Enzymatic Hydrolysis of Cellulose Coupled With Electricity Generation in a Microbial Fuel Cell

Transcription

1 ARTICLE Enzymatic Hydrolysis of Cellulose Coupled With Electricity Generation in a Microbial Fuel Cell Farzaneh Rezaei, 1 Tom L. Richard, 1 Bruce E. Logan 2y 1 Department of Agricultural and Biological Engineering, The Pennsylvania State University, University Park, Pennsylvania 2 Department of Civil and Environmental Engineering, The Pennsylvania State University, University Park, Pennsylvania 16802; telephone: ; fax: ; blogan@psu.edu Received 30 March 2008; revision received 19 May 2008; accepted 29 May 2008 Published online 11 June 2008 in Wiley InterScience ( DOI /bit ABSTRACT: Electricity can be directly generated by bacteria in microbial fuel cells (MFCs) from a variety of biodegradable substrates, including cellulose. Particulate materials have not been extensively examined for power generation in MFCs, but in general power densities are lower than those produced with soluble substrates under similar conditions likely as a result of slow hydrolysis rates of the particles. Cellulases are used to achieve rapid conversion of cellulose to sugar for ethanol production, but these enzymes have not been previously tested for their effectiveness in MFCs. It was not known if cellulases would remain active in an MFC in the presence of exoelectrogenic bacteria or if enzymes might hinder power production by adversely affecting the bacteria. Electricity generation from cellulose was therefore examined in two-chamber MFCs in the presence and absence of cellulases. The maximum power density with enzymes and cellulose was mw/m 2 ( W/m 3 ), compared to only mw/m 2 ( W/m 3 ) in the absence of the enzymes. This power density was comparable to that achieved in the same system using glucose (102 7 mw/m 2, W/m 3 ) suggesting that the enzyme successfully hydrolyzed cellulose and did not otherwise inhibit electricity production by the bacteria. The addition of the enzyme doubled the Coulombic efficiency (CE) toce ¼ 51% and increased COD removal to 73%, likely as a result of rapid hydrolysis of cellulose in the reactor and biodegradation of the enzyme. These results demonstrate that cellulases do not adversely affect exoelectrogenic bacteria that produce power in an MFC, and that Graduate Student. Associate Professor of Agricultural and Biological Engineering. y Kappe Professor of Environmental Engineering. Correspondence to: B.E. Logan Contract grant sponsor: Department of Agricultural and Biological Engineering of Pennsylvania State University Contract grant sponsor: NREL Contract grant number: RFH Additional Supporting Information may be found in the online version of this article. the use of these enzymes can increase power densities and reactor performance. Biotechnol. Bioeng. 2008;101: ß 2008 Wiley Periodicals, Inc. KEYWORDS: microbial fuel cell; enzymatic hydrolysis; cellulose hydrolysis; complex substrate Introduction Cellulose is the most abundant biopolymer on the earth, and is the focus of considerable interest as a renewable energy resource. Cellulosic materials are desirable feedstocks for alternative fuels and energy carriers such as ethanol, biodiesel, or hydrogen since they are renewable and abundant (Mielenz, 2001; Ni et al., 2006; Powlson et al., 2005). Cellulose can also be used as a fuel in a microbial fuel cell (MFC) for direct electricity production (Ren et al., 2007; Rezaei et al., 2007; Rismani-Yazdi et al., 2007) using exoelectrogenic bacteria. The complex structure and very low hydrolysis rate of cellulose presents challenges for using this material in an MFC (Ren et al., 2007; Rezaei et al., 2007). Both cellulolytic and exoelectrogenic microorganisms are needed in an MFC as no single strain has yet been shown to be able to generate electricity directly from cellulose. Maximum power densities produced in MFCs using cellulose have been lower than those achieved using soluble substrates. For example, a co-culture of cellulolytic fermentor Clostridium cellulolyticum and the exoelectrogen Geobacter sulfurreducens generated electricity, and the power generated using soluble cellulose (carboxymethyl cellulose; CMC), was comparable (143 mw/m 2, R ¼ 1,000 V) to that obtained with glucose. However, the maximum power density generated using insoluble cellulose (MN301) was 58% less (59 mw/m 2 ) than that with CMC. The removal of insoluble cellulose based on chemical oxygen demand (COD) was <30% and cellulose degradation was <50% ß 2008 Wiley Periodicals, Inc. Biotechnology and Bioengineering, Vol. 101, No. 6, December 15,

2 (Ren et al., 2007). Power generation and cellulose degradation were lower when a mixed culture (sludge) was used instead of a co-culture. A consortium of rumen bacteria was found to produce electricity using cellulose (Rismani-Yazdi et al., 2007) at a power density (55 mw/m 2 ) comparable to that obtained using mixed soluble carbohydrates in the same reactor set up. Although, the cellulose degradation rate was not measured in this study, it was claimed that power output was not limited by cellulose hydrolysis rate. It should be noted that in both of these previous studies by Rismani-Yazdi et al. (2007) and Ren et al. (2007) that ferricyanide was used as electron acceptor at the cathode. The absence of any oxygen being used in the system can be important, as oxygen diffusion into the anode chamber has been found to affect the enrichment and operation of MFCs for some substrates (Kim et al., 2007). One way to improve power generation is to use enzymes to increase the hydrolysis rate of cellulose. Enzymatic hydrolysis is widely used in bioethanol production to achieve higher rate of cellulose to sugar conversion. Cellulase usually refers to a mixture of different enzymes involved in cellulose hydrolysis to glucose, which can include endoglucanase, cellobiohydrolase, and b-glucosidase (Reese, 1975). One limitation with using cellulase is that there is a reduction in rates due to end product (cellobiose and glucose) inhibition. Simultaneous saccharification and fermentation (SSF) overcomes this problem by hydrolyzing cellulose and fermenting the hydrolysis product at the same time (Sun and Cheng, 2002). We hypothesized that enzymatic hydrolysis of cellulose could improve electricity generation rate in MFC s using cellulosic substrates. It has not been previously shown, however, whether the presence of the exoelectrogenic bacteria in the MFC would inhibit enzyme activity. Also, it is not known if the enzyme might adversely affect the exoelectrogenic bacteria. The main objective of this research was therefore to show that adding cellulases would increase the hydrolysis rate of cellulose without interfering with the activity of electrochemically active bacteria and consequently, that adding these enzymes would increase power output. To show the effect of enzyme on power output, power generation with both the cellulose and the enzyme (cellulose þ enzyme) was compared to that with only cellulose, and with reactors fed either only enzyme or glucose. Furthermore, to monitor the effect of microorganisms on enzymatic hydrolysis of cellulose, the hydrolysis rate in an active MFC was compared to the hydrolysis rate in an abiotic control. Materials and Methods Medium The medium was a phosphate buffered solution (PBS; ph ¼ 7) containing (g/l): NH 4 Cl, 0.31; KCl, 0.13; NaH 2 PO 4 H 2 O, 2.45; Na 2 HPO 4, 4.576; and mineral (12.5 ml/l) and vitamin (12.5 ml/l) solutions (Lovley and Philips, 1988). Cellulose, cellulase, cellulose þ cellulase (50% each), and glucose were used as substrates, with each total substrate added at a concentration at g-cod/l (Table I). Cellulose was microcrystalline insoluble cellulose, 15% amorphous and 85% crystalline (Fan et al., 1980), of type 50 50, cotton linters, having a 50-mm particle size (Sigmacell 1, Sigma-Aldrich Co., St. Louis. MO). The term cellulase as used here refers to a combination of Novozyme 188 (b-glucosidase) and Celloclast 1.5L (synonymous with a Novozyme cellulase from Trichoderma reesei ATCC 26921) (Sigma-Aldrich Co.; Ramos et al., 1993). Cellulase was added to the cellulose with the loading rate equal to 30 cellobiose units (CBU) of Novozyme 188 and 15 filter paper units (FPUs) of Celloclast 1.5L per gram of cellulose (Ramos et al., 1993). MFC Construction and Operation A two-chamber MFC was used that consisted of two media bottles (200 ml capacity, Penn State glass workshop, University Park, PA), joined together via a glass tube on either side of a cation exchange membrane (Nafion 117, Dupont Co., Newark, DE) as previously described (Oh and Logan, 2006). Oxygen was used as the electron acceptor in the cathode by air sparging. The cathode was 1 cm 10 cm (A cat ¼ m 2 ) carbon paper coated on one side with a 10% Pt/C catalyst applied at a rate of 0.35 mg Pt/cm 2 (10% of Pt/C catalyst, 30% wet-proofing; E-TEK, Somerset, NJ). A graphite brush electrode (25 mm diameter 25 mm length, A an ¼ 0.22 m 2 ) was used as the anode (Logan et al., 2007). Each reactor was inoculated once with sludge (20%) from secondary clarifier obtained from the Pennsylvania State University Wastewater Treatment Plant and 50 mm PBS medium. Reactors were then run in fed-batch mode for several cycles using only media to achieve at least two Table I. Four different treatments with their total initial substrate concentrations. Parameter Cellulose Enzyme Cellulose þ enzyme Glucose Particulate COD concentration (g/l) Soluble COD concentration (g/l) Cellulose (g/l) Enzyme (ml) 276 C and 21.2 N 138 C and 10.6 N Glucose (g/l) 1.0 C ¼ Celloclast; N ¼ Novozyme Biotechnology and Bioengineering, Vol. 101, No. 6, December 15, 2008

3 consecutive batches with the same maximum power density. The medium was then changed to a 200 mm PBS medium for the rest of experiments. Medium was replaced when the voltage dropped to <50 mv. Abiotic Rate of Cellulose Hydrolysis A control experiment was conducted in the absence of bacteria (abiotic control) to determine the effect of bacteria on enzymatic hydrolysis rates. The amount of glucose produced and the amount of cellulose consumed were measured in a flask supplemented with cellulose and cellulase (same loading rate as above) under sterile conditions. Samples (3 ml) taken over 168 h were analyzed for cellulose, glucose, and cellobiose concentrations. At the end of experiment, a sample was placed on an YPD (yeast, peptone, dextrose) agar plate to monitor bacterial contamination. The ph of the last sample was also monitored to ensure constant ph conditions and the lack of ph changes due to microbial activity. Analytics and Calculations COD was measured using a colorimetric assay (HACH, DR2010 spectrophotometer) based on a standard method (APHA et al., 1992). Soluble COD (SCOD) was measured after centrifugation (Eppendorf centrifuge, 5403) and filtration of the supernatant using a 0.2 mm pore-diameter syringe filter. Particulate COD was measured after removing bacterial biomass by centrifuging and washing the samples with 10 mm PBS (10 mm Na 2 HPO 4, 130 mm NaCl, ph 7.2) followed by 0.2 N sodium hydroxide (NaOH) as previously described (Ren et al., 2007). Samples were taken from the reactors at 0, 3, 24, and 168 h as well as at the end of batch. Cellulose concentrations in the samples were measured by removing bacterial biomass as described above. Cellulose pellets were solubilized in 67% (v/v) sulfuric acid (Updegraff, 1969), and the concentration measured using by the phenol-sulfuric acid calorimetric method (Dubois et al., 1956). Glucose and cellobiose concentration were measured using ion chromatography, with the PA20 column and 10 mm NaOH as the mobile phase (Dionex model ICS-3000, Sunnyvale, CA). Enzyme activities of Celloclast 1.5L and the mixed solution (mixture of the two enzymes, Celloclast 1.5L and Novozyme 188) were measured based on the FPU using standard methods (Ghose, 1987; Mandels et al., 1981). Initial activity of Novozyme 188 was measured based on the cellobiose assay (CBU; Sternberg et al., 1977). Samples taken from MFC were filtered (0.2 mm pore-diameter) to remove microorganisms prior to measurement of enzyme activity and concentration. The total enzyme concentration was estimated based on measurement of soluble protein using the Bradford protein assay (Bradford, 1976). Activity of the enzyme was measured based on FPU for each sample taken from cellulose þ enzyme reactor and enzyme alone reactor. Cell voltages (V) were measured across the external circuit containing a resistor (1,000 V, except as noted below) using a data acquisition system (2700, Keithly, Cleveland, OH) connected to a computer. Voltages were recorded every 20 min. Current (i) (A) was calculated as i ¼ V/R, where R is the external circuit resistance (V). Power ( P) (W) was calculated as P ¼ iv, and normalized by the cathode area (A cat )(m 2 ). The maximum power was calculated from the voltage measured over n points at the beginning of the cycle when the highest voltage was stable for the indicated period of time. Power density (mw/m 2 ) was calculated by dividing the power (1,000 iv for the given units) by the one side of the cathode surface area (containing the catalyst) as the cathode is the main factor limiting power production in an MFC (Rabaey and Keller, 2008; Rismani-Yazdi et al., 2008). The volumetric power density (W/m 3 ) was calculated based the volume of the anode chamber ( m 3 ). Polarization curves were obtained by varying the circuit load by changing the external resistor from 4 V to 200 kv after leaving the reactor under open circuit potential (OCP) for an hour, with each measurement taken under maximum power output conditions (with fresh substrate) after the voltage had stabilized (20 min). The internal resistance calculated as the slope of the voltage current graph obtained by varying the external resistance. Coulombic efficiency (number of electron recovered as current in the reactor to the total electron stored in biomass) was calculated based on COD removal as previously described (Cheng et al., 2006). Results Power Generation With Different Substrates When cellulose was added to MFCs with cellulase, the average maximum power density was mw/m 2 ( W/m 3 )(n¼1,377, based on voltages measured over 489 h). This power density was comparable to that achieved using the same system with glucose (102 7 mw/ m 2, W/m 3 ; n ¼ 1,447, 482 h). Power generation by the reactor fed cellulose þ enzyme was not limited by the concentration of cellulose. Increasing the initial cellulose concentration from 0.5 to 1 g/l did not increase the maximum power density (102 8 mw/m 2, W/m 3 ; data not shown). Enzymes were used as a substrate for power generation, as shown power production of mw/m 2, W/m 3,(n¼536, 178 h) which is similar to that obtained in the other two reactors (Fig. 1). The longevity of the reactor with only enzyme, however, was 10 days less than that of the cellulose þ enzyme reactor. These results demonstrate that cellulases not only increased power generation to levels similar to that with glucose, but that they were subsequently used as an additional substrate for power generation. The voltage curves for all treatments changed with the first two cycles, and were then constant for the third and Rezaei et al.: Electricity Generation From Cellulose 1165 Biotechnology and Bioengineering

4 Figure 1. Power density (normalized to cathode surface area, constant 1,000 V load) as a function of time from two-chamber reactor fed with cellulose þ enzyme, glucose, cellulose alone, and enzyme alone. [Color figure can be seen in the online version of this article, available at fourth cycles except for the reactor fed only cellulose (control). Repeated attempts to re-start the reactor using cellulose and no enzyme did not result in stable power generation after multiple batch cycles. Typically, the reactor fed only cellulose would generate powering the first cycle, and then it would fail to produce power when refilled. The reactor could be re-started by adding fresh inoculum, but power in the second cycle was not stable. Power generated during this first cycle was 48% less (52 2 mw/m 2 ( W/m 3 ), n ¼ 451, 150 h) than that produced in repeatable cycles using the cellulose þ enzyme or glucose MFCs, but part of this power generation could be due to the organic matter in the wastewater inoculum. Others (Ren et al., 2007; Rismani-Yazdi et al., 2007) have found power generation using in two-chamber reactors, but oxygen was not used in the second chamber and thus there was no potential for oxygen contamination of the system. We hypothesized that the lack of power generation on subsequent feeding cycles was due to: (1) oxygen leakage into the anode chamber; (2) the removal after each batch cycle of microorganisms needed to rapidly scavenge oxygen; and (3) removal of microorganisms needed for cellulose hydrolysis that were likely attached to the cellulose. Therefore, we modified our procedure to leave ml of solution from each batch cycle in the reactor for the next batch cycle. With this modification, we achieved stable and reproducible power generation over successive batch cycles, with a maximum power density of mw/m 2 ( W/m 3 ; n ¼ 16; Fig. 1). This is substantially less power than that generated using enzymes. Figure 2. A: Power as a function of current for three treatments, cellulose þ enzyme, enzyme, and glucose (error bars are SD based on duplicate measurements) obtained by varying external resistance and (B) voltage as a function of current for three treatments, cellulose þ enzyme, enzyme, and glucose (error bars are SD based on duplicate measurements) obtained by varying external resistance. [Color figure can be seen in the online version of this article, available at wiley.com.] Maximum Power Density Polarization curves were obtained to determine if the different systems could achieve higher power densities using lower external resistances. The maximum power densities for the MFCs using cellulose þ enzyme ( mw/m 2, W/m 3 ) or glucose (103 3 mw/m 2, W/m 3 ) were essentially unchanged relative to the constant resistance results tests. The maximum power density obtained using enzyme alone was about the same level (114 2 mw/m 2, W/m 3 ) as glucose and cellulose þ enzyme, where the slight difference in power could be a result of cathode performance (Fig. 2A). The internal resistance was 1, V (95% confidence interval) for the cellulose þ enzyme reactor, with a lower value found for the reactor fed only enzyme ( V; Fig. 2B). Enzymatic Hydrolysis Rate of Cellulose in the Presence and Absence of Bacteria In absence of microorganisms approximately 20% of the cellulose was hydrolyzed during the first 24 h (Fig. 3A). After 168 h, 51% of the cellulose was degraded. The cellobiose concentration was initially low (<0.1 g/l) and was not 1166 Biotechnology and Bioengineering, Vol. 101, No. 6, December 15, 2008

5 Figure 4. Enzyme activity in MFC reactors with and without cellulose. [Color figure can be seen in the online version of this article, available at wiley.com.] Figure 3. A: Enzymatic hydrolysis of cellulose without microbes. B: Cellulose concentration remaining inside the reactor during enzymatic hydrolysis without microbes (abiotic) versus simultaneous saccharification and oxidation in an MFC. Data are cellulose concentrations inside the reactor relative to the initial cellulose concentration. [Color figure can be seen in the online version of this article, available at measurable (<0.002 g/l) after 72 h (data not shown). The final and initial phs of the reactor were the same (6.8), suggesting no substantial activity by fermentative microorganisms. Moreover, there were no colonies on the YPD agar plates, confirming that the conditions remained free of microorganisms throughout the experiments. In the presence of bacteria, using samples obtained directly from the MFC, 50% of the initial cellulose was degraded during the first 24 h (Fig. 3B). This is 1.5 times higher than that achieved in the abiotic environment. After 168 h, 60% of the initial cellulose was degraded showing that the hydrolysis rates decreased after 24 h (Fig. 3B). Comparing the cellulose degradation inside the MFC with the abiotic reactors showed that cellulose degraded faster inside the MFC. Although the concentration of end product that is inhibitory to the enzyme varies with different substrates and enzymes (Gruno et al., 2004), there is a possibility that the higher degradation rate inside the MFC was due to removal of chemicals by microorganisms that may have contributed to end-product inhibition of enzyme activity. The concentration of cellobiose or glucose was not detectable (<0.002 g/l) in the samples taken from MFC reactor. Cellulase Activity, Cellulase Consumption, and Cellulose Degradation The activity of the remaining enzyme in the cellulose þ enzyme MFC declined by 25% during the first 24 h of the experiment, whereas the reactor containing only enzyme lost more than 60% of its activity during the same period (Fig. 4). At the end of a batch cycle, the activity of the remaining enzyme inside the cellulose þ enzyme reactor was 66% of the initial activity, whereas in the reactor containing only enzyme it was 26% of the initial activity (Table II). The enzyme concentration based on soluble protein content was compared in the MFC containing cellulose þ enzyme and MFC containing only enzyme. The percent concentration of remaining enzyme in solution was lower in reactor containing only enzyme (15%) as compared to cellulose þ enzyme reactor (36%; Table II). A lower enzyme concentration shows that more enzyme was consumed inside the reactor containing only enzyme than in the cellulose þ enzyme reactor. In the cellulose þ enzyme MFC, 74% of the initial cellulose was degraded at the end of cycle. In the reactor fed only cellulose 15% of the initial cellulose was degraded at the end of cycle, suggesting that adding enzyme enhanced cellulose degradation and that it provided more soluble substrate to be oxidized by the exoelectrogenic microorganisms. Table II. Summary table of percent substrate consumption, COD removal, enzyme activity lost, and coulombic efficiency. Parameter (%) Cellulose Enzyme Cellulose þ enzyme Glucose Soluble COD removal N/A Particulate COD removal 14 N/A a 73 N/A Cellulose degraded 15 N/A 74 N/A Coulombic efficiency b Enzyme consumed N/A N/A Enzyme activity lost N/A N/A a N/A, not applicable. b Based on COD removal (soluble COD removal for enzyme alone and glucose, particulate COD removal for cellulose and total (particulate þ soluble) COD removal for cellulose þ enzyme). Rezaei et al.: Electricity Generation From Cellulose 1167 Biotechnology and Bioengineering

6 Coulombic Efficiency Coulombic efficiency measured for cellulose þ enzyme was slightly higher (51%) than the CE from the reactor containing only enzyme (44%) (Table II). Both values were comparable to that obtained using glucose (48%). Coulombic efficiency from the cellulose þ enzyme MFC was higher than that using only cellulose (23%) in a two-chamber MFC. COD Removal Soluble COD removal was 69% from the cellulose þ enzyme reactor and it was 86% in the reactor fed only enzyme (Table II). This amount of COD removal indicated that the cellulose þ enzyme reactor did not generate power from only cellulose or enzyme. Cellulosic COD removal was 73% in the cellulose þ enzyme reactor (Table II). The COD removal for an MFC containing only cellulose was 14%, which is 421% less than that achieved here when using enzymes with the cellulose. Discussion Previous reports have suggested that power generation in MFCs from insoluble cellulose can be limited by a low hydrolysis rate (Ren et al., 2007; Rezaei et al., 2007). Our results demonstrated that using enzymes to hydrolyze cellulose improved the hydrolysis rate, overall cellulose and COD removal, and increased the power density and coulombic efficiency of an MFC. Power production using insoluble crystalline cellulose was mw/m 2, a level that was comparable with that generated using glucose (102 7 mw/m 2 ) in the same system. This power density with cellulose þ enzyme was almost an order of magnitude greater than that achieved in this experiment using cellulose alone. It was also two times greater than that previously found using cellulose by others (Ren et al., 2007; Rismani- Yazdi et al., 2007) even though ferricyanide was used in these previous tests. It is well known that ferricyanide increases power production in MFCs compared systems using dissolved oxygen (Logan, 2008; Oh and Logan, 2006). The presence of microorganisms did not inhibit the hydrolysis rate of cellulose. In fact, the hydrolysis rate in the MFC producing electricity was slightly greater than that in an abiotic reactor likely due to less end-product inhibition through reduced levels of cellobiose and glucose. These observations are consistent with the findings of others that simultaneous saccharification and fermentation increases the hydrolysis rate of cellulose by reducing end-product inhibition (Sun and Cheng, 2002). Although, recycling the enzyme can effectively increase the yield of cellulose conversion and reduce the enzyme cost (Mes-Hartree et al., 1987), in our reactors there is little possibility that the enzyme added to the reactor could be reused as power was generated in the reactor from the enzymes. This degradation of the enzyme is not surprising as it has been shown that electricity can be generated in an MFC using proteins as substrates (Heilmann and Logan, 2006). There are additional possibilities for improving cellulolytic rates using enzymes, such as varying the temperature, optimizing the ph, using multiple reactors and continuous flow, or adding surfactants. We conducted our experiments at room temperature but the optimum temperature is 308C for fermentation and C for hydrolysis (Philippidis and Smith, 1995). Running the reactor at a higher temperature should therefore improve the kinetics. The optimum ph for enzymes used here are in the range of (Eriksson et al., 2002; Saha and Cotta, 2007), but low ph can adversely affect bacteria. Rates could be improved by determining a ph that optimizes both enzyme activity and microbial activity. Alternatively, cellulose hydrolysis could be conducted in a tank separate from the one used for power generation, perhaps allowing the recovery of enzyme from that first system. Although the rate of hydrolysis might be slower due to end-product inhibition, this disadvantage might be outweighed by the economic benefits of enzyme recovery. Immobilizing the enzyme and running the system in a continuous mode, rather than fed-batch mode, could also improve process performance and perhaps reduce the need for enzymes and thus decrease costs. Improving the activity of the enzyme would also make the process more economical. One of the problems with using cellulase is that it can become deactivated by adsorption onto cellulose (Converse et al., 1988). One way to prevent the irreversible deactivation of enzyme is the addition of a surfactant that can modify the surface of the cellulose by increasing desorption of cellulase (Sun and Cheng, 2002). These results demonstrate the technical feasibility of enzymatic hydrolysis of cellulose for direct electricity generation. It has also recently been shown that cellulose can be used to generate hydrogen gas at yields higher than that possible via fermentation in a microbial electrolysis cell (MEC; Cheng and Logan, 2007). Thus, it is likely that enzymatic treatment of cellulose could also be used as a method for hydrogen production in a similar system. Whether the use of cellulases will be economical in comparison to direct cellulose utilization will depend on costs of the enzyme and further improvements in current densities. The cost of enzymes are being reduced and power densities in MFCs are being increased, which combined may allow for new methods of renewable electricity generation in MFCs or hydrogen production in MECs. Support was provided by the Department of Agricultural and Biological Engineering of Pennsylvania State University, and NREL contract RFH to BEL. The authors are grateful to Megan Marshall and Zhiyong Ren for their helpful advice. References APHA, AWA, WPCF Standard methods for the examination of water and wastewater, 18th edn. Washington DC: American Public Health Association Biotechnology and Bioengineering, Vol. 101, No. 6, December 15, 2008

7 Bradford MM A rapid and sensitive method for quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: Cheng S, Logan BE Sustainable and efficient biohydrogen production via electrohydrogenesis. Proc Natl Acad Sci 104(47): Cheng S, Liu H, Logan BE Increased power generation in a continuous flow MFC with advective flow through the porous anode and reduced electrode spacing. Environ Sci Technol 40(7): Converse AO, Matsuno R, Tanaka M, Taniguchi M A model for enzyme adsorption and hydrolysis of microcrystalline cellulose with slow deactivation of the adsorbed enzyme. Biotechnol Bioeng 32: Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F Colorimetric method for determination of sugars and related substances. Anal Chem 28(3): Eriksson T, Börjesson J, Tjerneld F Mechanism of surfactant effect in enzymatic hydrolysis of lignocellulose. Enzyme Microb Technol 31(3): Fan LT, Lee Y-H, Beardmore DH Mechanism and enzymatic hydrolysis of cellulose: Effects of major structural features of cellulose on enzymatic hydrolysis. Biotechnol Bioeng 4: Ghose TK Measurement of cellulase activities. Pure Appl Chem 59: Gruno M, Valjamae P, Pettersson G, Johnsson G Inhibition of the Trichoderma reesei cellulases by cellobiose is strongly dependent on the nature of the substrate. Biotechnol Bioeng 86(5): Heilmann J, Logan BE Production of electricity from proteins using a single chamber microbial fuel cell. Water Environ Res 78(5): Kim JR, Jung SH, Regan JM, Logan BE Electricity generation and microbial community analysis of ethanol powered microbial fuel cells. Bioresource Technol 98(13): Logan BE Microbial fuel cells. New York: John Wiley & Sons, 200 p. Logan BE, Cheng S, Watson V, Estadt G Graphite fiber brush anodes for increased power production in air-cathode microbial fuel cells. Environ Sci Technol 41(9): Lovley DR, Philips EJP Novel mode of microbial energy metabolism: Organism carbon oxidation coupled to dissililatory reduction of iron and manganese. Appl Environ Microbiol 54(6): Mandels M, Andreotti R, Roche C Enzymatic hydrolysis of cellulose: Evaluation of culture filtrates under use conditions. Biotechnol Bioeng Symp 23: Mes-Hartree M, Hogan CM, Saddler JN Recycle of enzyme and substrate following enzymatic hydrolysis of steam pretreated aspenweed. Biotechnol Bioeng 30: Mielenz JR Ethanol production from biomass: Technology and commercialization status. Curr Opin Microbiol 4: Ni M, Leung DYC, Leung MKH, Sumathy K An overview of hydrogen production from biomass. Fuel Process Technol 87: Oh S-E, Logan BE Proton exchange membrane and electrode surface areas as factors that affect power generation in microbial fuel cells. Appl Microbiol Biotechnol 70: Philippidis GP, Smith TK Limiting factors in the simultaneous saccharification and fermentation process for conversion of cellulosic biomass to fuel ethanol. Appl Biochem Biotech 51/52: Powlson DS, Riche AB, Shield I Biofuels and other approaches for decreasing fossil fuel emissions from agriculture. Ann Appl Biol 146: Rabaey K, Keller J Microbial fuel cell cathodes: From bottleneck to prime opportunity? Water Sci Technol 57(5): Ramos LP, Breuil C, Saddler JN The use of enzyme recycling and the influence of sugar accumulation of cellulose hydrolysis by Trichoderma cellulases. Enzyme Microb Technol 15: Reese ET Biological transformation of wood, In: Liese W, editor. Berlin: Springer, p Ren Z, Ward TE, Regan JM Electricity production from cellulose in a microbial fuel cell using a defined binary culture. Environ Sci Technol 41(13): Rezaei F, Richard TL, Brennan RA, Logan BE Substrate-enhanced microbial fuel cells for improve remote power generation from sediment-based system. Environ Sci Technol 41(11): Rismani-Yazdi H, Christy AD, Dehority BA, Morrison M, Yu Z, Tuovinen OH Electricity generation from cellulose by rumen microorganisms in microbial fuel cells. Biotechnol Bioeng 97(6): Rismani-Yazdi H, Carver SM, Christy AD, Tuovinen OH Cathodic limitations in microbial fuel cells: An overview. J Power Sources 180: Saha BC, Cotta MA Enzymatic hydrolysis and fermentation of lime pretreated wheat straw to ethanol. J Chem Technol Biotechnol 82(10): Sternberg D, Kumar PV, Reese ET b-glucosidase: Microbial production and effect on enzymatic hydrolysis of cellulose. Can J Microbiol 23: Sun Y, Cheng J Hydrolysis of lignocellulosic materials for ethanol production: A review. Bioresource Technol 83:1 11. Updegraff DM Semimicro determination of cellulose in biological materials. Anal Biochem 32: Rezaei et al.: Electricity Generation From Cellulose 1169 Biotechnology and Bioengineering