Inhibition of Biohydrogen Production by Undissociated Acetic and Butyric Acids STEVEN VAN GINKEL AND BRUCE E. LOGAN* Department of Civil and Environmental Engineering, 212 Sackett Building, Penn State University, University Park, Pennsylvania 16801 Glucose fermentation to hydrogen results in the production of acetic and butyric acids. The inhibitory effect of these acids on hydrogen yield was examined by either adding these acids into the feed of continuous flow reactors (external acids), or by increasing glucose concentrations to increase the concentrations of acids produced by the bacteria (self-produced). Acids added to the feed at a concentration of 25 mM decreased H 2 yields by 13% (acetic) and 22% (butyric), and 60 mM (pH 5.0) of either acid decreased H 2 production by >93% (undissociated acid concentrations). H 2 yields were constant at 2.0 ( 0.2 mol H 2 / mol glucose for an influent glucose concentration of 10- 30 g/L. At 40 g glucose/L, H 2 yields decreased to 1.6 ( 0.1 mol H 2 /mol glucose, and a switch to solventogenesis occurred. A total undissociated acid concentration of 19 mM (self-produced acids) was found to be a threshold concentration for significantly decreasing H 2 yields and initiating solventogenesis. Hydrogen yields were inhibited more by self-produced acids (produced at high glucose feed concentrations) than by similar concentrations of externally added acids (lower glucose feed concentrations). These results show the reason hydrogen yields can be maximized by using lower glucose feed concentrations is that the concentrations of self-produced volatile acids (particularly butyric acid) are minimized. Introduction One way to make renewable biohydrogen production costs competitive with hydrogen production from fossil fuels is to produce H2 using inexpensive substrates in reactors that have low capital costs. Economical biohydrogen production can be accomplished using high-strength wastestreams such as food processing wastewaters in reactors operating at short hydraulic retention times (1-3). However, cultures fed high sugar concentrations are susceptible to product inhibition. The products are primarily H2 and CO2 gases and acetic and butyric acids. Many authors have studied the effects of these acids on solvent production in the traditional acetone- butanol-ethanol (ABE) fermentation (4-6). It has been shown that hydrogen fermentation changes to a solvent- forming reaction once the undissociated acid concentration reaches a critical threshold, but the threshold cannot be well predicted as it varies substantially over the range of ∼2-30 mM (undissociated acid concentration) (5-9). The undis- sociated form of the acid is known to be the most important factor that causes the switch from H2 production to solvent production, although the total acid concentration also affects solution ionic strength (5-10). Inhibition is caused by nonpolar undissociated acids being able to cross the cell membrane at a low pH that then dissociate in the cell at the higher internal pH releasing a proton inside the cell (10). The uptake of protons in this way uncouples the proton motive force, which causes an increase in maintenance energy requirements to maintain the intracellular pH near neutrality (10). The uptake of acid also causes a decrease in the available coenzyme A and phosphate pools which decreases the flux of glucose through glycolysis (4). A detailed description of the mechanisms of undissociated acid inhibi- tion can be found in Jones and Woods (10). A major factor affecting H2 or solvent production is the pH. The optimum pH for H2 production is 5.5, while the optimum pH for solvent production is ∼ 4.5 (10-12). The pH can affect the form of the acids produced during hydrogen production. The concentrations of the undissociated forms of acetic or butryic acid are greater at a pH of 4.5 than at pH 5.5 and thus higher amounts of the undissociated form are present at the lower pH to cause inhibition. Undissociated acid concentrations can also be increased by increasing the substrate concentration from which the acids are produced. The effect of the undissociated acid on the bacteria therefore sets an upper limit on the influent substrate concentration, as a function of pH, for hydrogen production. The concentration of undissociated acids that inhibits hydrogen production has not been specifically studied, although it can be inferred from the literature that acid concentrations that promote solventogenesis inhibit hydro- gen production (4-10). Due to a wide range in concentrations reported to promote solventogenesis (2-30 mM), however, the concentrations that inhibit hydrogen production using a heat-treated soil innoculum cannot be predicted based on the literature. In this study, we therefore determined acid concentrations that caused a significant decrease in H2 yields using a continuous flow reactor system. Acetic and butyric acids were varied by either adding them directly to the feed at low glucose feed concentrations, or by increasing the glucose feed concentration to increase the acids produced by the bacteria as fermentation end products. Experimental Section Startup and Feeding. Two identical stirred fermentors (2-L New Brunswick BioFlo, Edison, NJ) operated in continuous flow mode were used in these experiments. Reactors were cleaned, filled to 2 L with distilled water, autoclaved for 1 h, and then sparged with N2 gas for approximately thirty minutes before inoculation. Each reactor was inoculated with 10 g of baked and sieved agricultural soil. The soil was baked for 2 h at 100 °C to select for spore-forming, H2-producing bacteria (11), a process which has been found to result in a microbial community primarily composed of clostridial species (13). Hydrogen production from this inoculum source has been shown to be consistent even when stored (-80 °C) over several months (3, 11, 14). All reactors were operated at 30 °C. During startup, L-cysteine (1 g/L) was added to the reactor when it was operated in batch mode to remove dissolved oxygen. Nutrients were added as described previously (14). Feed bottles (10 L) were continuously sparged with N2 gas and mixed (∼250 rpm) to achieve anaerobic conditions in the feed bottle. Reactor Operation. Several conditions were systematically employed to determine the effect of undissociated acid concentrations on H2 yields. The influent glucose concentra- * Corresponding author e-mail: blogan@psu.edu; phone: 814- 863-7908; fax: 814-863-7304. Environ. Sci. Technol. 2005, 39, 9351-9356 10.1021/es0510515 CCC: $30.25  2005 American Chemical Society VOL. 39, NO. 23, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 9351 Published on Web 11/02/2005