Among all immobilization methods, physical adsorption has been elected by most researchers due to its ease, absence of expensive and toxic chemicals, ability to retain the activity and feasibility of regeneration. On the other hand, poor adsorption of the enzyme results in its leaching off the support surface, which favors other means of enzyme immobilization such as covalent bonding, entrapment and encapsulation. It is possible to strengthen the attachment between the water-soluble enzyme and the water-insoluble surfaces by using multifunctional agents that are bifunctional in nature and have low molecular weight, such as glutaraldehyde (Shamel et al, 2005; Shamel et al., 2007). Nevertheless, physical adsorption remains the most attractive method industrially, because of its simplicity and economical effectiveness.

It has been shown that the adsorption of lipases from M. miehei on porous polysulfone surface (Shamel et al., 2005) and on modified regenerated cellulose hollow fiber membranes (Shamel et al., 2007) can be described by the Langmuir isotherm (Eq. [6.1]), which relates the amount of adsorbed lipase activity, aads, to that present in the supernatant solution, afree, at equilibrium.

A convenient way to express the temperature-dependent parameters Kads and aadsmax takes advantage of Van’t Hoff’s relationship between the equilibrium constant and the standard enthalpy change associated with the process under consideration:

Dhads

Kads = b eXp [ f

aads, max = a (1 + єТ). [6.3]

Experimental results showed that, unlike the general behavior of physical adsorption, increasing the temperature results in an increase in the equilibrium amount of enzyme adsorbed on both surfaces. This was a result of the increase in the diffusion of lipase into the micropores due to expansion of the pores and the reduction of the solution viscosity at higher temperatures.

The cellulose and hemicellulose part of lignocellulosic feedstock needs to be detached from the lignin part and this is possible either by physical and/or chemical and/or biological pre-treatment. In the physical pre-treatment technique no chemicals are involved. Physical pre-treatments can include: communition i. e. dry, wet and vibratory ball milling; irradiation i. e. electron beam irradiation, or microwave heating and also steam explosion (Wyman, 1996).

In the irradiation pre-treatment method electron beam irradiation or microwave assisted depolymerisation are used to separate cellulose or hemicelluloses from lignin. Electron beam irradiation has some effect on fatty and resin acids in the wood material and changes the physical properties of sawdust (Finell, Arshadi, Gref, Knolle, and Lestander, 2009). However, the irradiation pre-treatment method for ethanol production is not commercial.

In the physical steam explosion pre-treatment, the chipped lignocellulosic materials such as hardwood is treated by high pressure saturated steam (autohydrolyses) and then by reducing the pressure quickly, the material undergoes an explosive decompression. This results in hemicellulose degradation and some changes in polymeric lignin structure with the cellulose becoming more accessible for hydrolysis.

The addition of dilute acid in the steam explosion method (so-called acid catalysed steam explosion) may improve enzymatic hydrolysis of the cellulose and facilitate removal of the hemicellulose. Steam explosion requires less energy than mechanical pre-treatment methods (communition) and is the most effective pre-treatment method for hardwoods and agricultural lignocellulosic products.

One disadvantage of the steam explosion pre-treatment method is the formation of some inhibitory compounds for enzymatic hydrolysis in the next step in ethanol production (Sun and Cheng, 2002).

There are other physical-chemical pre-treatment methods such as ammonia fibre explosion (AFEX) where lignocellulosic feedstock is exposed to liquid ammonia at high temperature and pressure over a period of time and then the pressure suddenly lowered. This method can be used for many different materials including corn stover, wheat straw, softwood newspaper, switch grass, alfalfa, etc. In contrast to acid catalysis steam explosion, the AFEX pre-treatment does not significantly solubilise hemicellulose and high hydrolysis of hemicellulose and cellulose has been obtained after this pre-treatment method. But the superheated ammonia vapour must be recovered to protect the environment (Sun and Cheng, 2002).

In the chemical pre-treatment method some chemical/s are added to the feedstock, e. g. concentrated acids, dilute acids, alkaline solutions. It is possible to use concentrated acids such as H2SO4 and HCl for pre-treatment (acid hydrolysis) of lignocellulosic feedstock. The acids effectively hydrolyse the cellulose. However the concentrated acids are toxic, corrosive and must be recovered after the process. Therefore, dilute acid hydrolysis (e. g. sulphuric acid, hydrochloric acid) as a pre-treatment has been used instead in many applications (softwoods, hardwoods, agricultural residues) with a high reaction rate and effective cellulose hydrolysis. But a neutral pH is necessary for enzymatic hydrolysis or fermentation. One advantage of dilute acid hydrolysis as a pre-treatment method in comparison to the steam explosion method is that the xyloses in hemicellulose remain intact with high yields (Wyman, 1996). These xylans can be utilised in value-added products.

Alkaline pre-treatment with sodium hydroxide alone or in combination with other chemicals like peroxide are most effective for agricultural residues rather than wood feedstock. By this method the lignin is effectively removed and some of the hemicellulose solubilised as well (Wyman, 1996). In the biological pre­treatment technique, microorganisms degrade the lignin (lignin solubilising microorganism) by producing lignin-degrading enzymes and no chemicals are needed. The method is slow which makes it less economical and sometimes consumes hemicellulose as well but it does not require a high energy input and only needs mild environmental conditions (Wyman, 1996).

Recently, Lignol Innovations Corporation has developed a method based on an ethanol-based organosolv pretreatment (i. e. delignification by extraction of lignin from the lignocellulosic biomass with organic solvents or their aqueous solutions) to separate lignin, hemicellulose components (e. g. xylose) and extractives from the cellulose part of the woody biomass (Arato, Kendall, and Gjennestad, 2005). In a review article the prospects and evaluation of different organosolv methods and mechanisms have been presented recently (Zhao, Cheng, and Liu, 2009)

In general, an optimal pre-treatment stage should improve the enzyme/ hydrolysis accessibility, should avoid the degradation of carbohydrates and should avoid the formation of by-products which may have inhibitory effects on the hydrolysis and fermentation steps. In another words, any pre-treatment method must be tailored to the specific lignocellulosic material with different chemical and structural compositions. The economic aspects are also very important in large scale industrial bio-ethanol production.

Mathematical models have been developed to improve understanding of the complex dynamics of the anaerobic digestion process and to predict the response of the anaerobic systems to changes in operating conditions (hydraulic retention time, organic load, temperature, etc.). Models are tools for process design, control strategies, diagnosis or prediction of system performance under conditions of increasing or decreasing load and variation of feeding characteristics.

There are many types of anaerobic models ranging from steady-state models to single — or double — or multi-step dynamic models. Steady-state models can be applied in systems where the fluctuations in the feed characteristics and organic loading are minimised. This basis of static design modelling has been employed in several text books (Tchobanoglous and Burton, 1991). In most cases, however, the model should provide information about the dynamics of the system towards changes in the input of the system. Dynamic models can be utilised successfully in control schemes or for simulation purposes. Depending on the purpose, the model should be simple enough including only the basic steps for describing the dynamics of the core process (control) or more complex including as many steps as possible making it widely applicable (simulation). Table 12.1 refers to various models developed in the last three decades.

The basis for simplifying a model is the ‘rate limiting step’ concept, that is, the last slow step in a sequence of reactions that determines the overall rate of a multistep process. The two slowest steps recognised in anaerobic systems are hydrolysis and acetoclastic methanogenesis (Gossett and Belser, 1982; Pavlostathis and Gossett, 1986, 1988). When the feedstock contains particulate organic matter (sludge, organic fraction of municipal solid wastes, solid residues, etc.), the rate of hydrolysis usually determines the overall rate. In this case, the steps that follow are usually considered to be at pseudo steady state and can be described by algebraic equations reducing the degree of complexity of the model. In the absence of particulate matter in the feedstock, acetoclastic methanogenesis is

Table 12.1 Steps involved in various models of anaerobic digestion developed in the last three decades Hydrolysis Acidogenesis Acetogenesis Methanogenesis Source Volatile fatty acids Graef and Andrews (acetate) -»■ CH4, C02 (1974) Particulate organics -»■ soluble organics Soluble organics -»■ volatile fatty acids Volatile fatty acids Hill and Barth (1977) (glucose) Glucose -»■ butyrate, propionate, Butyrate, propionate (acetate) -»■ CH4, C02 Acetate -> CH4 Hill (1982) acetate -»■ acetate H2, co2 CH4 H2, C02 -»■ acetate Particulate organics -»■ aminoacids, sugars, Aminoacids, sugars, fatty Propionate -»■ acetate Acetate -> CH4 Bryers (1985) fatty acids acids -»■ propionate, acetate H2, co2 CH4 Glucose -»■ butyrate, propionate, Butyrate, propionate Acetate -> CH4 Mosey (1983) acetate -»■ acetate H2, C02 CH4 Pullammanappallil etal. (1991) Particulate organics (fats, carbohydrates, Soluble organics -»■ acetate Acetate -» CH4 Kleinstreuer and proteins) -»■ soluble organics Soluble organics (glucose) -»■ volatile fatty acids (acetate) Acetate -> CH4 Powegha (1982) Moletta et al. (1986) Easily biodegradable biomass -»■ soluble Soluble organics -»■ volatile fatty acids Volatile fatty Smith et al. (1988) organics Glucose -»■ lactate, butyrate, Butyrate, propionate acids -> CH4 Acetate CH4 Costello et al. (1991) propionate, acetate Lactate -»■ propionate, acetate -»■ acetate H2, co2 CH4 Particulate carbohydrates -»■ soluble Soluble carbohydrates -»■ butyrate, Butyrate, propionate Acetate -> CH4 Angelidaki etal. carbohydrates propionate, acetate -»■ acetate (1993) Particulate carbohydrates, proteins, fats -»■ Aminoacids and sugars -»■ Propionate -»■ acetate Acetate CH4 Siegrist et al. (2002) aminoacids, sugars, fatty acids propionate, acetate fatty acids -»■ acetate H2, co2 CH4 Particulate carbohydrates, proteins -»■ Soluble carbohydrates and proteins, Propionate -»■ acetate Acetate CH4 Gavala et al. (1996) soluble carbohydrates and proteins other organics -»■ propionate, acetate Particulate organics^carbohydrates, Aminoacids and sugars -»■ butyrate, Butyrate, propionate Acetate CH4 Batstone etal. (2002) proteins, fats -»■ aminoacids, sugars, fatty propionate, acetate -»■ acetate H2, C02 CH4 acids Fatty acids -»■ acetate

the rate limiting step, considering the preceding steps to be at a pseudo steady state.

On the other hand, in the case of multistep models, the steps usually included are:

• Hydrolysis of particulate matter: Although the mechanisms of the individual hydrolysis steps are known, the hydrolysis step is usually lumped as a single first order process (Eastman and Ferguson, 1981; Pavlostathis and Giraldo — Gomez, 1991).

• Acidogenesis of soluble organic matter: Modelling of sugar fermentation is challenging due to the variety of the possible fermentation products and the determination of the stoichiometry (subjected to the regulation mechanisms prevailed in the heterogeneous group of acidogens). The main pathways acknowledged to take place are towards formation of butyrate, acetate, ethanol and acetate, as well as propionate and acetate as end products (Ren et al, 1997; Batstone et al., 2002). Lactate has been also considered important to be in­cluded among the sugar fermentation products (Costello et al., 1991). In mixed fermentation processes, the mechanisms that regulate the composition of the fermentation product mixture have not been elucidated completely and as a result, modelling of this step has not yet been effective (Mosey, 1983; Costello et al, 1991; Ruzicka, 1996). This limitation has become critical due to the increasing interest concerning the production of biohydrogen produced along with the other sugar fermentation metabolic products. As far as the modelling of amino acid fermentation is concerned, the pathways based on Stickland reactions have been proposed (Ramsay and Pullammanappallil, 2001).

• Acetogenesis and methanogensis: Both steps have been extensively and successfully simulated. However, the incorporation of hydrogen, free ammonia and pH effects on the kinetics of both steps can be further improved.

In the biochemical part of the model, the kinetic relationships expressing the bioreaction rates are very important. There is a wide range of kinetics that can be applied in each step of the anaerobic digestion (Pavlostathis and Giraldo-Gomez, 1991), but the most common relationship is the Monod kinetics:

where p is the consumption rate of the substrate, km is the maximum specific consumption rate constant, KS is the saturation constant, S is the concentration of the substrate and X is the concentration of the microorganisms that consume the substrate.

Equation 12.1 can be extended to include any inhibition or regulation mechanisms if required (Batstone, 2006):

where Ij, I2, . . In are functions expressing inhibition mechanisms can include classic non-competitive or competitive inhibition, or empirical formulas. Modification of Monod kinetics to account for all kinds of product, cell and substrate inhibition has been extensively applied in biochemical engineering (Levenspiel, 1980; Han and Levenspiel, 1988).

Moreover, apart from the biochemical part of the model, it is important to include a physicochemical part to assess the gas transfer and calculate the pH (if required in the biochemical part). The gas transfer can be modelled by applying the gas-liquid transfer theory for each gas. Equilibrium can also be assumed for those gases that are practically insoluble in water, such as hydrogen and methane. The total gas production rate can be calculated as the sum of individual gas production rates. Gas flow can also be derived by setting a pressure difference between the headspace and the atmosphere (Batstone, 2006). pH calculation requires solving algebraic equations derived from the equilibrium of weak acids and bases as well as charge balance. Dissociation of acids and bases can also be considered as dynamic processes evolving at a high rate.

In 2002, a group of scientists expert on anaerobic digestion modelling constructed the anaerobic digestion model (ADM1) to be a frame model basis for several applications in anaerobic digestion (Batstone et al., 2002). The model has been used as a reference basis for many extensions made by several researchers afterwards to utilise it in specific applications, such as, the anaerobic digestion of brewery wastewater in a full scale high rate system (Ramsay and Pullammanappallili, 2005).

Depending on the bioreactor design (homogeneous or heterogeneous system), simple hydraulic or more complex models taking into account mass transfer phenomena can be developed. Mass transfer is important in the case of ‘biofilm’ bioreactors where microorganisms are attached on the surface of an inert material (anaerobic filters) or attached on each other (UASB). There are different degrees of complexity that can be entailed in modelling biofilm bioreactors. Several parts of the bioreactor can be considered to be homogeneous, as in UASB reactors modelled by Bolle et al. (1986), thus a non homogeneous system can be depicted by a combination of the homogeneous systems connected. In a more complex model design, the layers composing the biofilm in a filter or the granule in a UASB are taken into account, with each layer being formed by a specific group of microorganisms. Many UASB models assume that the granules are spherical and the relative concentration of the acidogens and methanogens remain constant in the granule. The density of the granules is also assumed to remain constant. Saravanan and Sreekrishnan (2006) review the various model approaches available for biofilm reactors extensively.

The current costs of providing biomass in Europe vary greatly depending on the biomass and range from €21 to €180 per tonne of dry matter (DENA, 2006). The variation is due to the different energy and moisture content as well as the origin of the feedstocks. Wood chips are at the upper end of the price range while waste wood and agricultural residues are at the lower end; the average feedstock costs are < €60 per tonne of dry matter. These costs include feedstock storage close to the field or forest (10 km) but not the transport costs to the processing plant. The delivery costs increase with the moisture content and transportation distance.

Table 3.5 shows a breakdown of the total costs of the feedstock for the example of switchgrass in the US (US EPA, 2009). As shown, just over half of the costs

(57%) are related to the cultivation of the feedstock and the rest are due to the storage and delivery to the ethanol plant. These costs compare well with forest residue costs (US EPA, 2009).

The majority of oils and fats after extraction (pressing, solvent extraction, combination pressing-extraction, rendering, etc.) are not suitable for the production of biodiesel, especially in large continuous plants using the traditional alkaline transesterification process. Undesirable products are non-triaclyglycerol compounds like free fatty acids (FFAs), phospholipids, oxidation products, metals, protein and carbohydrate residues, waxes, moisture and inorganic matter. Two refining routes are used (chemical and physical refining) which refer to the methodology for FFA removal.

In both refining procedures the first step is a degumming process in which the hydratable phospholipids are removed by water washing and the non-hydratable ones are discarded by treatment with citric or phosphoric acid. Enzymatic degumming becomes more attractive due to the lower losses.

In the chemical process, FFAs are removed as soaps (soapstock) by neutralization with NaOH. In the physical refining, FFAs are discarded by stripping (deodorization). Chemical refining is using a lot of water and the soapstock has to be acidified (production of acid oils) and is environmentally not recommended. In addition, there are considerable losses of lipids during the separation of the soap layer. As 85% of the production costs of biodiesel is based on the feedstock, losses should be kept to a minimum.

A bleaching step (adsorption with activated clay or silica) can be necessary for highly colored oils or high contaminated oils (removal of Ca, Fe, Cu, traces of soaps and phospholipids).

Deodorization, the last step of refining, is performed at 210-260°C, with one to two per cent steam (1 Mbar) for the removal of the FFAs and oxidation products (physical refining). In addition pigments and unwanted contaminants (PAHs are pesticide residues) are degraded. Deodorization can be omitted on condition that FFA content is very low which can be case for freshly extracted soy and rapeseed oils. A detailed overview of refining oils and fats is given by O’Brien et al. (2000).

A feedstock purification technology which may replace the conventional degumming process is Ambersep™ B19 (amberlyst resin) developed by Daw. The product removes proteins and polysaccharides, traces of phospholipids and soaps. By doing a purification step first, the lifetime of Ambersep™ B20 is extended during the esterification of FFAs in crude feedstocks.

Ecodiesel® is a biofuel incorporating glycerol, produced by enzymatic technology and patented by the University of Cordoba (UCO).37 It is composed of two moles of FAEEs and a mole of MG. Particularly, Ecodiesel® is obtained using pig pancreatic lipase (PPL), in both free and immobilized form, to achieve the 1,3 selective transesterification of TGs to produce the corresponding 2-monoacyl derivatives of glycerol (MG) and two moles of FAEEs. Ethanol is the alcohol employed in the process (Fig. 7.2).

2 RCOOCH2CH3

Fatty acid ethyl ESTERs (FAEEs)

for Ecodiesel®

It is interesting to note that the enzymatic transesterification process can also be carried out with different short-chain alcohol (ethanol, 1- and 2-propanol, 1- and 2-butanol, etc.) and their mixtures, and it is not, in principle, restricted to the use of methanol, as it is normally under conventional chemical reactions (with acidic or basic catalysis).

Many reports on biodiesel preparation using free38 or immobilized lipases can also be found.1117 In particular, PPL has been widely employed in the last decades for the resolution of mixtures of chiral enantiomers, either by enantioselective hydrolysis39,40 or by alcoholysis or transesterification.41

The recent work of Luna et al. and their patents37,42 show the entrapment of the PPL in demineralized sepiolite and its activity in the alcoholysis reaction of TGs contained in sunflower oil. Demineralized sepiolite is a clay mineral (a complex magnesium silicate) with a microporous structure and a channel dimension of

11.5 x 5.3 A. Its structure moves along fibres that confer a high specific surface area to the solid, similar to that of the AlPO-5.43,44 The extraction of the ions (Mg2+, Al+3, etc.) by acid treatment significantly increases the size of the pores, making them comparable to those of amorphous silica45 or even to a mesoporous structure similar to MCM-41.46 These voluminous pores are able to trap some macromolecules including various enzymes.47,48

Results obtained by employing immobilized PPL compared to the free enzyme are reported in Table 7.2. Different temperatures, oil/alcohol ratios and oil/ immobilized PPL ratios have been also investigated and included in Table 7.3.

Note: Data corresponds to the number of reuses (no.) of the biocatalyst, as a continuation of Table 7.2, under different reaction conditions.

The efficiency of the PPL can be obtained by comparing the turn-over frequency (TOF) values of free and immobilized PPL (Table 7.2), both obtained under the same experimental conditions and temperature. The efficiency of PPL was reduced to 42.5% [(24.5/57.7) x 100 = 42.5] after immobilization, due to a potential steric effect of the immobilized enzyme in the reaction and/or to the deactivation of the active sites of the enzyme in the entrapment process.

The TOF values showed that a decrease in the oil/alcohol molar ratio from 1:10 (Table 7.2) to 1:2 (Table 7.3) leads to an increase in the efficiency of the immobilized enzyme, in good agreement with the results obtained for the free enzyme. The results also pointed out that in any case, even with an excess of ethanol, a maximum 66% yield could be obtained, corresponding to a 1,3 selective enzymatic process. Of note was the enzyme stability and recyclability. Although the efficiency was reduced compared to the free form, the immobilization through physical entrapment of the PPL guaranteed the lifespan of the lipases. The free PPL was found to be completely deactivated in 48 hours, whereas the immobilized enzyme was active for several weeks, even after successive reuses preserving over 90% of the initial activity.

Another important advantage of the enzymatic process is the possibility of using various alcohols apart from methanol or ethanol. The effect of different short-chain alcohols on composition, yield, conversion and TOF of Ecodiesel-100, obtained in the alcoholysis of pure and waste frying sunflower oil, is reported in Table 7.4.

The biofuels could smoothly be obtained using the various alcohols employed, obtaining quantitative TGs conversions and selectivity to FAEE higher than 50% in most of the cases. The reaction typically takes 8-12 hours to complete, and the selectivity to FAEE increases with the time of reaction as expected.

A potentially useful biofuel blend of FAEE, MG and traces of DG, in varying proportions (depending on the conversions), can be obtained. The FAEE/MG ratio was around 2:1 molar at quantitative triglyceride conversion.

In conclusion, the alcoholysis of TGs with short-chain alcohols using 1,3-regiospecific lipases can play an advantageous role, compared to the conventional base-catalyzed process, to obtain new biofuels incorporating glycerine and to minimize the waste production by improving the reaction conversion under greener conditions. Milder reaction conditions were employed and a cleaner biofuel (Ecodiesel-100) was obtained. The efficiency of PPL was remarkably increased at a higher pH in contrast with the reported results describing a poor activity of the enzymes at that pH. The immobilized PPL was highly stable, although the efficiency was reduced (42%) compared to the free enzyme. The catalyst can easily be recycled (11 times), almost preserving the initial catalytic activity.

10.4.1 Modeling

First attempts to model the ABE fermentation were already undertaken in the 1980s (Voturba et al., 1986) based on mass balances for the substrate, biomass, key intermediates and products of C. acetobutylicum batch cultures. Better models were proposed when various on-line measurements (Chauvatcharin et al., 1998; Junne et al., 2008) and the genome sequences of C. acetobutylicum (Nolling et al., 2001) and later on C. beijerinckii (JGI, 2005) became available. Meanwhile, the study of transcriptome (Alsaker and Papoutsakis, 2005; Alsaker et al, 2004; Jones et al., 2008; Shi and Blashek, 2008; Tomas et al, 2003a, 2003b; Tomas et al., 2004; Tummala, Junne, Paredes, et al., 2003), proteome (Schwarz et al, 2007; Sullivan and Bennet, 2006) and metabolome (Shinto et al., 2007) of different solventogenic clostridia leads to more complex and comprehensive models (Junne et al., 2008; Senger and Papoutsakis, 2008a, 2008b; Shinto et al., 2007), which supported further understanding of the clostridial metabolism and allowed predictions of metabolic fluxes and end-products for several scenarios. However, the biphasic nature of the metabolism and the complex regulatory networks (see Section 10.2) still cause some problems that can result in incorrect outputs. The assumed reversibility of enzymatic activities lacks experimental evidence in several cases. Also, the obvious pH influence on the shift to solventogenesis has been neglected in some models. A transnational research network is currently focusing on elucidating systems biology of solventogenesis in C. acetobutylicum (project COSMIC within the SysMO program, www. sysmo. net).

Photoheterotrophic or photo-fermentative hydrogen production refers to the microbial process, during which organic substrates are oxidized under anaerobic conditions in the presence of light, generating hydrogen and carbon dioxide. Photo-fermentative hydrogen production is generally carried out by prokaryotic microorganisms called purple non-sulfur bacteria (PNS) (Basak and Das, 2007), although lately the process has also been reported to be carried out by eukaryotic microorganisms, that is, green algae (Hemschemeier and Happe, 2005). Photosynthetic microorganisms convert light energy into chemical energy in the form of chemical bonds, via the pathway of photosynthesis.

Contrary to dark fermentation (see Section 13.3), in which the enzyme hydrogenase catalyses hydrogen production, nitrogenase is the key enzyme for the photo-fermentative process of PNS. Under nitrogen-deficient conditions, nitrogenases can also catalyse the generation of molecular hydrogen using light energy and reduced compounds (such as organic acids) as the electron donors, where ferredoxin acts as the electron carrier (Das and Veziroglu, 2001). Light as an energy source is necessary for such reactions to take place, since their Gibbs energy is positive and thus they are not thermodynamically favored:

CH3COOH + 2H2O + “hv” ^2CO2 + 4H2 DGo= +75.2 kJ/mol [13.6]

As shown in Table 13.3, in most photo-fermentative biohydrogen studies pure cultures of the genera Rhodopseudomonas, Rhodobacter and Rhodospirillum have been used, whereas studies with other genera such as Rubrivivax (Li and Fang, 2008) and Rhodobium (Kawaguchi et al., 2001), as well as with mixed cultures (Zhang et al., 2002; Fang et al, 2005) have also been reported. Malate and glutamate were commonly selected as carbon and nitrogen sources, respectively. However, the use of other carbon sources such as the acids lactic, succinic, acetic, propionic and butyric, or their salts, has also been investigated for their potential to be converted into hydrogen either in the form of synthetic substrates or as parts of actual waste streams.

In order to evaluate the performance of a photo-fermentative hydrogen production system, the efficiency with which light energy is converted to energy in the form of hydrogen, the so-called photochemical efficiency (PE) or light conversion efficiency, has to be taken into consideration (Akkerman et al., 2002). It thus becomes obvious that the efficient utilization of light energy, provided either by a physical (sunlight) or an artificial source, is of extreme importance for the construction of a feasible energy production system (Miyake and Kawamura, 1987). Factors affecting PE include wavelength and intensity of light, cell concentration in the culture, surface to volume ratio of the culture (reactor geometry) and light penetration in the reactor. It is widely accepted that optimal light utilization is indispensable for maximal hydrogen production.

As shown in Table 13.3, in most studies one or more artificial light sources have been selected among florescent lamps, halogen lamps, optical fibers, neon tubes, light-emitting diodes and photosynthetically active radiations (PARs), which however could become a hindrance to the overall economic viability of a full-scale application. Sunlight on the other hand, is a free and abundant light source, which

Microorganism Substrate Reactor operation/ configuration Nitrogen source Condition of microorganisms Light source/lenergy Reference Rhodopseudomonas palustris Malic acid, acetic acid Fed-batch/ cylindrical glass Glutamic acid Suspended Lamps light source from one and two sides/460 |iE m-2 s-1 Carlozzi and Lambardi, 2007 Palm oil mill effluent (POME) Batch/serum bottles No addition, TKN ~60 mg I-1 Suspended Tungsten light bulbs/2-5 klux Jamil ef a/., 2009 Glycerol Batch/serum bottles Glutamate, 2-6 mM Ammonium, 0-4 mM Suspended Panel of 50 W halogen spotlights/- Sabourin- Provost and Hallenbeck, 2009 Glucose Continuous/flat, polymethyl methacrylate Ammonium Immobilized in (PVA)-boric acid gel LED mounted on the topside/3-11 klux Tian ef a/., 2009 Rhodopseudomonas faecalis Sodium Acetate Fed-batch/serum bottles Sodium glutamate, 10 mM Suspended 60 W incandescent lamps/4 klux Ren ef a/., 2009a Glucose Batch/serum bottles Sodium glutamate, 1 g I-1 Immobilized in agar gel 60 W incandescent lamps/4 klux Ding ef a/., 2009 Sodium acetate Batch/serum bottles Glutamate, 10 mM Suspended 60 W incandescent lamps/4 klux Ren ef a/., 2009b Rhodobacter capsulatus Acidified Miscanthus hydrolysate, acetate, lactate, fructose Batch, sealed glass bottles Sodium glutamate, 2 mM Suspended 150 W halogen lamp/1,370 [imol photons/m2/s Uyar ef a/., 2009 Lactate Batch/flat glass Sodium glutamate, 7 mM Suspended Sodium-vapour lamp 600 W/ Obeid ef a/., 2009 Malate Batch Sodium glutamate, 2 mM Suspended lamps/4 klux Ozturk ef a/., 2006 IContinued)

Microorganism Substrate Reactor operation/ configuration Nitrogen source Condition of microorganisms Light source/lenergy Reference Rhodobacter sphaeroides DL — malate Batch/triple jacketed vertical, cylindrical, glass Glutamate, 2 mM Suspended Tungsten filament lamp placed in central axis of reactor/15± 1.1 W m-2 Basak and Das, 2009 Mixed substrate of acetate, butyrate, ethanol Batch/water jacket glass column Sodium glutamate, 10 mM Suspended 100 W lamps 5.5±0.5 klux Nath ef a/., 2008 Raw, acidified or clay pretreated olive mill wastewater Batch/glass vessels No addition Suspended 150 W tungsten lamp/4 klux Eroglu efa/., 2006 Succinate Batch Ammonium chloride 0.04 w/v Suspended Lamps/2.4 klux Chalam ef a/., 1996 Rhodospirillum rubrum Sodium succinate Batch Glutamate 3 mM Suspended Fluorescent and incandescent light bulbs/60 W m-2 Melnicki efa/., 2008 Acidified cassava wastewater Batch/serum bottles Glutamic acid, ammonium nitrate Suspended Fluorescence lamp/6000 candela/m2 Reungsang ef a/., 2007 Mixed substrate of acetate, malate Batch/cylindrical, glass L-glutamate 23 mM Immobilized in agar gel Lamps/20 klux Planchard ef a/., 1989 Lactate, cheese whey Continuous, HRT 74 h rectangular L-glutamate 15 mM Suspended 100-W spot-light tungsten Zurrer and Bachofen, 1979

can be used for direct irradiation of the bioreactor or amplified by the use of solar — energy-excited optical fibers (Chen et al., 2008a). A drawback of using sunlight could be the periodicity of the light source; an obstacle that could be surpassed by the addition of solar-energy-excited optical fibers, accompanied by light-dependent resistors, which can ensure the stability of light energy (Chen et al., 2008a).

In order to develop commercially viable processes, the influence of many other factors has to be taken into consideration. The nitrogen source is one of the most critical parameters for effective photo-fermentative hydrogen production. An organic nitrogen source, such as glutamic acid or inorganic salts or more complex organic nitrogen sources such as yeast extract, seems to be necessary for efficient hydrogen production regardless of the species of microorganism used. The effect of the type and concentration of the carbon source used as substrate (Carlozzi and Lambardi, 2009), the C/N and C/N/P ratios (Reungsang et al., 2007) as well as the physicochemical conditions of growth such as pH (Tian et al., 2009) and temperature (He et al., 2006) have been widely studied and optimized, since they seem to have a severe effect on both the final hydrogen yield and the hydrogen production rate. Regarding pH, the optimum value is reported to be 7 in most cases, whereas the optimum temperature is reported to be 30°C. A general conclusion from all these studies is that the photo-fermentation processes seem to be favored by a high ratio of C/N, irradiation with light of saturating intensity, under anaerobic conditions with optimal temperature and pH, depending on the specific microorganism used.

There are three major types of photo-bioreactors developed for hydrogen production that is tubular, flat panel and bubble column reactors. The features of these photo-bioreactors have been reviewed by Akkerman et al. (2002) and the importance of PE in hydrogen production was strongly emphasized. The main advantage of tubular and column photo-bioreactors is that their geometry allows for quite efficient mixing of the culture, and thus the exposure of the microbial cell to light is more equally distributed. The way to scale-up is to connect a number of tubes via manifolds. Flat panel reactors consist of a rectangular transparent box with a depth of only 1-5 cm. The height and width can be varied to some extent, but in practice only panels with a height and width both smaller than 1 m have been studied. The advantage of these systems is the large surface that can be illuminated either using sunlight or artificial means. The main disadvantage of such type of reactors is the high consumption of energy used for maintaining efficient air supply and mixing of the liquid. Many scaled-up versions of photo-bioreactors consist of repeating many of the smaller photo-bioreactor units, with its practical implications.

Finally, the quantitative description of photo-fermentative hydrogen production seems to be quite complex, due to the large number of parameters that have to be taken into account. Simple models such as the Luedeking-Piret model (Basak and Das, 2009), the logistic model (He et al, 2009), the Monod equations (Obeid et al., 2009) and the Gompertz equation (Nath et al., 2008) have been used in order to fit experimental results regarding biomass growth and cumulative hydrogen generation, but so far very few studies have dealt with the development of complex structured kinetic models, properly incorporating specialized for photo-fermentative hydrogen production parameters such as light intensity and wavelength influence. A simple kinetic model for photo-fermentative biohydrogen production has been developed by Gadhamshetty et al. (2008) for batch bioreactors, where it was assumed that sufficient light intensity and optimal C/N ratio were available under stressful nitrogen concentrations. The proposed model used Rhodobacter sphaeroides as the model biomass and contained 17 parameters to describe cell growth, substrate consumption, and hydrogen evolution as well as inhibition of the process by biomass, light intensity, and substrate. Based on sensitivity analysis performed with the validated model, only 6 of the 17 parameters were found to be significant and it was indicated that the range of optimal light intensity for maximum hydrogen yield from malate by R. sphaeroides was 150-250 W/m2.

Chemical conversion involves a number of widely known and extensively employed processes since the nineteenth century. In fact, the chemical process currently in use for the preparation of biodiesel from biomass (transesterification of oils) is the same as has been used for many years. Feedstocks utilised for the preparation of biofuels are also very similar, with peanut, hemp, corn oil and animal tallow been partially replaced by soybean, rapeseed, recycled oil, forest wastes, trees and sugar cane.

First-generation biodiesel is currently the most common example of a biofuel prepared by chemical conversion. It is currently the most widely developed biofuel in Europe. In 2007, 19 biodiesel plants in the new EU member states were starting operations or were under construction/planning. Relatively large plants (with capacities of 100 000 tonnes/year) can be found in Lithuania, Poland and Romania.

The conventional methodology for the production of biodiesel involves the transesterification of triglycerides (TG) from vegetable oils (palm, corn, soybean, rapeseed, sunflower, etc.) with short-chain alcohols, including methanol and ethanol, to yield fatty acid (m)ethyl esters (FAM/EE) and glycerol as by-products (Scheme 1.1).

However, non-edible feedstocks, including Jatropha, Brasicca species and microalgae oil, are becoming increasingly important nowadays for the production of biodiesel and are considered to be an important asset for future biodiesel production. The methods of biodiesel preparation can be classified into three

types: chemical catalytic (base or acid catalysis: homogeneous and/or heterogeneous), biocatalytic (enzyme catalysis: homogeneous and/or heterogeneous) and non­catalytic processes. Several reviews on the preparation of biodiesel from different feedstocks utilising various technologies can be found in the literature.8-12

The production of related biofuels via chemical processes (i. e. (trans)- esterifications) has also been reported. These biofuels have been specifically developed in research institutions, and commercial processes for their implementation as transport fuels are still under development (see Chapter 7 for more details). For more specific details, the readers are referred to Part II of the book (Chapters 5 and 6 as well as some related content in Chapter 22), in which more detailed information about processes, technologies and biofuels produced will be given.

In temperate areas, annual oilseeds such as soybean, canola and sunflower have been largely used as biodiesel feedstocks, while palm oil trees have been used as feedstock in the tropics. However, the use of non-edible, low-input, low-cost and sustainable vegetable feedstocks compatible with good quality biodiesel (to achieve both customer and vehicles manufacture trust) should be the scientific community target. According to the previous requirements, the following section presents the most suitable vegetable raw materials for biodiesel production. The selection has been prepared considering low input and most promising crops according to their fuel properties (Dorado, 2008). Oleaginous crops to produce biodiesel, such as Bahapilu, castor, cotton seed, cuphea, Jatropha curcas, karanja seed, linseed, mahua, nagchampa, neem, rubber seed, tonka bean; low-cost edible oils, such as cardoon, Ethiopian mustard, Gold-of-pleasure, tigernut; and potential oil-bearing crops and trees such as allanblackia, bitter almond, chaulmoogra, papaya, sal, tung and ucuuba have already been revised by the authors and an extensive revision can be found in a previous work (Dorado, 2008).