There are several approaches for optimization of thermophilic dark fermentation:
Promising substrates contain carbohydrates and should be achieved from sustainable sources. They should enclose enough energy for enabling fermentative conversion to hydrogen. Additionally, substrates should not require any pre-treatment or processing in order to keep down costs. Regarding hydrogen production as part of energy supply, waste materials should be considered as substrates. In dependence of used substrate, the supplementation of medium with nutritive substances can be optimized. The optimal medium should contain sufficiently the elements nitrogen, phosphor, sulfur and iron, as hydrogenases need these elements for hydrogen production. Substrates, such as starch, cellulose or hemicellullose require complex nitrogen sources in the medium, in order to cover nutritional requirements (Hawkes et al., 2002).
Another important parameter concerning thermophilic dark fermentation is the choice of microorganisms. Facultative anaerobes produce less hydrogen than stringent anaerobic bacteria. The latter are much more sensitive to oxygen and therefore more difficult to breed. It has been shown, that thermophilic-living microorganisms produce higher hydrogen yields than microorganisms, living in mesophilic areas (Nath et al., 2004). In order to maximize hydrogen yield, the metabolic pathways concerning hydrogen production should be optimized and the biomass production, which is evolutionary more important for bacteria, should be reduced. Additionally, bacterial metabolisms of alcohols and reduced acids should be switched to production of volatile fatty acids. Energy-efficient microorganisms can be produced by genetic modification, which can produce optimal hydrogen yields (Levin et al., 2004). For reduction of negatively acting methanogenic bacteria, there are different possibilities: Heat impact of the inoculum destroys methanogenic bacteria and the spore-forming hydrogen producer acquire an evolutionary advantage. Additionally, methanogenic bacteria can be limited by low pH-values and short retention periods, which lead to elution of slowly growing methane-producers.
For thermophilic hydrogen production, pH-values between 5.5 and 6.7 as well as retention periods of 8-12 h are optimal (Hawkes et al., 2002). Dilution rates, which are higher than 0.075 h^-1^, enhance hydrogen production. This corresponds to retention periods less than 13.33 hours (Chen et al., 2001). Short retention periods lead to possible reduction of reactor size and therefore cost reduction. Retention period is also substrate-specific, which means, that complex feedstocks (e.g. starch or industrial waste) require longer retention periods than e.g. glucose.
Concerning bioreactors, there is large optimization potential. On the one hand, systems with biomass retention and low hydraulic retention periods show the highest volumetric hydrogen production rates. On the other hand, systems with long hydraulic retention periods possess the highest yields. The optimal reactor for thermophilic hydrogen production should enable a complete conversion of used substrates (meaning high yields), as well as high hydrogen production rates (Schönherr, 2006).
Oxidation reaction of reduced ferredoxin among release of electrons in the form of molecular hydrogen is reversible and depends on hydrogen partial pressure pH2in the liquid phase. If hydrogen concentration in the fermentation broth increases, hydrogen synthesis will decrease. Metabolism of organisms switches to production of higher reduced products, such as lactate or ethanol (Hallenbeck et al., 2002). For continuous hydrogen synthesis, a partial pressure of < 50 kPa at 60 °C, of < 20 kPa at 70 °C and of < 2 kPa at 98 °C is necessary (Levin et al., 2004). Reduction of partial pressure can be achieved by stirring and mixing of the suspension.
An additional possibility is stripping of the fermenter content with inert gases such as argon, carbon dioxide or nitrogen (Kim et al., 2006), with nitrogen stripping, increased hydrogen yields of 66-68 % could be reached (Hussy et al, 2005; Mizuno et al., 2000). Decrease of hydrogen partial pressure with inert gas stripping causes contamination of product gas. Initial conditions for purification of hydrogen gas is caused, as fuel cells require cleanness of > 99 % (Levin et al., 2004). Extraction of hydrogen with membranes show an increase of hydrogen production rate of 10 % and of yield of 18 %. This technology does not lead to a dilution of the product gas (Liang et al., 2002). End product inhibition of hydrogen production constitute the major determining factor, which can be reduced by application of different technologies.
In order to ensure an optimal energy recovery, a second process step should be succeeded after thermophilic dark fermentation. There are 3 possibilities for this subsequent stage, which leads to either more hydrogen, or to methane or to electricity (Hawkes et al., 2007). For evaluation of energy production potential and of effluent purification by a combined dark fermentation-MBZ system, further research efforts are indispensable. For second fermentation stage, anaerobic digestion with aid of methanogenic bacteria can be adopted. This technology is already used in industrial scale, where after acidification step, methane is produced. Thereby, higher stabilities, capacities and efficiency are reached. Third possibility for increased energy yield out of substrates is photo fermentation after dark fermentation (Claassen et al., 2005). In first step, hydrogen and organic acids are produced from used biomass. This reaction is limited in yield with 4 mol hydrogen per mol hexose. During photo fermentation, the produced acids are further converted from phototrophic not-sulfuric purpurbacteria under action of light. After complete conversion in the 2 process steps dark-fermentation and photo-fermentation, 12 mol hydrogen of one mol hexose can be produced.