At present, development of a practical and efficient hydrogen generation process is a growing concern among the research community. In the last decade, several methods such as mutagenesis, genetic modification or metabolic pathway control have been shown to improve hydrogen yield in laboratory scale experiments. These methods are based on the metabolic pathway and the enzymes which are involved during the fermentative hydrogen production process. Hydrogen evolution follows the NADH (Nicotinamide Adenine Dinucleotide) pathway described by the reaction [13.10] which is catalyzed by the enzyme of hydrogenase. Increased hydrogen yields could be achieved by shifting the chemical reaction so as to increase the amount of NADH usable for hydrogen production:

NADH + H+ ^ NAD+ + H2 [13.10]

NADH is usually generated by the catabolism of glucose to pyruvate via glycolysis. In general, the yield of hydrogen produced upon mixed acid fermentation of carbohydrates is quite lower than the maximum theoretical yields since sugars fermentation, in addition to VFAs, also leads to the formation of various reduced end-products, such as ethanol, butanol and lactate. These compounds contain additional hydrogen atoms that are not liberated as gas. Therefore, in order to maximize the yield of hydrogen, bacterial metabolism should be directed away from alcohols and reduced acids and towards VFAs production. The conversion of pyruvate to ethanol, butanediol, and lactic acid involves oxidation of NADH. The concentration of NADH would increase if the formation of these alcoholic and acidic metabolites could be blocked (Das and Veziroglu, 2001). Kumar et al. (2001) reported enhanced hydrogen yields by blocking the pathways of organic acid formation using the proton-suicide technique with NaBr and NaBrO3. A similar enhancement of hydrogen yield using

E. aerogenes HU-101 was reported by blocking the formation of alcoholic and acidic metabolites by both allyl alcohol and the proton-suicide technique (Mahyudin et al., 1997).

Operation conditions such as pH, HRT, temperature and hydrogen partial pressure are reported to have a significant effect on metabolic balance. C. acetobutyricum has the ability to produce solvents at pH values lower than 5 and under phosphate and iron limiting conditions. In order to obtain high hydrogen yields using C. acetobutyricum, a pH above 5, phosphate and iron concentrations above the limiting levels and glucose concentration below 12.5% are recommended (Dabrock et al., 1992). In addition, clostridia produce VFAs and hydrogen in the exponential growth phase and rapid alcohol production occurs in late growth phase (Lay, 2000). In order to shift the metabolic pathway towards VFAs production and away from solventogenesis, an application of a low HRT should be essential.

It is also reported that a hydrogen partial pressure higher than 60-100 Pa inhibits the hydrogen production process and in order to obtain maximum hydrogen yields, the hydrogen produced should be removed from the reactor system. For this reason many approaches have been proposed. Mizuno et al. (2000) showed that gas sparging with nitrogen enhanced hydrogen yield, while Voolapalli and Stuckey (1998) developed an applicable technique based on a submerged silicone-membrane dissolved gas extraction system, removing hydrogen and carbon dioxide from the reactor volume. Another potentially efficient method for removing hydrogen from the gas stream based on a heated palladium-silver membrane reactor has been proposed by Nielsen et al. (2001).

Another strategy for enhancement of hydrogen production by existing pathways can be sought by increasing the flux through gene knockouts of competing pathways or increased homologous expression of enzymes involved in the hydrogen-generating pathways. Up to now, the majority of attempts in laboratories employ the metabolic engineering of E. coli, because its genome can be easily manipulated, its metabolism is the best understood of all bacteria and it readily degrades a variety of sugars. For example, Yoshida et al. (2005) performed genetic recombination of E. coli in conjunction with process manipulation to elevate the efficiency of hydrogen production in the resultant strain SR13. The genetic modification resulted in 2.8-fold increase in hydrogen productivity of SR13 compared with the wild type strain. However, it is still unclear what pathways function under what environmental conditions and what the substrate specificities of all hydrogenase-coupled pathways are involved in E. coli (Laurinavichene et al., 2002b).

Metabolic engineering of other native hydrogen-producing microorganisms has so far been limited because there is poor knowledge regarding the existing pathways involved in hydrogen-production system (Jones, 2008). Despite this fact, there are few recent noteworthy examples of improvement in fermentation — based hydrogen production by either genetic or chemical engineering strategies of mesophilic hydrogen producing strains. For example, mutants of both E. aerogenes and E. cloacae have been isolated after subjecting wild-type strains to chemically selective media, requiring alterations in fermentation product metabolism for survival. In each case, this has resulted in substantial increases in the yield of hydrogen per glucose consumed (Kumar et al., 2001; Ito et al., 2004). In addition, overexpression of a native ferredoxin-dependent hydrogenase in Clostridium paraputrificum also resulted in a near doubling of fermentative hydrogen yield (Morimoto et al., 2005). However, progress in the field of metabolic pathway engineering needs to be made in order to develop optimized microorganisms producing high yields of hydrogen, at competitive rates and being able to utilize broader substrate ranges. In this respect, lab-scale hydrogen production will be soon scaled up and applied in real systems converting rich in carbohydrates feedstocks to hydrogen.