Although the available body of information on production from hydrates is still limited, there is sufficient information to begin identifying particular features, prop­erties, conditions, and production methods that are linked to a higher gas production potential and increase the desirability of hydrate deposits, and to use this informa­tion to develop a set of guidelines for the selection of promising production targets.

5.1 Desirable Features and Conditions

These include the following [125]:

• Large formation k and j, which are almost invariably associated with sandy and gravely formations that are characterized by low P, S and S, leading to

cap ’ wir’ gir

relatively high permeability to gas and aqueous flow.

• Medium SH (i. e.,30% < SH < 60%) corresponds to optimal QP in terms of magnitude and length of time to attain it. The effect of SH on production is not monotonic, but a complex function of SH and the timeframe of observation. A lower SH has the advantage of higher keff, leading to an earlier evolution of gas and a larger initial QP [131,132]. The disadvantages of a lower SH may include a larger water production and a lower total gas production because of early exhaustion of the resource. A high SH leads to slower evolution of gas and lower initial QP, but a higher maximum QP and total production.

• The most desirable targets can be easily identified from the inspection of the phase diagram (Fig. 14). The larger T provides a larger source of sensible heat to support the endothermic dissociation, and a larger initial P allows a larger pres­sure drop, leading to larger production rates. Thus, (a) hydrates that occur along the equilibrium line are very desirable, and (b) the desirability increases with an increasing equilibrium P (and, consequently, T). The production potential decreases as the stability of the hydrate deposit at its initial conditions increases (as quantified by the pressure differential DP = P-Pe at the prevailing reservoir T). In practical terms: we target the deepest, warmest reservoirs that are as close as possible to equilibrium conditions. In addition, the deeper reservoirs have larger overburdens and are therefore less prone to adverse geomechanical impacts.

• For reservoirs with the same hydraulic properties, SH, and P: the warmest possi­ble deposit is the most desirable. For reservoirs with the same hydraulic proper­ties, SH, and T: the reservoir with the lowest possible P is the most desirable.

• In terms of deposit classes: All other conditions and properties being equal, Class 1 deposits appear to be the most promising targets for gas production because of the thermodynamic proximity to the hydration equilibrium. Additionally, the existence of a free gas zone guarantees gas production even when the hydrate contribution is small.

• Class 2 and Class 3: Class 2 deposits can attain high production rates, but are also burdened by longer lead times of very little gas production; Class 3 deposits may yield gas earlier and can attain significant production rates, but there are indica­tions that these are lower than in Class 2. The relative merits of these two types will likely be determined by site-specific conditions.

• All classes: The difficulties of site access notwithstanding deeper and warmer oceanic accumulations appear to be more productive than permafrost ones they can have (a) a higher T (14°C is the maximum equilibrium temperature observed in permafrost-associated deposits) and a larger sensible heat available for disso­ciation and (b) a higher P, increasing the depressurization effectiveness, in addi­tion to (c) the beneficial dissociating effect of salt.

• All classes: The importance of impermeable or near-impermeable upper bound­aries cannot be overemphasized.

• In terms of production method: Depressurization appears to have a clear advantage in all three classes.

Similar to water resources, plants grown as open systems require significant fertil­izer loads to optimize biomass yield (Table 2). Brazilian sugarcane utilizes as little as 20% of applied nitrogen [34]; the remainder is largely lost to water supplies with the attendant negative impact on the surrounding environment [14]. The require­ment for nutrient complexity within soil often requires crop rotation and precludes year-over-year high biomass productivities available with closed-system bioreactors.

To the extent that growth can be separated from production, rapid CO2 fi xation is possible in an Electrofuels approach with low nutrient requirements, since relatively little biomass is produced per volume of fuel. These requirements are further dimin­ished through nutrient recycling within the biorefinery. At least some other advanced biofuels systems also address this issue; notable examples include slow growing woody biomass (with no external nutrient requirements) and other closed system photobioreactors [49, 60].

We are optimizing the chemical properties of soluble Mo-polypyridine electrocata­lysts for water-to-hydrogen conversion to increase turnover frequency and stability in microbial growth media. We are also seeking improvements to minimize electro­chemical overpotential for H2 production. In particular, we are targeting the incor­poration of electron-withdrawing groups at the axial and equatorial pyridine donors as well as methine bridges. Additionally, we are developing synthetic routes to more potential ligand frameworks where all the pyridines can be differentially functional­ized. The overall goal is to move the reduction potentials of these complexes to more positive values, which may lead to a decrease in the overpotential for catalytic reduction of water to hydrogen. We are also testing conditions and electrode materi­als that will allow R. eutropha to grow autotrophically with H2 generated in situ from electricity by the MoPy5 electrocatalyst.

As mentioned in the Introduction, we are pursuing two strategies to use these catalysts to generate H2 electrochemically for autotrophic growth and biofuel pro­duction. These strategies involved tethering the electrocatalyst to the electrode sur­face and to the surface of R. eutropha by expressing heterologous proteins on the surface that will bind metal complexes (Fig. 5). These two strategies will be evalu­ated in comparison to standard methods for electrolysis of H2O to generate H2 and O2 for R. eutropha growth.

Acknowledgments This work is funded by the ARPA-E Electrofuels program. We would also like to thank the Joint BioEnergy Institute (JBEI) for the use of its facilities and equipment; JBEI is supported by the Office of Science, Office of Biological and Environmental Research, of the U. S. Department of Energy. Work at Lawrence Berkeley National Laboratory is performed under the auspices of the U. S. Department of Energy through contract DE-AC02-05CH11231.

[1] The interest to be paid on these loans was lower than the rate of inflation which resulted in a nega­tive real interest rate.

[2] Theoretically one can produce 0.684 L of ethanol with 1 kg of sugar which is fairly close to the value established by the decree 76,593.

[3] Enzyme inactivation can result from a variety of irreversible processes, such as nonspecific adsorption to substrate, aggregation related to unfolding, and covalent chemical modification. These irreversible processes are favored by the relatively harsh conditions (high temperature and/ or acidic or alkaline pH) used for pretreatments that increase digestibility.

[4] This is particularly notable given that most GH families in Table 1 include numerous enzymes with noncellulolytic activities.

[5] Since modification is likely to be somewhat random, the odds of getting a block of i unmodified sugars in a row is nonzero, but it is likely to be quite small. Assuming each of the three hydroxyls per anhydroglucose are equally reactive and that their reaction does not influence reactivity of proximal hydroxyl groups, the probability is (1 — 0.7/3)3i = 0.673i. For i = 3, this is 0.09, and drops by a factor of 0.45 for every additional unmodified anhydroglucose.

[6]Endocellulase activity should strongly decrease the viscosity of CMC solutions by greatly decreasing chain length, whereas exocellulase activity should have little effect on viscosity.

[7] enzymes

[8]LD is lethal dose. The LD50 is the dose that kills half (50%) of the animals tested. The animals are usually rats or mice, although rabbits, guinea pigs, hamsters, and so on are sometimes used.

[9] Primary energy consumption comprises commercially traded fuels only, and excluding fuels such as wood, peat, animal waste, geothermal, wind and solar power generation.

T. M. Mata (*) • C. A.V. Costa

Laboratory for Process, Environmental and Energy Engineering (LEPAE),

Faculty of Engineering, University of Porto (FEUP), R. Dr. Roberto Frias, s/n, 4200-465 Porto, Portugal e-mail: tmata@fe. up. pt

A. A. Martins

Center for Transport Phenomena Studies (CEFT) , Faculty of Engineering,

University of Porto (FEUP) , R. Dr. Roberto Frias, s/n, 4200-465 Porto, Portugal

[11] K. Sikdar

National Risk Management Research Laboratory, Office of Research and Development,

U. S. Environmental Protection Agency, 26 West Martin Luther King Drive,

Cincinnati, OH 45268, USA

N. S. Caetano

Laboratory for Process, Environmental and Energy Engineering (LEPAE),

Faculty of Engineering, University of Porto (FEUP), R. Dr. Roberto Frias, s/n, 4200-465 Porto, Portugal

Department of Chemical Engineering, School of Engineering (ISEP),

Polytechnic Institute of Porto (IPP), R. Dr. Antonio Bernardino de Almeida, s/n, 4200-072 Porto, Portugal

J. W. Lee (ed.), Advanced Biofuels and Bioproducts, DOI 10.1007/978-1-4614-3348-4_31, 745

© Springer Science+Business Media New York 2013

G. J. Moridis (H) • J. Rutqvist • T. Kneafsey • M. T. Reagan • M. Kowalsky Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA e-mail: GJMoridis@lbl. gov

T. S. Collett

U. S. Geological Survey, Denver, CO 80225, USA

R. Boswell

National Energy Technology Laboratory, Morgantown, WV 26507, USA

[13] Hancock

RPS Energy, Calgary, AB, Canada T2P 3T6 C. Santamarina

Georgia Institute of Technology, Atlanta 30332, GA M. Pooladi-Darvish

University of Calgary, Calgary, Canada T2N 1N4 E. D. Sloan • C. Coh

Colorado School of Mines, Golden, CO 80401, USA

J. W. Lee (ed.), Advanced Biofuels and Bioproducts, DOI 10.1007/978-1-4614-3348-4_37, 977

© Springer Science+Business Media New York 2013

[14] The NASA estimate, based on a Congressional Budget Office report, A Budgetary Analysis of NASA’s New Vision for Space, found the Apollo program cost in 2005 dollars to be approximately $170 billion. The estimate includes costs for research and development; procurement of rockets, command and lunar modules; management; facilities, including construction and upgrading; and flight operations.

Chiara Samori and Cristian Torri

Abstract Lipid extraction is a critical step in the development of biofuels from microalgae. The use of toxic and polluting organic solvents should be reduced and the sustainability of the extraction procedures improved in order to develop an industrial extraction procedure. This could be done by reducing solvent amounts, avoiding use of harmful solvents, or eliminating the solvent at all. Here we describe two new processes to extract hydrocarbons from dried and water-suspended samples of the microalga Botryococcus braunii. The first one is a solvent-based procedure with switchable polarity solvents (SPS), a special class of green solvents easily convertible from a non-ionic form, with a high affinity towards non-polar compounds as B. braunii hydrocarbons, into an ionic salt after the addition of CO2 , useful to recover hydrocarbons. The two SPS chosen for the study, based on equimolar mixtures of 1,8-diazabicyclo-[5.4.0]-undec-7-ene (DBU) and an alcohol (DBU/octanol and DBU/ethanol), were tested for the extraction efficiency of lipids from freeze-dried B. braunii samples and compared with volatile organic solvents extraction. The DBU/octanol system was further evaluated for the extraction of hydrocarbons directly from algal culture samples. DBU/octanol exhibited the highest yields of extracted hydrocarbons from both freeze-dried and liquid algal samples (16 and 8.2%, respectively, against 7.8 and 5.6% with traditional organic solvents). The second procedure here proposed is the thermochemical conversion of algal biomass by using pyrolysis; this process allowed to obtain three valuable fractions, exploitable for energy purpose, fuel production, and soil carbon storage: a volatile fraction (37% on dry biomass weight), a solid fraction called biochar (38%) and, above all, a liquid fraction named bio-oil (25%), almost entirely composed by hydrocarbon-like material, thus directly usable as fuel.

C. Samori (*) • C. Torri

Interdepartmental Research Centre for Environmental Sciences (CIRSA), University of Bologna, via S. Alberto 163, 48123 Ravenna, Italy e-mail: chiara. samori3@unibo. it; cristian. torri@unibo. it

J. W. Lee (ed.), Advanced Biofuels and Bioproducts, DOI 10.1007/978-1-4614-3348-4_27, 651

© Springer Science+Business Media New York 2013

1 Introduction

The need to replace fossil fuels with fuels derived from renewable biomass is cur­rently focused on biodiesel from oleaginous plant seeds and ethanol from sugar — cane/corn; however, this first-generation biofuels, primarily produced from food crops and mostly oil seeds, are limited in their ability to achieve targets for biofuel production, climate change mitigation, and economic growth; moreover, the recent dramatic increase of food stocks prices has become a worldwide emergency. Because of these environmental and social concerns, the attention is recently shift­ing towards the development of next-generation biofuels mainly produced from non-food feedstock [1], by converting for example the highly abundant and wide­spread non-edible lignocellulosic fraction of plants. A further exploitable source of biofuels relies on the aquatic environment, specifically on micro and macroalgae; lipids, which include acylglycerols and hydrocarbons, represent the most valuable fraction of microalgal biomass as their high energy content per mass unit is similar to conventional fuels. Several oleaginous microalgae (with lipid content exceeding 20% of their dry weight) have been exploited to this purpose [2], and the biodiesel obtained has been claimed to be more convenient than conventional biodiesel from plant seeds [3, 4]. Benefits rising from the utilization of aquatic over terrestrial bio­mass include: (1) higher sunlight use efficiency (about 5% vs. 1.5% [5]), (2) utiliza­tion of marginal areas (e. g. desert and coastal regions), (3) possible coupling with other activities (e. g. wastewater treatment, CO2 sequestration) [6-9], (4) minor dependence on climatic conditions, (5) availability of a larger number of species, and (6) easier genetic manipulation to modify chemical composition (e. g. lipid con­tent) [10]. However, the industrial development of fuels from microalgae is still hampered by higher overall costs with respect to both fossil fuel and first generation of biofuels counterparts: operating open ponds and bioreactors are expensive and the harvesting of algal biomass is energy costly [11]. For this reason, the net energy balance from microalgae cultivation is still debated [12, 13]. Moreover, besides the cost of growing and collecting microalgae, downstream processes are to be taken into account to evaluate the overall productivity. Botryococcus braunii is a freshwa­ter colonial green microalga proposed as a future renewable source of fuels because it is capable of producing high levels of liquid hydrocarbons [14]. There are three main B. braunii races, each one synthesizing different types of olefinic hydrocar­bons: the A, B, and L races. The A race (Fig. 1) accumulates linear olefins, odd numbered from C23 to C31, chiefly C27, C29, and C31 dienes or trienes; some studies have revealed that oleic acid is the direct precursor of these specific olefins [15] and that decarboxylation of very long chain fatty acid derivatives, activated by a b-sub — stituent, is the final step which leads to the formation of the terminal unsaturation [16]. The B race produces polyunsaturated triterpenes (botryococcenes), while the L race synthesizes one single tetraterpenoid hydrocarbon named lycopadiene [17, 18]. Both A and B races contain similar amounts of lipids (approximately 30% on a dry weight basis), but with a very different composition: in the A race hydrocarbons, non-polar lipids and polar lipids are, respectively, 25, 60, and 15% of the total lipids,

Fig. 1 Botryococcus braunii, A race

whereas in the B race the percentages are 71, 9, and 20%, clearly indicating that one quarter of the dried biomass of the B race is composed by hydrocarbons [19].

Specifically for B. braunii, the bulk ofhydrocarbons is located in external cellular pools and it can be recovered from algal biomass by means of physical process, named cold press and typically used to extract more traditional food oils as olive oil, and by means of chemical process (extraction with solvents) or both [20]. The chemical pro­cess, mainly used for the extraction of industrial oils such as soybean and corn oils, is generally based on an extraction with n-hexane, to obtain vegetable oil in higher yields and with a faster and less expensive process [21] . However, the existing solvent approach is characterized by several problematic aspects, such as the high solvent/ biomass ratio, solvent hazard (including solvent toxicity, volatility, and flammability) and large solvent losses (e. g. in the extraction process of soybean oil, n-hexane losses are 1 kg per tonne of beans processed [22]). Because of this general lack of “green­ness” in the chemical extractive processes, in the last years different efforts have been made to reduce the use of toxic and polluting organic solvents and to improve the sustainability of the extraction procedures from aquatic and terrestrial biomass, for example by using supercritical fluids [23, 24].

Here we present two novel methods for the extraction of lipids from B. braunii, comparing the extraction efficiency of the new processes with those of traditional organic solvents. The first method [25] is a solvent-based process, more sustainable than the traditional solvent extraction because it involves the use of switchable polar­ity solvents (SPS) [26, 27], a “new” class of green solvents, considerable as reversible ionic liquids, with the unique and advantageous feature of having switching solubility behaviour, correlated with reversible polarity. This feature can be successfully exploited in practical applications as extraction procedures or chemical reactions, bypassing the cumbersome need to change solvent in each step of the process itself.

The second method is based on the thermochemical conversion of B. braunii biomass by using pyrolysis [28], in order to obtain, directly and in one step process, a liquid fraction rich in lipids, a gaseous fraction useful for energy purposes, and a soil-amending co-product called biochar [29] .

The solar receiver tubing must have a specified length so that photosynthetic growth can be optimised. It has been shown that the maximum tube length relies on three parameters: liquid velocity, dissolved oxygen concentration and the rate of oxygen production by photosynthesis. Generally, a tube run in a photobioreactor should not surpass 80 m. However, the maximum length of tubing is dependent on solar inten­sity, biomass concentration, liquid flow rate and initial oxygen concentration at the tubing entrance [22]. Molina Grima et al. [23] states that “other than “scale up” by multiplication of identical tubular modules, the only way to increase volume is by increasing length and/or diameter”. Ten years on, the debate over scale-up is still prevalent, with no clear solution readily available. A possible solution to scale-up is to make use of current cultivation designs and employ several cultivation units to produce a significant amount of biomass. However, the process must produce enough biomass such that it will offset extensive equipment costs.

The annual costs represent a range of expenses incurred in the running of the plant from payroll charges to maintenance costs. The most significant of these costs is depreciation, which is calculated using the FCI based on 10-year plant life as rec­ommended in Peters et al. [26] . All other annual costs were calculated using the methods specified in Molina Grima et al. [21] with the exception of labour, supervi­sion, wastewater treatment, and goods and services tax. Labour was assumed to be constant, with 12 employees working during the day and 3 working at night, each charged at the standard labour hourly rate given in ENR (US $34.16). Supervision was also assumed to be constant, with two managers working during the day and one at night, charged at the skilled labour hourly rate again outlined in ENR (US $44.99). Wastewater treatment cost was also estimated based on the costing data reported by Molina Grima et al. [21]. Finally, goods and services taxes were charged at a rate of 10%, reflecting the tax codes applicable to Australia.

In all LC studies, a reference flow is needed to which all other modeling flows of the system will be related [13]. This flow must be a quantitative measure and for some industries, e. g., steel, the choice is usually obvious like X kg steel at the foundry. In other cases, including algae-to-energy systems, this decision can be more complicated. Recent studies have selected a wide variety of functional units (FU) including volume of biodiesel, dry mass of algae produced, kilometers of truck transport, and total energy embedded in the algae assuming the biomass is burned (see Table 2). All of these FUs are valid bases from which to evaluate algae LC, but this diversity in FUs does not make for straightforward comparison between studies. The lack of consensus on a standard FU reflects the lack of indus­try agreement on what the best products to make with the algae will be. Some of the assumptions about goal and scope setting carry over into the functional unit since a FU of liters of biodiesel will inherently exclude the value that could come from a by-product such as ethanol.

LEmEMD

5% RMEE Boundary

Fig. 2 Many studies assume that the upstream impacts of delivering fertilizers and carbon dioxide should not be included. A cut off of 5% was assigned to LC contributions that would be neglected in the analysis (from Sander and Murthy [24])

In LCA more broadly, FUs sometimes require that a performance constraint be applied in order to normalize between dissimilar systems. A carpet, for example, is quieter than a wood floor, even if the latter is more durable. Using a square meter of flooring as the functional unit may overlook performance characteristics (noise buffering and durability) that will ultimately impact the analysis [1]. In the case of algae, performance constraints are certainly limiting in a few important ways. When benchmarking algae to other terrestrial crops, it is useful to apply an FU that is com­monly accepted by the biofuels industry. Though bushels of corn or liters of ethanol do not apply directly, analogs are possible. For example, algae might be compared in terms of dried biomass generated per unit area or liters of biodiesel produced per unit area per time. Energy content can be used as an FU, though it can overlook important differences between biomass. Algae may have a high heating value com­parable to switchgrass though in practice, converting algae to usable fuel is quite a bit more straightforward.

Joao C. M. Carvalho, Raquel P. Bezerra, Marcelo C. Matsudo, and Sunao Sato

Abstract This chapter comments on fed-batch cultivation of Arthrospiraplatensis under different carbon and nitrogen sources, pH, temperature, light intensity, type of photobioreactor and typical parameters of the fed-batch process, such as feeding time, addition protocol and flow rate. Inexpensive nitrogen sources, such as urea, ammonium salts and nitrogen-rich wastewaters can be used for A. platensis cultiva­tion, with results that can be comparable to those with classical nitrate sources. Closed photobioreactors are useful for preventing ammonia loss. The use of organic carbon sources needs to be carried out under aseptic conditions, and it is necessary to evaluate the best supplying conditions when using fed-batch process. The addi­tion of CO2 ensures the control of pH and, at the same time, supply of the carbon source into the culture medium. The fed-batch process can be useful for the produc­tion of A. platensis using CO2 from industrial plants, particularly from industrial alcoholic fermentation.

1 Introduction

The cultivation of microalgae and cyanobacteria is an important current issue because of the possibility of supplying human needs related to food production and removal of atmospheric or industrial carbon dioxide. Arthrospira (Spirulina) plat­ensis, Dunaliella salina and Chlorella vulgaris are among the most studied photo­synthetic microorganisms, but several other cyanobacteria and microalgae have been investigated lately, mainly for biodiesel production.

J. C.M. Carvalho (H) • R. P. Bezerra • M. C. Matsudo • S. Sato

Department of Biochemical and Pharmaceutical Technology, University of Sao Paulo, Av. Prof. Lineu Prestes 580, Bl. 16, Sao Paulo 05508-900, SP, Brazil e-mail: jcmdcarv@usp. br

J. W. Lee (ed.), Advanced Biofuels and Bioproducts, DOI 10.1007/978-1-4614-3348-4_33, 781

© Springer Science+Business Media New York 2013

The increasing demand for protein sources and other high biological value products, such as polyunsaturated fatty acids and pigments, associated with the need of the development of new technologies that contribute to the mitigation of environmental pollution indicates that the market for microorganisms such as A. platensis is going to increase in the coming years.

The previous uses of photosynthetic microorganisms as food are related to events in China 2,000 years ago, where Nostoc was used in periods of food shortage. Additionally, Spirulina sp. was consumed by the Aztecs in the Mexico Valley and by people living near Chad Lake in Central Africa [54]. They have been consumed by Africans, where French researchers first reported in 1940 the use of Spirulina platensis as food [54] .

Currently, the correct scientific designation for S. platensis is A. platensis [95] . Despite this, in this chapter, it was maintained the denomination given by the authors of the cited works. The genus Arthrospira (family Cyanophyceae) encompasses the photosynthetic cyanobacteria with helically coiled trichomes along the entire length of the multicellular filaments and visible septa (Fig. 1). The last characteristic dif­ferentiates this genus from true Spirulina which has invisible septa [16].

Arthrospira (Spirulina) platensis is one of the most promising microorganisms, among microalgae and cyanobacteria, not only to be used as food but also for other industrial applications because of its composition.

It contains a great amount of polyunsaturated fatty acids and pigments such as phycocyanin and zeaxantine [24]. Palmitic, linoleic, g-linolenic, and oleic acids are the predominant fatty acids in S. platensis. g-Linolenic is only found in significant amounts in breast milk, some fruits, species of fungi and cyanobacteria [25].

S. platensis is also an interesting source of chlorophyll, since this microorganism synthesizes only chlorophyll a, which is more stable than chlorophyll b, very com­mon in vegetables. Moreover the cell wall is composed of mucopolysaccharides and therefore easily digested [44], which is an advantage for the bioavailability of cell components.

This cyanobacterium shows low nucleic acid content in dry biomass (4-6%) in comparison with yeasts (8-12%) and other bacteria (20%) [3], so the daily intake of this biomass would not cause any damage to the human body [73].

Besides the high protein content in dry biomass, S. platensis shows a satisfacto­rily balanced amino acid content, possessing even methionine, which is absent in most microalgae [35]. About 20% of the cellular protein is represented by the main pigments in this microorganism, called phycobilins [82] .

This biomass also contains important vitamins such as cyanocobalamin (B12), pyridoxine (B6), thiamin (B1), tocopherol (E), and phylloquinone or phytonadione (K) [12]. Moreover, recent studies indicate that some trace elements such as chro­mium III [57] and selenium [19] can be accumulated in S. platensis biomass depend­ing on the cultivation conditions.

In fact, Spirulina spp. are noted in the literature as an alternative protein source, due to the high protein content in dry biomass (reaching as high as 70%), good digestibility, low nucleic acid content, and presence of vitamins, polyunsaturated fatty acids, immunomodulatory polysaccharides, pigments, and antioxidants [24] .

S. platensis is mainly used as a food supplement. One of the applications is the use of this microorganism as a source of pigments for food industries [33, 69]. S. platensis was also shown to act as a prebiotic, improving the growth in vitro of lactic acid bacteria such as Lactobacillus lactis, Lactobacillus delbrueckii and Lactobacillus bulgaricus [73]. For application in animal feed, some researchers have studied the use of S. platensis in aquaculture, to feed shrimp larvae, for instance [46].

Another important aspect of this microorganism is the possibility of obtaining bioactive compounds [21]. Since the 1980s, several studies have evaluated the use of S. platensis as a dietary supplement for intestinal disorders [39] , diabetes melli — tus, hyperglycemia [74], hyperlipidemia [67, 85], anemia [17], and hypertension [102). Moreover, it can act as an anti-inflammatory )101]. Recent studies also focused on the isolation of fatty acids, particularly the polyunsaturated ones, and pigments from photosynthetic microorganisms. Such characteristics indicate that this microorganism can be a source of molecules with potential use in pharmaceuti­cal and cosmetic industries as well. Besides, Arthrospira (Spirulina) spp. have been used for the removal of heavy metals from wastewater [27, 45], and it is important to emphasize its potential for CO2 biofixation [114], including CO2 from ethanol production plants [15, 40, 63].

In the large-scale production process, S. platensis can be easily cultivated due to the fact that it grows at high alkalinity and high salinity inorganic medium, with high content of carbonate and bicarbonate. These characteristics make it possible to inhibit or prevent contamination. Large-scale cultivation can thus be carried out in open ponds, which is very common in algae cultivation farms, where 5,000 m2 ponds are employed [93], even though there are several studies about their cultiva­tion in closed bioreactors. A. platensis biomass recovery is also facilitated due to its fi lamentous morphology.

Tacon and Jackson [104] list the following advantages for cultivation of S. platensis: they are able to use both organic and inorganic carbon sources; they exhibit a short generation time under optimum growth conditions; and they are easily cultivated in small areas. Cyanobacterial strains are carefully selected among collections around the world, which are periodically sub-cultured in the laboratory in order to maintain actively growing cells. Major criteria in the selection of strains are growth rate, biochemical composition, and resistance to environmental stress at each production site. It must be emphasized, however, that a strain that shows a good performance in the laboratory does not always display the same behavior in an outdoor open pond operation [93] .

Even though microalgae and cyanobacteria have been used by humans for a long time, microalgal biotechnology has only begun in the middle of the last century. Currently, 5,000 tons of dry microalgal biomass is marketed per year, representing up to US$1.25B [99]. The production of photosynthetic microorganisms consider­ably increased in the world due to the possibility of using this kind of culture for oxygen production and as a source of protein for food in space travels [9].

At the end of the 1970s, Sosa Texcoco Co., in Mexico, was the first responsible for large-scale Spirulina production [23, 93]. Afterward, several countries such as Taiwan, India, USA and Japan also started producing this cyanobacterium in open ponds [93]. Among the different cyanobacterial species, A. platensis stands out due to its characteristics related to cell composition, cell growth and cell recovery.

PLE is another technique that, nowadays, is regarded as an advanced extraction technique, due to the advantages that presents over other traditional extraction mechanism. PLE is based on the use of high temperatures and pressures so that the solvent is maintained in the liquid state during the whole extraction procedure. As a result of the application of these particular conditions, faster extraction processes are obtained in which generally the extraction yield is significantly higher than that obtained using traditional extraction techniques, besides, using lower amounts of organic solvents. Moreover, most of the instruments used for PLE are automated, allowing the development of less labor intensive methods and improving reproducibility.

The principles governing this kind of extraction and providing the above men­tioned characteristics are: (a) the mass transfer rate is improved as a result of the increment on the solubility of the compounds as a consequence of the increase of the extraction temperature; (b) under the PLE experimental conditions, the sur­face tension of the solvent is reduced, allowing a better penetration of the solvent into the sample matrix, increasing likewise the mass transfer; (c) the effect of the pressure theoretically could help to matrix disruption, increasing again the mass transfer rate.

Method development in PLE is by far easier than in SFE, since less parameters influencing the extraction should be considered. Once the solvent has been selected according to the nature of the compounds to be extracted, only two parameters are of significant importance: extraction time and extraction temperature. Although the extraction pressure could help to disrupt the matrix enhancing the mass transfer of the analytes contained on it, as it has been already mentioned, in practice, several reports have shown that the influence of this parameter is not significant once the pressure is high enough to maintain the solvent in the liquid state. The extraction temperature has to be optimized always keeping in mind the possible thermal deg­radation effects that might occur over the interesting extracted compounds. Although generally an increase in the temperature produces the subsequent increase in the extraction yield, for bioactive compounds, too high temperatures might lead to the degradation of these compounds. Therefore, this value should be carefully maxi­mized just to the level in which the interesting compounds start to get degraded. On the other hand, the extraction time has to be minimum enough to have an adequate mass transfer. Longer extraction times would result on slower extraction procedures and could also favor the thermal degradation, once the solvent solution is saturated with analytes from the food matrix. Therefore, quite simple experimental designs, such as full factorial designs with two factors and three levels can be useful to opti­mize the bioactives PLE extraction conditions.

Compared to SFE, the possibility of choosing among a high number of solvents causes PLE to be more versatile in terms of polarity of the bioactive compounds to be extracted and thus, the solvent will be selected depending on their nature. However, this technique is considered by far less selective than SFE. Therefore, it is important to keep in mind, that even if the extraction of the bioactives is attained, it would be possible to find other interfering compounds in the obtained extract. To avoid this problem, other steps can be included. For instance, an extraction step using hexane/acetone as solvent was performed before the PLE of phenolic com­pounds from several algae species using 80% methanol in water at 130°C for 20 min (two 10 min cycles) [132]. Ethanol has been selected to extract antioxidants from different species, such as Synechocystis sp. and Himanthalia elongata [143] or anti­microbial compounds from H. pluvialis [165]. Generally, the best extraction condi­tions in these applications were obtained at mild temperatures, around 100°C.

Moreover, PLE can be applied using a wide variety of extraction solvents, although GRAS extraction solvents, like ethanol, are most commonly used. When the extraction solvent is water, this technique is commonly called subcritical water extraction (SWE). The principles of extraction are the same, but in this case, another parameter has critical importance, the dielectric constant of water. This property of water is greatly modified with the increasing temperature when water is maintained in the liquid state. In fact, the value of dielectric constant of water (e) can vary from 80 at room temperature to values around 25 when is submitted to temperatures of ca. 250°C. This value is similar to the one presented by some organic solvents at room temperature, such as ethanol or methanol, and thus, the use of SWE could be an alternative to the use of this type of solvents in some applications. This technique has been already used to explore the possibility of obtaining antioxidants from dif­ferent microalgae species [52, 55]. However, the wide development of novel appli­cations for the extraction of bioactives from algae by using SWE has not been fully explored so far.

The pH is another important environmental factor for the ADP. Different groups of methanogens have different ranges of optimum pH. The acidogens exhibit maxi­mum activity at pH 5.5-6.5 while the optimum for methanogens is pH 7.8-8.2 [68]. Since the methanogens are more sensitive to pH variation, the pH in anaerobic digesters is usually maintained in the range of 7-8. Rapid inhibition of methanogens at pH higher than 8 can be caused by dissociation of NH4+ to the neutral NH3 form [69]. The presence of alkalinity is an important marker of pH persistence in anaero­bic digesters. The bicarbonate alkalinity buffers the fluctuations in the generation of VFAs and carbon dioxide at pH close to neutral. A stable ADP is characterized by the bicarbonate alkalinity in the range from 1,000 to 5,000 mg/L as CaCO3 [70].

The ratio between VFAs to alkalinity should be in the range of 0.1-0.25. A further increase of the ratio of VFAs to alkalinity indicates possible process deterioration and requires the OLR to decrease in order to lower the VFA formation rate.