Bacterial hydrogenases have been harnessed to the removal of heavy metals from solution by reduction to less soluble metal species. For Pd(II), its bioreduction results in the deposition of cell-bound Pd(0)-nanoparticles that are ferromagnetic and have a high catalytic activity. Hydrogenases can also be used synthetically in the production of hydrogen from sugary wastes through breakdown of formate produced by fermentation. The Bio-H2 produced can be used to power an electrical device using a fuel cell to provide clean electricity. Production of hydrogen from confectionery wastes by one organism (Escherichia coli) can be used as the electron donor for the production of Bio-Pd0 from soluble Pd(II) by a second organism. The resulting Bio-Pd0 can then be used as a bioinorganic catalyst in the remediation of Cr(VI)-contaminated solutions or polychlorinated biphenyls at the expense of Bio-H2, as a hydrogenation catalyst for industry or as a component of a fuel cell electrode.
- Cr(VI) reduction
- fuel cell
- hydrogenation catalyst
- palladium catalyst
- polychlorinated biphenyl
Hydrogenases and bioreduction of metals
Many microorganisms are capable of reducing toxic metals to less harmful forms by a variety of mechanisms including hydrogenase activity [1,2], e.g. in the reduction of U(VI)  and Se[IV] . The ability to reduce the problematic radionuclide 99Tc(VII) to insoluble 99Tc(IV) was attributed to the activity of the hydrogenase-3 component of the FHL (formate hydrogenlyase) complex in Escherichia coli  and to the nickel–iron hydrogenase of Desulfovibrio fructosovorans . Decontamination of Tc(VII) was also achieved by other Desulfovibrio spp. and demonstrated in flow-through decontamination processes [7,8]. Metal reductase activity was also extended to the reduction and recovery of precious metals from wastes at the expense of H2 [9–11], and conclusively identified as hydrogenase-mediated by the use of mutants of D. fructosovorans deficient in one or more of the three major hydrogenases . Pd(II) was reduced to Pd(0) which was held as biotemplated and supported metallic nanocrystals on the cell surface (Figure 1).
Catalytic activity and properties of cell-bound Pd(0)
The microbially reduced Pd(0) (Bio-Pd0) had a high catalytic activity, comparable with a commercial supported metal catalyst, in a standard hydrogenation reaction  and in its ability to reduce toxic Cr(VI) to less toxic Cr(III), at the expense of H2, under conditions where bacteria alone or chemically reduced Pd(0) were ineffective [14–17]. This activity was seen also in Bio-Pd0 biorecovered from industrial wastes and harnessed at acidic pH [16,18]. Bio-Pd0 was also able to reductively dehalogenate extensively polychlorinated biphenyls, which are recalcitrant to microbiological or chemical attack [19,20]. The reason for the high efficacy of the biomaterial is still under investigation but the Bio-Pd0 has magnetic properties diagnostic of nanocrystals (Figure 2). If the magnetic moment is examined as a function of the applied magnetic field a hysteresis is visible (Figure 2), which is indicative of a ferromagnetic material. From the magnetic result, the size of the population subset of nanoparticles can be calculated; in this case, it is approx. 5 nm . It is well known that nanoparticles have properties that differ from bulk crystals but chemical synthesis is difficult and expensive. First, in the preparation of nanoparticles, nucleation must be achieved, whereas growth is controlled to prevent the formation of larger particles. Secondly, the nanoparticles must be physically separated either by ‘capping’ (coating with a protective layer) or attachment to a support [21–23] during or after preparation. This prevents uncontrolled particle growth, which occurs either by coalescence, the fusing of two or more smaller particles to form a larger one or by ripening, i.e. the growth of large particles by migration of material from smaller ones. Either of the two solutions above is required to preserve the nanoparticles and maintain their particular physical properties for any length of time. The bacterial system is unique in that it provides nucleation functions, support on the biomatrix and, importantly, the rates and extents of nanocluster growth are controllable. Thus, according to requirements, Bio-Pd0 can be prepared with variable nanoparticle size, as illustrated by the magnetic (Figure 2) and also the catalytic  properties. The activity of Bio-Pd0 was also extended to its use as an electrode material for a low-temperature PEM (proton-exchange membrane) fuel cell (L.E. Macaskie, C. Levine and K.K. Kendall, unpublished work), which makes electricity from hydrogen.
Fuel cells and sustainable energy 
Low-temperature PEM fuel cells use hydrogen as an energy vector. The incoming H2 is split at the anode that comprises precious metal nanoparticles supported on electrically conducting graphite. Electrons derived from the H2 generate current to power an electrical load. The protons from the H2 move across an intermediate proton conducting layer and are reunited with the electrons from the circuit at the cathode and, with oxygen from the air, form H2O as the only product. Consequently, a fuel cell is environmentally clean, but the major disadvantage of the PEM fuel cell is that it requires a clean supply of hydrogen . Provision of high quality, cheap H2 is one of the objectives of the developing hydrogen economy. H2 can be made by (e.g.) electrolysis of water or by anaerobic digestion of organic wastes [25,26]. In the latter case it is difficult to exclude other gaseous products (e.g. H2S) that would poison the fuel cell catalyst or methane (the end-product of anaerobic digestion), which is not used in a PEM fuel cell; a mixed gas stream would require a prefiltration step.
Bioproduction of fuel cell quality H2 by fermentation of industrial wastes
Enterobacterial strains, in common with other bacteria, synthesize hydrogen through the hydrogenase-3 component of the FHL complex. The FHL complex comprises formate dehydrogenase and hydrogenase-3 and its function is to oxidize formate (an end-product of the mixed acid fermentation) to equimolar amounts of CO2+H2, coupling formate oxidation by formate dehydrogenase H to proton reduction . The conversion of formate is a homoeostatic response to acidic extracellular pH and a mechanism to dispose of this acidic end-product of fermentation . Expression of the FHL system is regulated by the constitutive repressor HycA, encoded by the hycA gene . An E. coli strain HD701, which cannot synthesize the FHL repressor and is, therefore, up-regulated with respect to FHL expression, was constructed . The up-regulated FHL expression allowed strain HD701 to reduce Tc(VII) at a higher rate (see above) when compared with its parental wild-type E. coli MC4100 . The hydrogenase-3 activity was also used synthetically in the generation of H2 by the fermentation of industrial confectionery wastes, using the up-regulated mutant . The head gas was passed through a NaOH trap to remove the CO2 component; H2 was the only component detected in the resulting gas stream and connection of the culture exit gas to a PEM fuel cell connected to a fan permitted continuous operation of the fan for the duration of H2 evolution by the batch culture [15 h; Figure 3 (see http://bst.portlandpress.com/bst/033/bst0330076add.htm)]. Subsequent tests showed that is was not necessary to remove the CO2 component of the culture exit gas, which could be passed directly into the fuel cell. The use of a monoculture ensures a constant gas composition and the use of an overproducing strain gives the maximum rate of H2 production, which was several-fold greater than that of the parent culture . However, there are three potential drawbacks in the use of the monoculture. The first is overgrowth by contaminant organisms, since it would be very difficult to run the process aseptically at the industrial scale. The E. coli strain was used as resting cells in buffer, it lacked an additional nitrogen source and hence negligible biomass growth was supported using sugary waste substrates. This type of waste is often derived from cane sugar (sucrose) but E. coli does not possess invertase activity and a large proportion of the waste is not used, reducing bioprocess efficiency. The genes encoding sucrose uptake and invertase activity were cloned into the H2-overproducing strain, conferring the property of H2 production from sucrose . The industrial acceptability of a genetically modified organism remains to be proved. The third problem is that enterobacterial strains perform mixed acid fermentation, i.e. H2 is produced along with other end-products. For maximum H2 yield per mole of substrate an additional downstream fermentation step is required to convert residual organic acids; this could be an anaerobic digester (for a secondary H2/CH4 stream) or a photobioreactor for the production of an additional H2 stream, e.g. as a side reaction of nitrogenase activity by Rhodobacter sphaeroides in the presence of light at the expense of outflow from the E. coli reactor. This was shown to be feasible (D.W. Penfold, M. Redwood and L.E. Macaskie, unpublished work).
Use of Bio-H2 in sustainable waste treatment processes: ‘closing the loop’
The use of hydrogenases in biotechnology shows one important aspect: they can be harnessed in the synthetic direction (for Bio-H2 production from wastes) or catabolically (at the expense of H2) for the hydrogenase-mediated reduction of metals by bacteria [e.g. Pd(II)→Pd(0); i.e. Bio-Pd0 synthesis] or Bio-H2 for the Bio-Pd0-catalysed reduction of Cr(VI), the reductive dehalogenation of PCBs (polychlorinated biphenyls) or indeed hydrogenation reactions for chemical manufacturing. Palladium metal splits H2, not only in fuel cells (above) but also within the Pd-crystal itself, to provide an excellent chemical catalyst. Therefore it should be possible to couple the Bio-H2 production by E. coli strain HD701 to processes which consume H2. A two-stage bioreactor was constructed with the first stage fermenter as shown connected to a fuel cell for electricity production (Figure 3) or with the gas outlet connected to a second stage reactor containing (i) Cr(VI)-reducing bacteria and Cr(VI); (ii) a reactor containing Desulfovibrio desulfuricans and Pd(II) or (iii) a reactor containing Bio-Pd0 and Cr(VI). In all cases the bio-H2 supported the reaction similarly to H2 from a commercial supply.
In conclusion, by coupling synthetic and catabolic hydrogenase activities it has been shown that it is possible to co-treat wastes from the food industry, using the product (Bio-H2) to treat precious metal waste solutions, producing a new class of bioinorganic catalyst (Bio-Pd0). This can be used in the treatment of a third [Cr(VI)] waste (produced by many chemical industries) or fourth (PCB) waste, providing the possibility to co-treat several classes of wastes from several industrial sectors simultaneously. In the field of sustainable energy, bacterial hydrogenase activity can produce a key component of a fuel cell and the means (Bio-H2) to generate electricity from it, whereas the use of Bio-Pd0 as a hydrogenation catalyst in industry could permit the use of H2 as a chemical reactant at lower temperatures and pressures compared with those currently used and, hence, a possible industrial application for Bio-H2 here also. In this context, it should be noted that for decoupling the rate of H2 production from its consumption, temporally or spatially, a suitable H-store is necessary and hydrogen biotechnology and H-store development must proceed together to realize the hydrogen economy. H-stores are now becoming available that will accept H2 at the low pressure of the culture head-gas (0.5–1.5 bar) and at ambient temperatures (D.W. Penfold, L.E. Macaskie and I.R. Harris, unpublished work) and these are currently being used in the coupling of the H-production and consumption steps, and in pooling the H-streams from dual synthetic bioreactor systems.
This work was supported by the EPSRC, BBSRC and the Royal Society (Industrial Fellowship to L.E.M.) in collaboration with C-Tech Innovation.
International Hydrogenases Conference 2004: Independent Meeting held at the University of Reading, 24–29 August 2004. Edited by R. Cammack (King's College London, U.K.) and F. Sargent (University of East Anglia, Norwich, U.K.). Organized by R. Cammack and R. Robson (University of Reading, U.K.). Sponsored by COST (European Cooperation in the field of Scientific and Technical Research), the European Science Foundation and the European Office of Aerospace Research and Development.
Abbreviations: FHL, formate hydrogenlyase; PEM, proton-exchange membrane
- © 2005 The Biochemical Society