conclusion and Future Directions
  8. References

Experimentally, hydrogenase activity can be measured by artificial electron mediators such as viologen dyes or the surface of an electrode. Hydrogenases also catalyze reactions with H2 which involve no net electron transfer, such as the isotopic exchange of 1H2 with 2H2 or 2H2O, forming 2H2 and 1H2H (48).

8.3.1 EPR and Related Methods EPR spectroscopy provided the first direct evidence that hydrogenases could contain redox-active nickel. Early studies of hydrogenases from sulfate-reducing and methanogenic bacteria revealed three types of EPR spectra, which were identified as due to nickel by substitution with 61Ni (49). As mentioned above, two signals, designated Ni-A and Ni-B, were observed in the oxidized state and one in the reduced state designated Ni-C (50). The facile reduction from Ni(III) to Ni(II) (Scheme 8.2) showed that the redox chemistry of the center takes place at the nickel. The iron atom in the NiFe center appears to bear little unpaired electron density, as seen by the very small hyperfine couplings to 57Fe shown by ENDOR of enzyme grown in 57Fe medium (51). Mediator titrations showed that the redox potentials of the nickel are pH dependent, consistent with the involvement of protons in the reduction of the center (52-54).

The paramagnetic states of the NiFe-hydrogenases have proved to be a valuable tool for observing hydrogen bound to hydrogenase, which cannot be seen by X-ray diffraction or FTIR spectroscopy. Hyperfine interactions with protons have been studied by pulsed EPR, ENDOR and related techniques (23). Geometric information was obtained by exploiting the orientation selectivity of the EPR spectrum and, more precisely, by studies of single crystals. The results are consistent with the model for the Ni-C oxidation state, in which the center has Ni(III) with a bridging hydride to the iron at the position indicated by X in Scheme 8.1a (55).

Further work with DFT calculations has helped to define the likely intermediates in catalysis. A combination of techniques is narrowing the search for the position where H2 binds to hydrogenases. Orientation-selective ENDOR spectroscopy, combined with single-crystal EPR and density-functional calculations, has led to the conclusion that the Ni-C state has a bridging hydride ligand (23, 29, 56).

8.3.2 FTIR Spectroscopy The identity of the unusual diatomic ligands to the iron was established by FTIR spectroscopy of Allochromatium vinosum NiFe-hydrogenase. Unusual infrared absorption bands were shown to shift in response to changes in the oxidation state of the nickel (57). The effect of these ligands, which are very unusual in biology, is to maintain the low-spin state of the ferrous iron ion of the dinuclear center. FTIR can observe these ligands in all oxidation states of the dinuclear center, in contrast to EPR, which only observes paramagnetic states (Scheme 8.2). The redox titration of the Ni-A–Ni-SU couple, monitored by FTIR, has also been reported for the D. fructosovorans, A. vinosum and D. vulgaris NiFe-hydrogenases (58-60). The pH dependences of the formal measured redox potentials for these standard NiFe-hydrogenases are indicative of a one-electron/one-proton step, in agreement with the small shift in the frequencies of the CO and CN ligands, which indicates that there is compensation of the charge density at the active site.

8.3.3 Protein Film Voltammetry (PFV) In the hands of Armstrong's group, protein film voltammetry (PFV) has provided an incisive method not only to measure redox potentials of the metal centers of hydrogenases, but also to observe the catalytic activity of the enzymes in all their forms and to follow the deactivation/activation processes in real time (61, 62). The hydrogenase is adsorbed as a film on the surface of an edge-cut graphite rotating electrode, which allows facile electron exchange. The electrode is mounted in an anaerobic chamber into which various gases may be introduced. The rate of diffusion of H2 molecules to the enzyme is controlled by the rotation rate of the electrode. The electric current is a measure of catalytic activity and the driving voltage can be altered at will. Analysis of the non-turnover cyclic voltammograms of the CO-inhibited hydrogenase allows the determination of the redox potentials of the iron-sulfur clusters and the electrode coverage (63). Rapid changes in applied voltage can be used to measure the instantaneous activity of the enzyme, while chronoamperometry, in which a current is measured during the steady application of a voltage, can be used to follow activation and inactivation (63). This strategy allows a total control of the redox state of the enzyme by modulating the potential of the electrode (62). By using cyclic voltammetry and potential step techniques and by controlling experimental conditions as pH, H2 partial pressure and scan rate, the mechanism of the interconversion from the ready to the active states was defined as a two-step mechanism involving a chemical step and an electrochemical step (64). The exposure of the enzyme to O2 at different potentials and the further analysis of the reductive reactivation showed that the unready form was mainly formed, hence it should contain a product of the partial reduction of O2, probably a peroxo species (65). The reactivation of this species by reducing agents is so slow in vitro that it could pose a problem to the organism. Lamle et al. showed that CO and H2 acted not only to drive the irreversible activation process, but also to displace the peroxo or hydroxo bridging species and thus facilitate reactivation of the unready species (66).

Using PFV, the ability of several hydrogenases to tolerate exposure to O2 was tested, showing a different behavior. Whereas the D. desulfuricans FeFe-hydrogenase was irreversibly damaged by the O2, the R. eutropha membrane-bound NiFe-hydrogenase showed reversible inactivation, followed by fast reactivation. A. vinosum and D. gigas [NiFe] hydrogenases showed a different reactivity with O2 depending on the reaction conditions, but it was clear that the reversibility of A. vinosum was higher than that of the D. gigas enzyme (67). The PFV methodology has also been used by Leger and co-workers for studying the kinetics of wild-type (68) and mutant samples of D. fructosovorans NiFe hydrogenase (69). The electrochemical technique allowed studies of the effects of mutations near the distal iron-sulfur cluster of the hydrogenase on intra- and intermolecular electron transfer (69).

8.4 Catalytic Mechanisms of Hydrogenases

  1. Top of page
  2. Introduction: The Future Hydrogen Economy
  3. Chemistry of Hydrogenase Catalytic Sites
  4. Experimental Approaches
  5. Catalytic Mechanisms of Hydrogenases
  6. Progress So Far with Biological Hydrogen Production Systems
  7. Conclusion and Future Directions
  8. References

Examination of the structures of NiFe-hydrogenases suggests that the components of the enzyme reaction (Equation 8.1), namely H2, H+ and electrons, are brought to the dinuclear center by different paths through the protein (Figure 8.3). It is possible to divide the hydrogenase catalytic cycle into six steps, which are, in the direction of H2 oxidation:

  1. Diffusion of hydrogen molecules from the surface of the protein to the active site. Although there are several ways in which H2 can diffuse through the protein (70), a particular series of channels were revealed by structural determination of an NiFe-hydrogenase saturated with xenon, which becomes localized in the channels (71).

  2. Heterolytic splitting of hydrogen molecule after binding to the bimetallic active site. This may be summarized as

    mathml alt image(8.3)
    mathml alt image(8.4)
    where “< >” represents a double-or triple-bridged dinculear center.

    It is noticeable that the NiFe-hydrogenases, the FeFe-hydrogenases and Hmd each contain a low-spin Fe(II) with CO and, usually, CN ligands at the active site (Scheme 8.1). As the three types of hydrogenases are not phylogenetically related, the strict conservation indicates that these elements are the key to H2 conversion to hydride.

  3. Oxidation of the hydride to H+. This step requires an acceptor that can take a transient two-electron reduced state, then donate electrons one at a time to the electron-transfer chain. The reduction step for the NiFe-hydrogenases might be inline image or for the FeFe-hydrogenases inline image or even inline image (22, 72).

  4. H+ transfer from the active site to the water solvent. In the FeFe-hydrogenases, the azapropane-1,3-dithiol bridging ligand has been modeled with its NH proton close to where H2 is presumed to bind, so it could act as an H+ acceptor (17). In the NiFe-hydrogenases, cysteine ligand of the Ni is generally considered as an H+ acceptor (17, 21, 22). A conserved histidine residue is in a position to promote a hydrogen bond to one of the NiFe bridging cysteines, which DFT calculations have predicted would favor the protonation of the Nɛ of the histidine (73). Possible pathways for hydrons leading from the active site have been proposed, including a glutamate residue (74) and numerous internal water molecules in the protein structure, including ligands to a magnesium ion (75).

  5. Electron transfer from the active site to the distal redox cluster. Most hydrogenases comprise a chain of iron-sulfur clusters, separated by distances of 1.2-1.4 nm, from the catalytic center to a point on the surface where an electron-transfer protein could bind (Figure 8.3). The distances between the clusters seem to be critical for efficient electron transfer (76), although the redox potentials do not. In D. gigas hydrogenase, the clusters are, in sequence, a inline image, a inline image and a inline image cluster. The central cluster has a much less negative reduction potential than those of the others and the H+/H2 potential, but this does not appear to affect the rate of reaction significantly (77). An important factor is that the transfer of electrons is accompanied by hydrons, to preserve charge neutrality. In hydrogenases, this requires that there is movement of protons to compensate for transfer of electron density. A corollary of this is that if there is no possibility of charge compensation, electron transport is blocked. Possibly for this reason, the distal inline image cluster has an essential histidine ligand and substitution of this by any other ligand decreases the catalytic rate (69). The FeFe-hydrogenases from eukaryotic algae such as Scenedesmus obliquus are small monomeric proteins, lacking the chain of iron-sulfur clusters (78). This shows that the H cluster can be self-sufficient for hydrogenase activity. These hydrogenases are believed to exchange electrons directly with the iron-sulfur clusters of ferredoxins.

  6. Intermolecular electron transfer from the distal cluster to the redox partner. This part of the hydrogenase structure shows the greatest variability, due to the wide variety of electron carriers used. For D. gigas hydrogenase, the acceptor is a positively charged cytochrome c3. A ‘crown’ of glutamate residues around the distal inline image cluster (27, 79) attracts and orients cytochrome c3 through the positive charges of lysine residues around the hemes (80-82). Hydrogenases that reduce membrane-bound quinones such as menaquinone generally do so through an additional heme-containing hydrophobic membrane anchor subunit. As will be discussed later, the membrane-bound hydrogenases are particularly suited for binding to the hydrophobic surfaces of carbon electrodes.

8.5 Progress So Far with Biological Hydrogen Production Systems

  1. Top of page
  2. Introduction: The Future Hydrogen Economy
  3. Chemistry of Hydrogenase Catalytic Sites
  4. Experimental Approaches
  5. Catalytic Mechanisms of Hydrogenases
  6. Progress So Far with Biological Hydrogen Production Systems
  7. Conclusion and Future Directions
  8. References

8.5.1 Fermentation Anaerobic bacteria such as E. coli and Clostridium species produce H2 as a product of fermentation of organic materials. Hydrogen production serves to maintain the redox balance of the metabolic cycles in their cells. These fermentations may be applied to generate H2 as a fuel from many types of biomass, including the waste products of agriculture and food production, which have a high biological oxygen demand (83-85). Usually the fermentation is associated with the production of other compounds such as organic acids and solvents, which can also be useful. The nature of the feedstock means that it is not practicable to use pure cultures of bacteria in order to select for H2 production. The production of H2 has principally been optimized by the selection of growth conditions such as pH, temperature, nutrients and dilution rate and by bioreactor design (86). For example, the growth of methanogens, which would produce methane from H2 and CO2, is prevented by operating outside the pH range in which methane production is energetically favorable (83).

8.5.2 Oxygenic Photosynthesis In principle, the cleanest hydrogen biotechnology is to use photosynthesis to produce reducing equivalents for hydrogen production and release O2. This requires the isolation and application of metabolic systems from algae or cyanobacteria. (87-90).

Photosynthesis involves the creation of oxidizing and reducing species by photolysis of chlorophylls, coupled to complex arrangements of electron carriers (91). In oxygenic photosynthesis, Photosystem II generates O2 and Photosystem I generates compounds with negative reduction potentials that could potentially generate H2. Light is a destructive factor in most biological systems and in particular the strong oxidizing potentials generated in Photosystem II mean that the D1 protein has a short half-life. This problem, known as photoinhibition, is overcome by continuous rapid synthesis and reassembly of the protein complex (92). Another problem when nitrogenase or an FeFe-hydrogenase is the source of H2 is that these enzymes are highly sensitive to O2 and must be kept separate from oxygen-evolving systems. Therefore, the common practice in biotechnology of using cells as ‘bags of enzymes’ (see for example (93)) will not work for photosynthetic water splitting. It requires the participation of the full machinery for protein synthesis and therefore a fully-functioning cell. A second problem for the use of photosynthetic organisms such as algae is that they have evolved to replicate themselves as efficiently as possible, conserving their resources such as hydrogenase. Existing organisms are certainly not optimized as hydrogen factories for us. Much genetic engineering of the cellular regulatory systems would be required.

Benemann et al. (11) conducted an early demonstration of the feasibility of this approach, using a combination of oxygenic photosynthesis with hydrogenase to produce detectable amounts of hydrogen and oxygen. For various reasons the yields of hydrogen were small. The systems used (spinach chloroplasts and Clostridium kluyveri hydrogenase) were not stable; the O2 produced was inhibitory to the hydrogenase and would reoxidize the reduced ferredoxin, thereby short-circuiting the system. Since then, steps have been taken towards improving the efficiency and longevity of the system, such as the use of thermophilic algae and cyanobacteria, in which the proteins are more robust (94), and embedding the components on surfaces or in gel matrices (95). These photochemical systems produce both H2 and O2 in a mixture and, moreover, in dilute aqueous solution. The mixture of gases is undesirable from the point of view of safety. It cannot be used directly in conventional fuel cells, because the platinum electrodes would catalyze the conversion of the mixture back to H2O. In order to obtain H2 as a fuel, it would have to be separated, for example by suitable permeable membranes, with inevitable loss of efficiency.

A further step is to avoid the use of cells or organelles altogether and to generate H2 and O2 from genetically engineered molecular systems. A chimeric protein complex has recently been constructed, comprising the PsaE subunit of Photosystem I from the thermophilic cyanobacterium Thermosynechococcus elongatus and the membrane-bound NiFe-hydrogenase from R. eutropha (96). The resulting hydrogenase-PsaE fusion protein associated spontaneously with ferredoxin and PsaE-free Photosystem I and was capable of light-driven electron transfer to H+, forming H2.

Alternatively, the processes of photochemical O2 production and anaerobic H2 production can be made to operate alternately over time, for example by switching the growth conditions. A novel intervention by Melis et al. (97, 98) in the eukaryotic alga Chlamydomonas reinhardtii was to suppress O2 formation by growing the algae under sulfur-limited conditions. This is because sulfur-containing amino acids are required for the constant renewal of the D1 subunit of the Photosystem II subunits, to repair damage done by light (92). The cells then switch to fermentative metabolism and consume their reserves of starch, producing H2. Restoration of sulfate to the growth medium then allows O2 production to resume and starch reserves to be recovered (99). In line with these results, progress has been reported on photosynthetic algae, genetically modified to avoid O2 evolution under light, which allows them to reach anaerobic conditions compatible with hydrogen evolution (100).

The hydrogenases of C. reinhardtii are of the FeFe type (101) and are sensitive to O2. Moreover, H2 inhibits the endogenous hydrogenases of these organisms and suppresses the biosynthesis of hydrogenase. By genetic engineering it should be possible to substitute the FeFe hydrogenases by ones that are less sensitive to O2, such as those from R. eutropha, and to modify the regulatory mechanisms for hydrogenase biosynthesis.

Cyanobacteria (prokaryotes) have the capacity to produce both H2 and O2 under appropriate conditions. The conditions under which hydrogen production may be optimized have been the subject of much investigation (94). H2 production by nitrogenases is a major source of hydrogen in cyanobacterial cultures. Nitrogenase is not as efficient a catalyst as hydrogenase and, moreover, it is highly sensitive to O2. In filamentous cyanobacteria, nitrogenase is spatially separated from the oxygenic photosynthetic reactions in specialized cells known as heterocysts. In N2-fixing organisms, the H2 produced is often recovered by specific membrane-bound hydrogenases, which oxidize it to recover energy in the form of ATP.

8.5.3 Anaerobic Photosynthesis Higher yields of H2 from organic wastes can in principle be obtained by the use of anaerobic photosynthetic bacteria, which use organic or inorganic electron donors instead of H2O to boost the production of H2 (102, 103) (Figure 8.2). Unlike oxygenic photosynthesis, this type of photosynthesis can use both near-infrared and visible wavelengths and is less sensitive to the destructive effects of light on the photosystems. Photobioreactors, using anaerobic photosynthetic bacteria such as Rhodobacter species, can use organic wastes as reductant. H2 has been produced on a pilot scale. In practice, most of the H2 appears to derive from nitrogenase rather than hydrogenase (104-106).

8.5.4 Emulation: Hydrogenase Model Compounds The active site of NiFe-and FeFe-hydrogenases is essentially an organometallic compound, a dinuclear center with CO and/or CN ligands. In the case of the FeFe-hydrogenase, the bimetallic center is bound to the rest of the protein only through a single cysteine residue. It sits in a cavity in the protein, analogous to the Fe-Mo cofactor in nitrogenase (107). Nevertheless the protein environment is one of the most important features of the hydrogenases, which minimizes the peaks and troughs in free energy during the course of the reaction of the catalytic centers with H2 (17). The protein can maintain a vacant site at which H2 or hydride can bind, while protecting it from inhibitory gas molecules. The position of the CO and CN ligands in the centers also appears to be critical. In the proteins, the ligands are located in pockets in the protein, with hydrogen bond partners to the CN ligands. This would assist the correct assembly of the centers, but would probably also assist the activation of H2/hydride at the vacant site.

Inorganic chemists have attempted to synthesize compounds resembling structural and catalytic features of the hydrogenase active site and examined their activity in O2 production. Pickett and co-workers synthesized a close analogue of the iron-only hydrogenase H cluster, linking a diiron complex to a inline image cluster (Scheme 8.4a) (108). The subsite model was capable of catalyzing proton reduction at a potential of −1.13 V (versus Ag/AgCl). Liu and Darensbourg explored derivatives of the classic symmetrical FeIFeI organometallic compound (µ-pdt)[Fe(CO)3]2 (pdt = propanedithiolate) as models of the iron dimer of the H cluster in FeFe hydrogenases. In the X-ray structures of the FeFe-hydrogenases, the ligands to one iron atom of the H cluster are rotated so that a CO ligand bridges to other atom of the pair. In a recent study, using the unique properties of an N-heterocyclic carbene, they synthesized an asymmetric mixed-valent FeIIFeI compound with a rotated state, which reproduced accurately the EPR signals and the IR spectra of the Has isolated and Hox cluster of the native enzyme (Scheme 8.4b) (109). In the same direction, Rauchfuss and co-workers synthesized another mixed-valence complex with spectroscopic properties and the reactivity resembling those of the Hox cluster (110). In this case they used diphenylphosphinovinylidene as a ligand on one iron and a trimethylphosphine on the other (Scheme 8.4c). The rotation of the ligands to one iron atom causes an asymmetry between the two iron atoms, which would facilitate binding of H2 to one iron atom. DFT calculations indicate that H2 or CO would be expected to bind to the terminal iron of the H cluster (17).

original image

Scheme 8.4. (a) H cluster framework synthesized by Pickett and co-workers (108). (b) Mixed-valent diiron compound with a rotated state according to Lin and Darensbourg (109). (c) Unsaturated mixed-valence diiron model of the Hox state synthesized by Rauchfuss and co-workers (110). (d) Nickel complex with diphosphine ligands incorporating a pendant base (111). (e) Ni compound with two pendant bases on the phosphines, synthesized by DuBois and co-workers (113). Depending on the substituents on nitrogen and phosphorus, the compounds were very active for H2 oxidation or production. (f) Carboxylic acid derivative for attachment to an amine modified electrode synthesized by Darensbourg and co-workers (117).

In the FeFe-hydrogenases, the dithiolate linker is a three-atom bridge of which the middle atom might be a nitrogen, which, as already mentioned, would be a pendant base well positioned to facilitate proton transfer out of the enzyme. DuBois and co-workers explored the chemistry of first-row metals with diphosphine ligands incorporating a pendant base as a proton relay in H2 activation (111). A complex was synthesized (Scheme 8.4d) which was a very effective catalyst for H2 production, displaying high rates and long lifetimes. The turnover number found, 350 s−1 at 22 °C, is comparable to those reported for NiFe-hydrogenases (500-700 s−1 at 30 °C (112)). Addition of H2 was the limiting step, oxidation of H2 being fast once it was incorporated to the coordination sphere (113). The effects of coligands on the pKa of such complexes was studied, demonstrating that there is considerable electronic communication between the nitrogen atom and the metal center (114).

Another model of the NiFe-hydrogenase was published by Ogo et al. Combining two aqueous solutions of Ni(S2N2) and [(C6Me6)Ru(H2O)3]2+, the resultant water compound reacted in water with hydrogen to form a hydride and protons (Scheme 8.4e) (115).

In order to investigate the electrocatalytic properties of such complexes, different strategies have been applied to attach them to electrodes. Pickett and co-workers reported the incorporation of iron-only hydrogenase model complexes into functionalized polypyrrole electrodes, obtaining electrocatalytic currents for hydrogen production when a proton source was provided (116). Electrocatalysis by this compound was measured, but at too high an overpotential for use in a commercial device. Another approach to linking model complexes to electrodes has been proposed by Darensbourg and co-workers, who modified model complexes with carboxylic acids to form amide bonds with amine functionalized electrodes (Scheme 8.4f ) (117). The bond was stable and all the reaction steps were identified. However electrocatalytic proton reduction on the electrodes was not observed due to the irreversible reduction of the iron complex.

A point to bear in mind with these biomimetic systems is that it may be necessary to avoid poisoning of the catalyst by the use of a selective filter membrane to exclude molecules such as O2, H2S and CO, as is done in fuel cells (4, 118).

A new approach to building hydrogenase models and modulating the properties of the metal site has been initiated by Jones et al., who reported the incorporation of a diiron compound with an α-helical synthetic peptide (119). The design of such hydrogenase maquettes offers the prospect of developing robust and cheap water-soluble H2 production catalysts (119).

8.5.5 Hydrogenases on Electrodes Yagi's group were pioneers in electro-enzymatic hydrogen generation with hydrogenases isolated from Desulfovibrio vulgaris adhering to glassy carbon electrodes (120). As already discussed, hydrogenases bound to electrodes are outstanding tools for the study of the kinetics, mechanisms of reaction and inactivation and reactivation processes of the enzymes. Other practical applications of hydrogenase electrodes include H2 sensors (121, 122). Hydrogenase electrodes have also been used as a source of reducing power in the electro-enzymatic preparation of chiral compounds (123). Armstrong et al. (124) reviewed the factors that allow electron transfer between electrode surfaces and the redox centers in redox proteins or enzymes. For renewable H2 production (Figure 8.1), several strategies have been employed to obtain a catalytic response from hydrogenase-modified electrodes, to accelerate the electrochemical reduction of hydrons to hydrogen in electrolysis and its reverse reaction, the oxidation of hydrogen in fuel cells (6, 125, 126):

  1. Indirect electron transfer between the redox sites in the protein and the electrode may be mediated by soluble redox carriers with redox potentials close to the H+/H2 couple, such as viologens (54, 127). This approach has been useful in estimating the midpoint potentials of the redox carriers in the protein which can be observed by spectroscopy (128). Some electron-transfer proteins may also act as mediators. Bianco and co-workers showed that the tetraheme cytochrome c3, which has hemes exposed to its surface and which is the natural substrate of Desulfovibrio desulfuricans FeFe-hydrogenase, could act as electron mediator in the electrocatalytic hydrogen evolution at pyrolytic graphite electrodes (PGEs) (129-131).

  2. Glassy carbon electrodes surface-modified with redox mediators could act as the electron source to soluble D. vulgaris hydrogenase for catalytic H2 evolution (132).

  3. Entrapment of hydrogenases in polymeric semiconductive matrices on PGEs was developed by Varfolomeyev and Bachurin (133, 134). Hydrogenases have been also entrapped in viologen polymers, which acted as electron conductors, generated in the electrode (135, 136), or in an amphiphilic bilayer assembly covering an electrode that also contains a viologen compound (137). De Lacey et al. reported a hydrogenase electrode formed by successive layers of the enzyme and a viologen compound immobilized by avidin–biotin affinity interactions, in which the current density of H2 oxidation was proportional to the quantity of hydrogenase layers (138). Another method of co-immobilization of hydrogenase and the redox mediator was based on intercalation of hydrogenase between two layers of a mixture of clay and a viologen polymer deposited on a glassy carbon electrode (121). A similar clay-based layer-by-layer deposition method has been reported in which the natural redox partner of the hydrogenase, cytochrome c3, replaced viologen as redox mediator (139). Redox-mediated electro-enzymic oxidation of H2 has also been reported for bacterial cells with hydrogenase activity adsorbed on glassy carbon electrodes (140).

  4. A mediated response for the hydrogenase can be monitored by immobilizing hydrogenase on a carbon electrode and using methylviologen as a redox mediator. This procedure allowed a qualitative study of the activation/inactivation processes (141). One step further is building a multi-monolayer hydrogenase electrode based on the avidin-biotin interactions and measuring the mediated current. With this kind of electrode, a quantitative analysis of the hydrogenase mechanism was carried out and the measured kinetics were in good agreement with previously published data (142).

  5. Direct, non-mediated, electrocatalysis by hydrogenases covalently immobilized on glassy carbon electrodes: Schlereth et al. reported on the activity of R. eutropha Z-1 hydrogenase covalently immobilized on a glassy carbon electrode (143). This complex enzyme has the capacity to reduce NAD+, affording the possibility of other NAD-dependent reactions of interest to biotechnology. It was able to support the electro-enzymatic reduction of NAD+ without external promoters or electron mediators. Bergel and co-workers described direct electrocatalysis of NAD+ reduction with R. eutropha H16 hydrogenase adsorbed on platinum electrodes (144, 145). The product of the electro-enzymatic reduction of NAD+ was demonstrated to be biologically active (146). R. eutropha hydrogenase has also been entrapped during the electro-polymerization of pyrrole on a Pt electrode and the modified electrode catalyzed the reduction of NAD+ (147). A fluidized bed reactor for the continuous reduction of NADP+ to NADPH with hydrogen has been developed by Greiner et al. (148), using hydrogenase I from Pyrococcus furiosus directly adsorbed on conducting graphite beds with excellent stability.

  6. Promoters, non-redox substances which assist the binding of proteins to surfaces, have been used to bind hydrogenases to electrodes and facilitate their electron exchange. This methodology was introduced by Eddowes and Hill (149) and Veeger and co-workers (150). Early examples of the alternative strategy to promote the adsorption of hydrogenases consisted of coating poly-l-lysine on either a mercury electrode surface (151) or a PGE (152). In these modified electrodes, the adsorbed polycationic film bonded the negatively charged molecule of D. vulgaris hydrogenase and promoted a direct electron transfer between the electrode and the redox centers of the enzyme, which was not detected in the absence of poly-l-lysine. Under low-overpotential conditions it was shown that the PGE modified with poly-l-lysine and hydrogenase behaved as an H2 electrode, the electrons rapidly exchanged between the electrode and the active site of the hydrogenase (152). By comparison, this electrode behaved as the mercury electrode did, but avoided the denaturation of the hydrogenase observed on its adsorption on the mercury drop surface (153). A similar stabilizing effect on the adsorption of proteins at a PGE was observed by Armstrong et al. with polymyxin B, a cyclic decapeptide with several amine groups (154). Films of A. vinosum hydrogenase PGE with polymyxin B showed a higher turnover number for unmediated H2 oxidation than those measured with soluble organic compounds such as methylene blue or methylviologen as electron acceptors. In these experiments, rapid catalytic hydrogen oxidation, close to mass transfer-controlled rates, were observed (63). Armstrong and co-workers showed that NiFe-hydrogenase molecules from A. vinosum adsorbed on edge-cut PGE, modified with polymyxin B, catalyzed hydrogen oxidation at rates comparable to those with electrodeposited platinum and, more importantly, are much more resistant to CO poisoning than the noble metal (8).

  7. Direct electrochemistry, without aid of promoters, has also been reported for NiFe-hydrogenases adsorbed on a carbon electrode (155, 156). The efficiency of binding of hydrogenases to electrodes, with and without promoters, can be rationalized on the basis of the structural motifs that control electron transfer in vivo between the NiFe-hydrogenases such as that from D. gigas and cytochrome c3. An analysis of the electrostatic potential distribution over the surface of this protein showed an asymmetric charge distribution with the most external inline image cluster located in the negative region of the protein surface, that is, at the origin of a strong dipole moment in the enzyme molecule (157). Using this rationale, D. gigas hydrogenase was covalently coupled to PGE modified with a monolayer of 4-aminophenyl groups in an orientation that allowed direct electron transfer to the electrode. As the enzyme was covalently bound to the electrode surface, these electrodes showed exceptionally high operational stability (157). Hydrogenase electrodes have also been constructed with FeFe-hydrogenases. Hagen and co-workers reported a direct electron exchange of the hydrogenase from Megasphaera elsdenii with glassy carbon electrodes (158). In this case, polymyxin did not stabilize the response of the electrodes and even attenuated the electrode response. As expected, this FeFe-hydrogenase electrode was more sensitive to CO poisoning that the NiFe-hydrogenase electrodes and their electrochemical response decayed rapidly after CO addition. FeFe hydrogenase from Desulfovibrio vulgaris adsorbed on PGE also catalyzed direct H2 production in the absence of any promoter (159); however, the catalytic current was lower than those observed with A. vinosum NiFe-hydrogenase adsorbed on an electrode that was surface modified with polymyxin B sulfate (63).

  8. To gain stability for the development of fuel cells, covalent attachment of the enzyme is needed. De Lacey and co-workers attached D. gigas NiFe-hydrogenase through its glutamate residues to an amine-modified carbon electrode, which increased the electrode stability from hours to weeks (157).

8.5.5.1 Sensitivity and Resistance of Hydrogenases to O2, CO and Other Inhibitory Gases O2, NO, H2S and CO are strong inhibitors of the active sites of hydrogenases, just as they are poisons for Pt electrodes. Inactivation by O2 is a serious drawback for the utilization of NiFe-and FeFe-hydrogenases on electrodes (34), although in the case of NiFe-hydrogenases the inhibitory effect can be reversed by a reductive treatment (160). However organisms that produce these gases metabolically (Figure 8.2) have evolved mechanisms to counteract the inhibitory effects.

In some hydrogenases from H2-oxidizing bacteria, the active site appears to be modified so as to restrict access to O2 (161, 162). The O2-resistant soluble NiFe-hydrogenase from R. eutropha is a complex protein in which there is FTIR spectroscopic evidence for the involvement of an additional cyanide ligand to the nickel (163). The enzyme from a mutant organism, lacking the hypX gene that is required for insertion of the extra CN ligand, is sensitive to O2 (163). The R. eutropha enzyme gives no Ni-C EPR signal and there is recent evidence that it contains an additional FMN group (164). It has been suggested that this is located so as to accept hydride directly from the NiFe center, possibly avoiding an O2-sensitive paramagnetic state (165).

A second oxygen-tolerant NiFe-hydrogenase occurs in R. eutropha, which is membrane bound. This hydrogenase maintains a significant amount of H2 uptake activity in the presence of high levels of O2 and full activity was recovered upon removal of H2 (10).

Karyakin et al. compared some advantages of hydrogenases in fuel cell electrodes with those of platinum (166). In agreement with the findings of Armstrong and co-workers (8), they showed that hydrogenases adsorbed on PGE could oxidize hydrogen at rates comparable to those for electrodeposited platinum and with less susceptibility to CO poisoning (6, 8). Hydrogenases are more selective than platinum for H2, so the problem of poisoning electrodes by impurities from the fuel gas can be solved using enzymatic electrodes. The enzymes adsorbed on bare or polypyrrole-modified PGEs were tested under different H2–CO gas mixtures; no CO inhibition was observed up to a CO partial pressure of 0.1%, and even at 1% CO the activity was decreased by just 10%. By contrast, platinum loses 99% of its activity after 10 min under 0.1% CO. Moreover, the inhibition by CO was reversible and, after exposing the enzyme to pure CO, 100% activity was recovered after flushing the CO with hydrogen, whereas recovery of noble metals poisoned by CO requires special procedures (166). The results with Na2S were similar and showed insensitivity of the hydrogenase up to 5 mM sulfide.

An important step was the realization by Vincent et al. (126) that if hydrogenases were used that are unreactive with O2, these effectively contain their own intramolecular barrier, obviating the need for a membrane to separate the O2-consuming side of a fuel cell from the H2-consuming side. This made possible the design of an entirely enzymatic fuel cell, using oxidases which are unreactive with H2. The cell constructed used the membrane-bound hydrogenase of R. eutropha at the anode for hydrogen oxidation and the robust copper-containing oxidase laccase from Trametes versicolor at the cathode (Figure 8.4a, b). Better performance was shown by R. metallidurans hydrogenase, leading to a fuel cell which operated in a non-explosive mixture of air with 4% H2. The cell was able to generate a usable electric current from 3% H2 in still air. In the presence of CO it achieved almost 1 V at open-circuit conditions, which represents a noticeable advantage over platinum electrodes (167). Three cells were connected in series and provided enough power to power a wristwatch for 24 h.

original image

Figure 8.4. (a) Schematic representation of a hydrogenase/laccase bio-fuel cell. (b) On the anode hydrogenase oxidizes hydrogen to protons. The electrons are used by the laccase on the cathode to reduce oxygen to water. When using a hydrogenase which is not affected by H2 as R. metallidurans, the fuel cell operates without the need for a proton transfer membrane for protecting the hydrogenase from oxygen inactivation (126).

All these devices have been based on enzyme adsorption on the electrodes and surface electrochemistry is a two-dimensional process. Therefore, in order to increase the catalytic rates and produce smaller fuel cells, an increased surface area is required. Multi-walled carbon nanotubes were grown on microscale electrodes. The nanotubes were modified with diazonium salts to cover them with amine functionalities and covalently attach the hydrogenase (Figure 8.5). The electrodes prepared in this way had an electroactive area of 130 C cm−2, 43 times higher than the area obtained upon modification by the same method of a polished PGE. The catalytic currents measured were 33 times larger than those at the PGE, meaning that most of the electroactive area was covered by the hydrogenase. This result was similar to the best results reported by other workers for hydrogenase directly adsorbed on carbon electrodes, which almost equaled the performance of Pt-based electrodes. The operational stability of those electrodes was tested under hydrogen electrocatalytic oxidation conditions. The covalent bonding of the enzyme to the electrode was crucial to gain an operational stability of more than 1 month, retaining 90% of initial activity (168).

original image

Figure 8.5. Carbon nanotube electrode for D. gigas hydrogenase. Carbon nanotubes were grown on a gold microchip. The nanotubes were electrochemically modified with amine functionalities for the orientation and covalent attachment of the hydrogenase (168).

8.6 Conclusion and Future Directions

  1. Top of page
  2. Introduction: The Future Hydrogen Economy
  3. Chemistry of Hydrogenase Catalytic Sites
  4. Experimental Approaches
  5. Catalytic Mechanisms of Hydrogenases
  6. Progress So Far with Biological Hydrogen Production Systems
  7. Conclusion and Future Directions
  8. References

We should not underestimate the challenges involved in trying to exploit biological hydrogen production for a major source of energy. Studies are at the stage of proof of principle and clearly a huge effort of research and development would be needed to develop practical renewable systems to replace fossil fuels.

8.6.1.1.1 Domestication of Hydrogenases The ideal organism for biological hydrogen production might be one that produces oxygen by photosynthesis and diverts a high proportion of the resulting reducing equivalents to generate hydrogen. Such organisms exist, but the yields of hydrogen are low. In view of the complexity of biosynthesis and regulation of hydrogenases, this would be a major project in cell biology. An analogy may be drawn with the domestication of animals and plants to support an agricultural existence, which took millennia of selective breeding. This process could be greatly accelerated by genetic engineering.

8.6.1.1.2 Emulation of Hydrogenases The features of the hydrogenase molecule that allow it to be so efficient and specific have become clearer as a result of recent research. The elucidation of the structures and mechanisms of hydrogenase action has revealed novel organometallic chemistry and novel applications of theoretical chemistry. A major challenge is to create a stable structure with a vacant site with the appropriate geometry for hydrogen binding and activation and controlled access for electrons, hydrons and H2.

8.6.1.1.3 Hydrogenase Electrodes The technology has been demonstrated to work on a small scale. Fuel cells could operate without the membranes, which are a weakness in current designs. Other attractive features are the lack of reliance on precious metals, room temperature operation, the use of non-explosive H2–air mixtures and resistance to pollutant gases. The challenges include creating greater stability and sufficiently high current densities.

Abbreviations
DFT

density functional theory

ENDOR

electron–nuclear double resonance

EPR

electron paramagnetic resonance

FMN

flavin mononucleotide

FTIR

Fourier transform infrared

Hmd

H2-forming methylene-H4MPT dehydrogenase

MPT

methanopterin

NAD

nicotinamide adenine dinucleotide

NADP

nicotinamide adenine dinucleotide phosphate

PFV

protein film voltammetry

PGE

pyrolytic graphite edge electrode

References

  1. Top of page
  2. Introduction: The Future Hydrogen Economy
  3. Chemistry of Hydrogenase Catalytic Sites
  4. Experimental Approaches
  5. Catalytic Mechanisms of Hydrogenases
  6. Progress So Far with Biological Hydrogen Production Systems
  7. Conclusion and Future Directions
  8. References