Biotechnol Lett (2011) 33:8995 DOI 10.

1007/s10529-010-0393-7

ORIGINAL RESEARCH PAPER

Dehalogenation of environmental pollutants in microbial electrolysis cells with biogenic palladium nanoparticles
Tom Hennebel Jessica Benner Peter Clauwaert Lynn Vanhaecke Peter Aelterman Ruben Callebaut Nico Boon Willy Verstraete

Received: 31 May 2010 / Accepted: 24 August 2010 / Published online: 24 September 2010 Springer Science+Business Media B.V. 2010

Abstract Purpose of work Hydrodehalogenation of persistent pollutants, such as the groundwater contaminants trichloroethylene and diatrizoate, are catalyzed by biogenic Pd nanoparticles. As H2 gas supply for the dehalogenation reactions is still the limiting factor, this study examines in situ H2 production in the cathode of a microbial electrolysis cell. In a biogenic Pd nanoparticle (bio-Pd) free microbial electrolysis cell (MEC), dechlorination of trichloroethylene (TCE) with concomitant chloride and ethane formation was achieved in the cathode compartment at a removal rate of 120 g TCE m-3 total cathode compartment (TCC) day-1, applying -0.8 V

Tom Hennebel, Jessica Benner contributed equally to this article.

with a power source. When the cathode granules were coated with 5 mg bio-Pd g-1 graphite, chloride and ethane formation increased to 151 g TCE m-3 TCC day-1 corresponding with a specic removal rate of 48 mg TCE g-1 Pd day-1. In both cases, formation of unwanted byproducts, such as vinyl chloride, was not signicant. When the same setup was applied for transformation of the iodinated contrast medium diatrizoate (diaI3), reduction in a catalyst-free cathode of a MEC resulted in a removal of 48 9% during the rst h corresponding to 3 g diaI3 m-3 TCC day-1. Coating the cathodic graphite granules with bio-Pd enhanced the transformation resulting in a 93 4% removal during the rst h corresponding to 6 g diaI3 m-3 TCC day-1. These results suggest that MECs can produce H2 in a sustainable way to provide an economical interesting reactant for bio-Pd catalyzed dehalogenation reactions. Keywords Bioelectrical system Dechlorination Catalysis Groundwater remediation Nanotechnology

Electronic supplementary material The online version of this article (doi:10.1007/s10529-010-0393-7) contains supplementary material, which is available to authorized users.
T. Hennebel J. Benner P. Clauwaert P. Aelterman R. Callebaut N. Boon W. Verstraete (&) Laboratory of Microbial Ecology and Technology (LabMET), Ghent University, Coupure Links 653, 9000 Ghent, Belgium e-mail: willy.verstraete@Ugent.be L. Vanhaecke Laboratory of Chemical Analysis, Ghent University, Salisburylaan 133, 9820 Merelbeke, Belgium

Introduction Nanopalladium catalysts can be synthesized by the precipitation of Pd on the surface of bacteria leading to the production of biogenic Pd nanoparticles (bio-Pd) (Hennebel et al. 2009a). For example, Shewanella oneidensis reduced Pd(II) and subsequently precipitated

123

90

Biotechnol Lett (2011) 33:8995

it as Pd(0) nanocrystals on their cell walls and in their periplasmatic spaces when a H2 donor is provided (De Windt et al. 2006). No expensive and harmful chemicals are required as the bacteria serve as reductant for the Pd salt and as stabilizer of the Pd nanoparticles. Bio-Pd catalyzes the dehalogenation reactions of polychlorinated biphenyls (Baxter-Plant et al. 2003; De Windt et al. 2005), polybrominated diphenylethers (Harrad et al. 2007; Deplanche et al. 2009), trichloroethylene (TCE) (Hennebel et al. 2009b, c), chlorophenols (Baxter-Plant et al. 2003), lindane and chlorobenzenes (Mertens et al. 2007) and diatrizoate (diaI3) (Hennebel et al. 2010a). In the latter conversions, bio-Pd was used as catalyst and an external hydrogen donor, such as H2 or formate, was added as a reactant. In large-scale applications, the use of H2 can give rise to large costs and technical difculties. Microbial electrolysis cells (MECs) can be used for the production of H2. However this innovative technology has never been considered to deliver H2 for bioPd catalyzed dehalogenation reactions. In MECs, organic material is oxidized by electrochemically active microorganisms at the anode. Subsequently, the microorganisms transfer the electrons resulting from this oxidation to the anode by extracellular electron transfer. Through an electrical circuit, the electrons are transported to the cathode, where they are consumed for H2 formation (Mu et al. 2009a). In contrast with microbial fuel cells (MFCs), in which electrical energy can be extracted from the electrical circuit, one needs to supply electrical energy to the electrical circuit of an MEC by means of a power source. MECs have been used for the chemical reduction of nitrobenzenes (Mu et al. 2009b), for the decolorization of azo dyes (Mu et al. 2009a) and for the biologically catalyzed reduction of nitrate and TCE in a biocathode (Aulenta et al. 2008; Clauwaert et al. 2009). In this study, the application of bio-Pd in the cathode of an MEC as a catalyst for pollutant reduction was compared with an MEC without bioPd and tested in the dehalogenation reaction of two model compounds: (i) the groundwater pollutant TCE and (ii) the iodinated contrast medium, diatrizoate (diaI3). Both contaminants were selected based on their recalcitrance, their environmental relevancy and their degradation using bio-Pd in earlier studies (Hennebel et al. 2009b, 2010a). The rst objective of this study was to examine the dehalogenation of TCE and diaI3 in a MEC without a

catalyst at the cathode and at several cathode potentials. Additionally, it was tested if different amounts of bio-Pd at the cathode could function as catalyst to enhance the dehalogenation reaction.

Materials and methods Microbial electrolysis cell construction and operation The MEC was made of two Plexiglas frames (12.5 9 8 9 1.5 cm3 per frame; 0.15 l total cathodic compartment (TCC) of which 0.07 l net liquid cathodic compartment (NCC). The anodic and cathodic frames were lled with 160 g graphite granules (type 00514, diam. 1.55 mm, Le Carbone, Belgium) and were connected to the external electrical circuitry with a graphite rod current collector (5 mm diam. Morgan, Belgium). A cation exchange membrane (CEM) (Ultrex CMI7000, Membranes International Inc.) was used between the anodic and cathodic frames. As electrolyte for anode and cathode a minimal medium, consisting of 6 g Na2HPO42H2O l-1, 3 g KH2PO4 l-1, 0.1 g (NH4)2PO4 l-1, 0.1 g Ca(PO4)2 l-1 and 0.5 ml l-1 of a tracemetal solution previously described by Clauwaert et al. (2009) was used. The electrolyte (400 ml) was sparged with N2 and pumped through cathode and anode in a separate recirculation mode at 24 ml min-1. The anode medium was inoculated with a mixture of anaerobic sludge and anode medium of an active microbial fuel cell and 1 g sodium acetate l-1 was added as feed. After the successful start-up of the MEC, the cathode medium was spiked with 100 mg TCE l-1 or 0.1 mg diaI3 l-1 for the dehalogenation experiments. The headspace and the electrolyte of the cathode were sampled via a sample port without opening the system up to 24 h. For operation in MEC-mode, an external power source (PL-3003D, Protek, Englewood, NJ, USA) was used to apply a potential difference between zero (MFC mode) and -0.8 V between anode and cathode. The voltage difference between the anode and the cathode was measured to verify the applied cell voltage. The voltage between the cathode and a Ag/ AgCl reference electrode was monitored to measure the cathode potential. The voltage drop over a 1 X resistor in circuitry was used to measure the current owing from anode to cathode. These voltage

123

Biotechnol Lett (2011) 33:8995

91

differences were measured with a 34970A Data Acquisition Switch unit (Agilent, Diegem, Belgium). Production of bio-Pd-coated graphite Bio-Pd was produced as previously described by De Windt et al. (2005). Different amounts of a 500 mg bioPd l-1 suspension were mixed with the graphite granules and dried for 48 h at 100C to result in 1 mg Pd g-1 graphite and 5 mg Pd g-1 graphite, respectively. The cathode compartment (0.15 l) was supplied with 160 g bio-Pd coated graphite granules. For the anode compartment, only non-coated granules were used. Analytical methods Analysis of TCE and lower chlorinated reaction products such as vinyl chloride and dichloroethylenes as well as Cl-, was by performed according to Hennebel et al. (2009b). DiaI3 and its metabolite diaH3 were analysed on an LCMS system as in detailed described in the supplementary material. Absorption of TCE and activity tests of bio-Pd To test absorption of TCE onto the cathode granules, batch tests with 160 g graphite granules coated with 1 mg bio-Pd g-1 C in 400 ml electrolyte were spiked with 100 mg TCE l-1. GC measurements directly after spiking and after 24 h were taken. In addition, possible degradation of TCE was monitored by measurement of the Cl- concentration. To determine the catalytic activity of the bio-Pd coated granules, H2 from an external source was added to the
Table 1 Comparison of chloride formation (mg l-1, mean standard deviation) as a function of time in bio-Pd free microbial electrolysis cell setups with various applied potentials Time (h)

headspace (101 kPa) of the same setup. TCE concentrations in the headspace as well as Cl- formation in solution were monitored.

Results TCE dechlorination in bio-Pd free MECs at varying applied voltages In a rst experiment, the dehalogenation of 100 mg TCE l-1 in the cathode of an MEC was studied in open circuit mode. After 24 h, 5 0.3 mg Cl- l-1 was formed which was considered as background value, since no TCE removal was expected. Subsequently, the dehalogenation was studied in the closed circuit mode without bio-Pd on the electrode material and with various applied voltages, an increasing concentration of Cl- over time could be observed in all cases (Table 1). The concentrations after 24 h with 0, -0.1, and -0.4 V did not differ signicantly. However, increasing the supplied voltage to -0.6 and -0.8 V resulted in much higher removal efciencies. At the end of the experiment, 25 2 and 50 3 mg Cl- l-1 were produced respectively, corresponding to TCE removal efciencies of 17 2 and 62 6% (estimated on the basis of 3 mol Cl- from 1 mol TCE). Hence, a strong relation between the applied voltage and the removal of TCE could be observed. Ethane and Cl- were the main degradation products, chlorinated byproducts, such as vinylchloride and 1,1-DCE, could only be detected in trace amounts (\0.1 mg l-1). When neglecting formation of these products and calculating the TCE removal based on the Cl- formation, the

Chloride formation [mg l-1] at various applied potentials OC 0V 0.4 0.2 31 41 41 41 52 42 51 71 -0.1 V 0.39 0.01 66 65 66 76 76 76 85 89 -0.4 V 22 64 85 n.a. 74 74 84 83 11 3 -0.6 V 3.4 0.4 62 n.a. 91 10 1 11.7 0.9 12.4 0.5 13.9 0.7 28 2 -0.8 V 64 10 4 15 9 20 8 23 9 30 7 30 10 32 10 56 7

0 1 2 3 4 5 6 7

0.28 0.08 0.4 0.1 32 41 3.1 0.1 43 42 42 5.0 0.4

Each experiment was performed in triplicate

24

123

92

Biotechnol Lett (2011) 33:8995

mass balance reached only up to 62%. Most likely this is due to transport of TCE to the anode, volatilization and adsorption on the graphite granules or the tubing and Perspex material used in the setup. As the Clvalues are neither affected by adsorption nor volatilization, efciencies of the different tested setups were mainly based on the Cl- formation. TCE dechlorination in MECs with bio-Pd as catalyst in the cathode In a second stage of the research, bio-Pd was coated on the graphite of the cathode to enhance TCE dechlorination. Two different Pd concentrations were

applied, i.e., 1 and 5 mg Pd g-1 graphite, and the Cland ethane formation were monitored over 24 h (Fig. 1). When applying a cell voltage of -0.8 V, a clear difference between the varying concentrations could be observed. One mg Pd g-1 graphite in the cathode did not improve the formation of Cl- respectively ethane in comparison to non bio-Pd containing graphite. The produced Cl- after 24 h amounted to 32 6 mg Cl- l-1, which was lower than the formation of 50 3 mg Cl- l-1 in the case no bioPd was used. The same trend was observed for ethane production. After 24 h, 12 6% of the added TCE could be recovered as ethane with 1 mg Pd g-1 graphite versus 17 7% in the case of non-coated (catalyst-free) granules. In contrast, coating the granules with 5 mg Pd g-1 graphite resulted in faster and higher formation of both Cl- and ethane. After 2 h, already 80 30 mg Cl- l-1 was formed versus 11 5 mg l-1 in the case of not-coated granules. At the end of the experiment 100 30 mg Cl- l-1 was detected in the cathode (Fig. 1a). Ethane was produced more rapidly and to a larger extent as well. After 4 and 24 h respectively, 18 3 and 26 4% of the initially added TCE amount could be recovered as ethane (Fig. 1b). Because the decreased activity in the case of the 1 mg bio-Pd g-1 graphite was observed, batch experiments were conducted to investigate whether the granules coated with bio-Pd as such had a catalytic activity for TCE dechlorination. Measurements of the TCE concentration by GC-FID showed no removal under N2 atmosphere. Also no Clformation was observed. However, when external H2 was supplied, a TCE decrease of 63% was observed after 24 h. The Cl- formation of 86 mg Cl- l-1 corresponding to app. 30% TCE (estimated on the basis of 3 mol Cl- from 1 mol TCE) conrmed the catalytic transformation of TCE. Dehalogenation of diaI3 in MECs with and without bio-Pd at the cathode

Fig. 1 Inuence of the presence of bio-Pd [(lled inverted triangle) for 0 mg, (open circle) for 1 mg, (lled circle) for 5 mg bio-Pd g-1 graphite] coated onto cathode graphite granules on a chloride (in mg/l) and b ethane (in mol.% of added trichloroethylene (TCE, 100 mg l-1)) formation versus time; microbial electrolysis cell setups were all applied with -0.8 V external potential. Each experiment was performed in triplicate and mean standard deviation are shown

Operating the MEC setup in open circuit did not result in any diaI3 degradation after 24 h. When an external potential of -0.8 V was applied, 48 9% of the initially added diaI3 was transformed after 1 h, even without bio-Pd being present. Coating the graphite granules of the cathode with 5 mg bio-

123

Biotechnol Lett (2011) 33:8995

93

Fig. 2 Relative removal of 0.1 mg diatrizoate (diaI3) l-1 in function of time at an applied potential of -0.8 V comparing an MEC with a cathode containing non coated graphite granules (lled circle) versus an MEC with a cathode containing 5 mg bio-Pd g-1 graphite (lled inverted triangle). Open circuit conditions (open circle) were used as reference. Every experiment was performed in triplicate and mean standard deviation are shown

Pd g-1 graphite increased the removal up to 93 4% after 1 h and a complete removal was obtained in less than 3 h (Fig. 2). A semi-quantitative analysis of the complete deiodinated metabolite (diaH3) revealed that only in the experimental setup with bio-Pd coated granules, complete deiodination was achieved. All samples were also analyzed for diaI3 metabolites with one (diaH) or two iodine (diaH2) atoms cleaved, but they could not be detected.

TCE. During closed circuit operation, no signicant removal could be observed in the MFC mode (applying no external potential) indicating that the cathode potential [-(0.0030.103) V vs. SHE] was not low enough to deliver electrons from the cathode to be used for TCE reduction. However, at -0.6 and -0.8 V Cl- formation was signicantly enhanced, and the cathode potential was -0.8 V vs. SHE and -0.95 V vs. SHE, respectively. H2 formation and consequently H? reduction was only visually observed at -0.8 V, demonstrating that the cathode was used as the electron donor for TCE reduction. Hence, when applying -0.6 V, electrons are delivered to TCE but applying -0.8 V may result in a competition between H? and TCE for electrons. The formation of 25 2 and 50 3 mg Cl- l-1 corresponds to a 38 and 76% TCE removal or 60 and 120 g TCE m-3 TCC day-1. The applied voltages in the range of -0.6 to -0.8 V used in the MEC were much lower compared to those required in previous studies in electrochemical cells. For example, removal levels of 75, 95 and 99% were obtained using 5, 10 and 15 V in electrolytic dechlorination of TCE on granular graphite (Al-Abed and Fang 2007). In comparison to these values, the energy requirements in our work applying only -0.8 V were greatly decreased. As a renewable source, the microbial oxidation of acetate at the anode provided the electrons for reductive TCE removal. Moreover, byproducts were only detected at concentrations around the detection limits. This was in contrast with previous studies in which for example chloromethane was found as major end-product (Al-Abed and Fang 2007). TCE dehalogenation in MECs containing bio-Pd

Discussion TCE dehalogenation in MECs In this study, we investigated the use of a MEC to remove halogenated compounds at a cathode coupled to microbial oxidation of acetate at an anode. In the case of TCE, dechlorination assays were only based on Cl- formation, as TCE was subject to adsorption on the graphite granules and volatilization to the anode. Under open circuit conditions, 5.0 0.3 mg Cl- l-1 was formed. This only accounts for 6% TCE removal and could be attributed to non-electrochemical processes of impurities on graphite granules (Freguia et al. 2007), which may catalytically reduce To investigate the inuence of bio-Pd on TCE dehalogenation, cathode granules were coated with bio-Pd at two different concentrations. In case of the 1 mg Pd g-1 graphite, the TCE removal deteriorated while using 5 mg Pd g-1 graphite enhanced the dehalogenation process. However, batch tests pointed out that the 1 mg Pd g-1 graphite coated granules were catalytically active for hydrodechlorination reaction of TCE. The total removal in the MEC trials using 5 mg Pd g-1 graphite was 151 g TCE m-3 TCC day-1 or 48 mg TCE g-1 Pd day-1. Visible and signicant H2 formation was observed starting from

123

94

Biotechnol Lett (2011) 33:8995

an applied voltage of -0.8 V. It is hypothesized that two mechanisms can occur at the same time: (1) electrochemical reduction of TCE and (2) catalytical dehalogenation with bio-Pd as catalyst and electrochemically formed H2 as the reactant. Coating the granules with bio-Pd might negatively affect the rst mechanism due to the bacterial matrix which is electrically less conductive, but can enhance the second mechanism. Hence, also the Pd concentration plays a crucial role as only 5 mg Pd g-1 graphite resulted in higher removal rates. Thus, opting for the second mechanism requires a certain threshold of bioPd catalyst (i.e., [1 mg Pd g-1 graphite) and a cathode potential that allows for H2 production. In this study, -0.8 V was applied to obtain H2 at a cathodic potential of about -1 V vs. SHE. However, Rozendal et al. (2006) obtained H2 production at an applied voltage of -0.5 V and expected to lower this value to -(0.30.4) V. In any case, biocatalyzed electrolysis achieves very low energy requirements for H2 production while water electrolysis in practice operates at applied voltages well above 1.6 V (Rasten et al. 2003). This means that biocatalyzed electrolysis requires four (Rozendal et al. 2006) to two (this study) times less energy expenditure. The cost of H2 produced through water electrolysis is strongly dependent on the price of the electricity consumed (Turner 2004). Hence, the lower consumption of electrical energy per unit of H2 is a strong advantage of biocatalyzed electrolysis. This opens perspectives for the economical and sustainable production of H2 for the bio-Pd catalyzed reaction, as H2 consumption was one of the major challenges related to bio-Pd technologies (Hennebel et al. 2009b, 2010a, b). The use of bioelectrochemically produced H2 would decrease the costs of this process with a factor of at least 24. Recently, Foley et al. (2010) demonstrated through life cycle analysis (LCA) that MECs producing H2O2 have a net positive environmental benet. A comparable LCA needs to point out if our suggested technology could obtain signicant environmental benets as well. DiaI3 removal in MECs with and without bio-Pd The removal of diaI3 in a MEC containing non-coated and bio-Pd coated granules was also studied. Under open circuit conditions, no removal of diaI3 was observed but applying an external voltage of -0.8 V

resulted in a complete removal with and without bioPd. As in the case of TCE, it was demonstrated that the cathode was used as electron donor for diaI3 reduction without bio-Pd. The complete deiodinated compound, diaH3, was only detected in the experiments with bio-Pd. Moreover, the dehalogenation with bio-Pd graphite occurred more rapidly. These observations support our hypothesis that the reduction mechanisms are different using bio-Pd coated granules versus noncoated granules. With bio-Pd, a simple catalytical reaction with bio-Pd and bioelectrochemically produced H2 takes place comparable to previous reports using bio-Pd or Pd as catalyst and molecular H2 as reactant (Knitt et al. 2008; Hennebel et al. 2010a). Without bio-Pd, diaI3 is bioelectrochemically reduced at the cathode to an undened compound not being diaH3 or the partially deiodinated diaH and diaH2. In conclusion, we have developed an innovative technology with potential for sustainable removal of halogenated pollutants from water coupling biogenic Pd nanoparticles to MECs.
Acknowledgments This work was part of a research project obtained from the EU Commission within the Energy, Global Change and Ecosystems Program of the Sixth Framework (FP62005-Global-4): EU Neptune project (036845, SUSTDEV2005-3.II.3.2). Tom Hennebel was supported by the Fund of Scientic Research-Flanders (FWO, 7741-02) and the German Academic Exchange Service (DAAD) is acknowledged for the postdoctoral fellowship of Jessica Benner.

References
Al-Abed SR, Fang YX (2007) Use of granular graphite for electrolytic dechlorination of trichloroethylene. Environ Eng Sci 24(6):842851 Aulenta F, Canosa A et al (2008) Trichloroethene dechlorination and H-2 evolution are alternative biological pathways of electric charge utilization by a dechlorinating culture in a bioelectrochemical system. Environ Sci Technol 42(16):61856190 Baxter-Plant V, Mikheenko IP et al (2003) Sulphate-reducing bacteria, palladium and the reductive dehalogenation of chlorinated aromatic compounds. Biodegradation 14(2): 8390 Clauwaert P, Desloover J et al (2009) Enhanced nitrogen removal in bio-electrochemical systems by pH control. Biotechnol Lett 31(10):15371543 De Windt W, Aelterman P et al (2005) Bioreductive deposition of palladium (0) nanoparticles on Shewanella oneidensis with catalytic activity towards reductive dechlorination of polychlorinated biphenyls. Environ Microbiol 7(3):314325

123

Biotechnol Lett (2011) 33:8995 De Windt W, Boon N et al (2006) Biological control of the size and reactivity of catalytic Pd(0) produced by Shewanella oneidensis. Antonie van Leeuwenhoek 90(4):377389 Deplanche K, Snape TJ et al (2009) Versatility of a new bioinorganic catalyst: palladized cells of Desulfovibrio desulfuricans and application to dehalogenation of ame retardant materials. Environ Technol 30(7):681692 Foley JM, Rozendal RA et al (2010) Life cycle assessment of high-rate anaerobic treatment, microbial fuel cells, and microbial electrolysis cells. Environ Sci Technol 44(9): 36293637 Freguia S, Rabaey K et al (2007) Non-catalyzed cathodic oxygen reduction at graphite granules in microbial fuel cells. Electrochim Acta 53(2):598603 Harrad S, Robson M et al (2007) Dehalogenation of polychlorinated biphenyls and polybrominated diphenyl ethers using a hybrid bioinorganic catalyst. J Environ Monit 9(4): 314318 Hennebel T, De Gusseme B et al (2009a) Biogenic metals in advanced water treatment. Trends Biotechnol 27(2): 9098 Hennebel T, Simoen H et al (2009b) Biocatalytic dechlorination of trichloroethylene with bio-Pd in a pilot-scale membrane reactor. Biotechnol Bioeng 102(4):9951002 Hennebel T, Verhagen P et al (2009c) Remediation of trichloroethylene by bio-precipitated and encapsulated palladium nanoparticles in a xed bed reactor. Chemosphere 76(9):12211225

95 Hennebel T, De Corte S et al (2010a) Removal of diatrizoate with catalytically active membranes incorporating microbially produced palladium nanoparticles. Water Res 44(5):14981506 Hennebel T, Simoen H, et al (2010b) Biocatalytic dechlorination of hexachlorocyclohexane by immobilized bio-Pd in a pilot scale uidized bed reactor. Environ Chem Lett. doi:10.1007/s10311-010-0295-x Knitt LE, Shapley JR et al (2008) Rapid metal-catalyzed hydrodehalogenation of iodinated X-ray contrast media. Environ Sci Technol 42(2):577583 Mertens B, Blothe C et al (2007) Biocatalytic dechlorination of lindane by nano-scale particles of Pd(0) deposited on Shewanella oneidensis. Chemosphere 66(1):99105 Mu Y, Rabaey K et al (2009a) Decolorization of azo dyes in bioelectrochemical systems. Environ Sci Technol 43(13): 51375143 Mu Y, Rozendal RA et al (2009b) Nitrobenzene removal in bioelectrochemical systems. Environ Sci Technol 43(22): 86908695 Rasten E, Hagen G et al (2003) Electrocatalysis in water electrolysis with solid polymer electrolyte. Electrochim Acta 48(2526):39453952 Rozendal RA, Hamelers HVM et al (2006) Principle and perspectives of hydrogen production through biocatalyzed electrolysis. Int J Hydrogen Energy 31(12):16321640 Turner JA (2004) Sustainable hydrogen production. Science 305(5686):972974

123