The nation’s reliance on fossil-based fuels creates problems for the environment and our national security. The production of a renewable source of motor fuels is required. We have designed a production pathway for synthesizing IBT biofuel from CO2, H2, and O2 using the genetically tractable and metabolically versatile bacte­rium, R. eutropha. The majority of the genes required for this pathway are already present in R. eutropha. Metabolic engineering strategies are being implemented to establish a semisynthetic pathway to produce IBT from CO2, H2, and O2. This IBT production pathway has the potential to affect two high-priority environmental con­cerns, capture of CO2 and production of an alternative nonfossil-based fuel.

Acknowledgments We thank John W. Quimby for critical review of this manuscript. D. S. is supported by the following foundations: Nijmeegs Universiteitsfonds (SNUF), Fundatie van de Vrijvrouwe van Renswoude te’s-Gravenhage, and Dr. Hendrik Muller’s Vaderlandsch Fonds. Other authors are supported, fully or in part, by the Advanced Research Projects Agency-Energy (ARPA-E) Electrofuels project. We wish to thank the ARPA-E directors and staff for their support.

1.1 Switchable Polarity Solvents SPS

The concept of “switchable compounds” and in particular of “switchable solvents” has been proposed for the first time by Jessop et al. few years ago [26]; this smart idea is based on the possibility to reversibly switch on and off some properties of a substance when a “trigger” is applied, from one version, with a specific set of properties, to another one, with very different properties. For solvents, such properties

can be polar/apolar, volatile/non-volatile, protic/aprotic. For the practical application of SPS, their switching solubility behaviour is a fundamental feature, correlated with their reversible polarity; an equimolar mixture of DBU (1,8-diazabicyclo-[5.4.0]- undec-7-ene) and an alcohol, for example, behaves as a slightly polar solvent, similar to chloroform, enabling to dissolve apolar compounds such as hydrocarbons, whereas the salt DBU alkylcarbonate, after CO2 treatment, is a polar liquid, very similar to dimethylformamide and immiscible with hydrocarbons (Fig. 5) [26].

The major “greenness” of SPS respect to volatile organic solvents as я-hexane relies mainly on the possibility to develop safer, more economic and more sustainable processes. Chemical processes often involve many steps which can require many specific solvents; after each step, the solvent has to be removed and replaced with a new one, more suitable for the following step, increasing the economic costs and environmental impact of the process itself. Switchable solvents represent a valid answer to these cumbersome procedures, because their properties can be adjusted for the following step, enabling the same solvent to be used for several consecutive reactions or separation steps. Moreover, the recyclability of a SPS is based on the principle of adding and removing CO2, and not on the recovery of the solvent by distillation; this means that the process of recycling is less expensive, less energy costly, and more environmentally friendly. Because of the peculiar recycle method, the solvents in the SPS formulation do not need to be volatile and this feature reduces the risks of fire and explosion, the release in the atmosphere, and the exposure for the operators.

Given the chemical nature of B. braunii lipid oil, the concept of “SPS” fits very well with the development of a new “greener” process of extraction [25]: the lipophi — licity of the non-ionic form of SPS is suitable to extract apolar materials as algal hydrocarbons, whereas the ionic form of SPS, obtainable by a simple bubbling of CO2, has a low affinity for these molecules, guaranteeing an easy recovery (Fig. 6).

The first important point in the development of the suitable SPS extraction process is the choice of the alcohol because it strictly influences the formation of the proper liquid carbonate anion. Previous works [40-42] have demonstrated that bicarbonate and methyl carbonate DBU salts have melting points lower than room temperature, whereas DBU salts with longer alkyl chains are liquids [43], therefore suitable for

hydrocarbons

the extraction process. For this reason in our extraction process of B. braunii, DBU ethyl carbonate (melting points of 35°C) and DBU octyl carbonate (melting points of 30°C) were investigated as suitable candidate solvents (Table 3). In particular, the choice of an alcohol as octanol, hydrophobic and low volatile, should guarantee a good affinity with the apolar matrix which has to be extracted, and a scarce solubility in water, useful in the case of liquid samples.

The kinetics of the extraction process (Fig. 7) clearly shows that DBU/octanol and DBU/ethanol exhibit approximately the same behaviour, with similar hydrocar­bons’ amount extracted after 240 min (14 and 13% yields, respectively). Moreover, the hydrocarbons extraction with DBU/octanol is quite efficient from the beginning, with a yield of 9% after 20 min (65% of the yield at the end of the extraction time after 240 min).

A second important factor for the development of the extraction process with SPS is related to the chemical features of the process itself, and specifically to the eventuality that the presence of specific chemical compounds could prevent the switching of the system and the formation of the right DBU alkylcarbonate salt, affecting the separation process. Free fatty acids in the algal oil for example could

of DBU/octanol and DBU/ethanol data, respectively form an ion pair with DBU, thus preventing the formation of a two-phases system when CO2 is added and altering the stoichiometry of the SPS. In the case of B. braunii, the percentage of free fatty acids is 0.6-0.7% on a dry weight basis (Table 1), thus an irrelevant amount (about 0.18 mg) if compared with the amount of DBU used for the extraction (1 g) eventually able to react with free fatty acids. However, this eventuality could be a problem for the extraction of other kinds of oil with a high content of free fatty acids (as waste cooking oils [44]) and should be taken into account in the development of the extraction system.

A third relevant factor is the recyclability of the system, in terms of feasibility, contaminations of the algal oil, and losses. The efficiency of a non-ionic/ionic cycle with the SPS DBU/octanol in recovering pure hydrocarbons is about 81% of the total amount of hydrocarbons extracted, with 8.1% of hydrocarbons retained in the ionic SPS phase in the second half of the cycle, and about a 10% mechanical loss probably due to small samples size (in our work the extraction procedures were
accomplished on 30 mg of freeze-dried algal samples [25]). These results clearly indicate that the recovery of the hydrocarbons is very good, since the ionic form of SPS retains a rather small amount of the hydrocarbons in the extraction phase; moreover, this amount can be still reduced by scaling up the process and increasing the size of the samples.

The GC-MS evaluation of oil quality after the first cycle of extraction indicates that the oil still contains small amounts of octanol (0.3%) and DBU (0.4%), but bubbling extra CO2 for 1 h at 40°C decreases the levels of contamination to undetectable values because of the precipitation of all the ionic liquid from the oil [25].

The cultivation process relies on photosynthesis. One of the major products from photosynthesis is oxygen; thus as the algae consumes carbon dioxide and photosyn — thesise, the culture experiences a significant increase in dissolved oxygen concen­tration. As mentioned above, excess dissolved oxygen within the culture can inhibit photosynthesis and cause photo-oxidative damage to cells. Molina Grima et al. [22] found that the maximum dissolved oxygen concentration within the culture should not exceed the standard air saturation of the culture by more than 400%. This param­eter constraint is one key issue involved with the scale-up of photobioreactors. Dissolved oxygen cannot be removed within the solar receiver, thus limiting the length of the tubular receiver.

Electricity consumption and raw materials usage were the major running costs resulting from biomass production. Electricity consumption was of particular importance in this analysis as this contributes directly to carbon emissions. All car­bon dioxide consumed by the system was assumed to be supplied free of charge through flue gas from the nearby power station. The analysis of electricity con­sumption centred on the pumping and mixing of fluids in each of the production stages, and the electricity consumed by the centrifuge or filtration equipment in dewatering the culture.

In the cultivation stage, the electricity consumed in pumping carbon dioxide and water throughout the system was estimated using electricity consumption data from

Sazdanoff [30]. Sazdanoff’s consumption data were scaled-up volumetrically to meet the requirements outlined in Table 1. Also considered in the cultivation section was the electricity consumed in mixing the algal culture: by airlift pump in the reactor-style systems or paddle wheel in the raceway ponds. Electricity consumed by the airlift pumps and the paddle wheels was estimated using data from previous studies by Acien Fernandez [1] and Sazdanoff [30], respectively. The only major raw material considered in the cultivation section was the cost of culture medium, where unit costs were based on Molina Grima et al. [21] and the quantity required was developed using information provided in Danquah et al. [8].

All pumping to the dewatering unit operation were assumed to be associated with the cultivation section, thus the only running costs involved in a single-stage dewatering process was the operation of the different dewatering systems. The energy consumption of the single-stage dewatering options was estimated primarily using data provided in Molina Grima et al. [21] . For the dual-stage process, other running costs such as the flocculant and the mixing of algal broth in the flocculation tanks were also considered. Chitosan was the preferred flocculant, with costs esti­mated at US $11/kg [9]. The electricity consumption in the mixing during flocculation was determined using the data provided in Sinnott [32] .

A number of materials and solvents were required to extract saponifiable lipids from the dewatered biomass including ethanol, hexane and water. The quantities of the raw materials required were based on work by Ramirez Fajardo et al. ] 27], whereas costs were based on Molina Grima et al. [21].

A number of common metrics to describe ecological consequences are included in most LCAs. The most common example is global warming potential (GWP) which normalizes greenhouse gas emissions into one number with units of mass emissions in carbon dioxide equivalents. Since several chemicals typically contribute to specific ecological impacts, metrics are very useful for consolidating data. Other examples of common metrics in this class are ozone depleting potential, eutrophica­tion potential, and acidification potential.

The most obvious ecological consequence to include in algae-to-energy studies is GWP since many algae-based energy systems are designed to produce intrinsi­cally low carbon neutral fuels. Because of this desire to produce low carbon fuels, many algae projects have used “sequestration” to describe their activities. In reality, algae-to-energy systems are not a sequestration technology. Sequestration implies that there is long-term storage of CO2 either as a solid carbonate mineral or in the subsurface under high pressure. In theory, algae could be grown and the biomass buried to sequester carbon, but it would be necessary to carefully control the condi­tions under which the carbon was buried such that the biomass was not simply digested by bacteria that could generate methane, effectively compounding the problem. What algae-to-energy systems can offer is a fuel that is closer to carbon neutral than conventional fossil fuels. That is, most of the carbon that will be emit­ted from the combustion of the fuel is not new carbon removed from the ground as in the case of coal or petroleum. This won’t help mitigate the impacts of climate change by reducing atmospheric concentrations, but it will reduce the increase in this concentration by not contributing new carbon. How much carbon these pro­cesses can keep out of the atmosphere is a current topic of investigation. It is impor­tant for the industry to adapt norms with regard to the way it treats carbon dioxide for full transparency.

T. suecica was cultivated in outdoor bag bioreactors using a modified F medium [16]. Each bioreactor contained up to120L of culture and was aerated with com­pressed air. Temperature and illumination depended on day-to-day weather condi­tions. Microalgal cultures from multiple bioreactors were harvested at the same time (with a concentration of ~0.5 g/L), concentrated via industrial centrifugation, and then mixed together to create a homogeneous culture from which all the bio­mass needed for the study was obtained.

3.2.2 Dewatering

The homogeneous culture was further dewatered in a laboratory centrifuge (Heraeus Multifuge 3S-R, Kendro, Germany) at 4,500 rpm for 10 min. The supernatant was discarded and the resulting microalgal paste was rinsed with deionised water to remove residual salts. The paste was then stored in the dark at either 4 or -20°C until further use.

3.2.3 Extraction Pre-treatment

In experiments where extraction was carried out on dried microalgae, microalgal paste (stored at either 4 or -20°C) was dried at 65°C in an oven (Model UNE 500 PA, Memmert GmbH + Co., Germany) for 16 h. A pestle and mortar was used to grind the dried biomass into powder. In experiments where extraction was carried out on the wet paste, the microalgal paste obtained from centrifugation (stored at 4°C) was used directly. The solid concentration of the paste (22.4 wt.%) was deter­mined by drying a portion of the paste and comparing its pre-dried mass to that of the corresponding dried biomass.

Algae are a large and very diverse group of organisms ranging from simple unicel­lular microalgae to giant macroalgae. The morphology of macroalgae or seaweeds resembles the terrestrial plants but the biochemical composition is significantly dif­ferent. The major carbon storage products in terrestrial plants are starch and fructo — san [13-15] that can be easily converted to biogas. However, the main components of terrestrial plant cell walls are cellulose/hemicellulose fibers embedded into a pec­tin matrix and cemented together by lignin [16-18]. This lignocellulosic complex is recalcitrant to biological degradation and requires intensive chemical (acid hydroly­sis, alkaline wet oxidation, ammonia fiber expansion) or thermal pretreatment (steam explosion, hot water) before biological conversion [19-22].

The major components of macroalgae are polysaccharides, algal cell wall lack lignin. The main components of cell envelopes are ulvan and xylan in green algae; carrageen, agar, and xylose in red algae; alginate and fucoidan in brown algae. Cellulose is a structural component of the cell wall in many genera, but only in some green algae is the ratio on a level comparable to terrestrial plants. The main storage polysaccharides in macroalgae are floridean starch in red algae; chlorophy — cean in green macroalgae; laminarin; and mannitol in brown macroalgae.

The biochemical composition of microalgae and cyanobacteria are significantly different from macroalgae. Often carbohydrates are a minor component of cell dry weight, whereas proteins and lipids account for the bulk of microalgal dry weight.

One of the challenges in AD of algae is significant variation in biochemical com­position not only among different phylum or genera, but also among similar species. Biochemical composition depends on many environmental factors, such as tem­perature, salinity, light intensity, and nutrient availability [23-27] .

The enzymatic hydrolysis of algal cell walls and other biopolymers is a promising alternative to energy-consuming mechanical pretreatment and chemical catalytic hydrolysis at high temperature. It has large potential to increase the digestion rate and methane yield. Treatment of WAS by carbohydrases increased the biogas yield by 13% [ 195] . Pretreatment with pancreatic lipases (250 units/mg protein, dose

0. 25 g/L at 25°C for 5.5 h) of slaughterhouse wastewater with pork fat particles resulted in 35% hydrolysis of the neutral fat, but did not significantly increase the fat hydrolysis rate in the anaerobic reactor (sequencing batch type, 25°C) and did not influence the methane yield [244]. The authors suggested that at relatively low temperature (25°C), anaerobic oxidation of LCFA is the rate-limiting step.

Natural hydrolysis pretreatment of green macroalgae in percolators has been extensively studied [175, 245-249]. This method can be viewed as a type of two — step ADP and is discussed in the subsequent reactor design subsection.

The endo-b-1,4-glucanase from Cellulomonas sp. YJ5 hydrolyzed Chlorella sorokiniana cell wall and caused cells lysis after 60-180 min of treatment [250]. Immobilized cellulases hydrolyzed Chlorella cells (reduced sugars yield 62%) and gave a twofold increase in lipids extraction efficiency [251]. Other advantages of enzymatic hydrolysis include an absence of inhibiting by-products and achievement of high selectivity [252]. While this method has a large potential, it is necessary to solve several technological blocks before it can be applied in the biofuel industry.

The major roadblocks are higher cost of enzymes production and their handling, high enzymes to substrate specificity, enormous diversity of algal cell envelope composition and structure.

Jorquera and colleagues compared the energy life-cycle analysis for the production of microalga Nannochloropsis sp. in flat-plate photobioreactors (FPPR), tubular photo­bioreactors (TPR), and raceway ponds (RP) [485] . The NER of these systems was estimated and compared as total energy produced (energy of biomass) over the energy content of the materials, energy required for construction, and the energy required for operation. A preliminary analysis (including energy costs for pumping, mixing, and gas transfer) showed that the NER value of biomass production in the TPR is 0.2. Consequently, the energy demand is larger than the energy content of the produced biomass in the TPR. The estimated NERs for biomass production in the FPPR and RP are 4.51 and 8.34, respectively [485] . The authors also performed a detailed energy life-cycle analysis using the GaBi program and achieved NER values 4.33 and 7.01 for the FPPR and RP, respectively. This result is in agreement with data achieved for the production of Dunaliella biomass where raceway ponds were found to be more efficient [486]. Indeed, the production of C. vulgaris in raceway bioreactors is more environ­mentally sustainable compared to the production in air-lift tubular reactors [487].

These include the following [125]:

• Class 4 deposits: Earlier studies by Moridis and Sloan [140] have indicated the hopelessness of such deposits under any combination of conditions and produc­tion practices.

• Fine sediments (i. e., rich in silts and clays), deformed fractured systems, and hydrates in veins and nodules despite high SH, because of the geomechanical instabilities in such systems under production.

• In Class 2 deposits: Inappropriate well configurations [131].

• In Class 2 deposits: Constant-P production, because it can lead to early break­through and massive water production; however, such an approach can be used in a short-term flow test to determine the HBL properties [55].

• In Class 2 deposits: Deep WZ, and/or permeable overburden and underburden, can drastically reduce gas production [160]. Additionally, the use of multi-well (five — spot) systems involving simultaneous depressurization (at the production well) and thermal stimulation (through warm water injection) appears disappointing [128].

• In all Classes: Permeable upper boundaries drastically reduce production [157,160].

• In all Classes: Pure thermal dissociation methods and/or inhibitor methods have high cost and limited (and continuously eroding) effectiveness [132].

• In all Classes: SH that are so high that the remaining fluids are below their irre­ducible saturation levels. Such hydrates may not be prone to depressurization — induced dissociation.

• In all Classes: Fracturing appears to have limited effect on increasing productivity from hydrate deposits [53,94].