Biofixation of carbon dioxide from coal station flue gas using Spirulina sp. LEB 18 and Scenedesmus obliquus LEB 22

1 Laboratory of Biochemical Engineering, College of Chemistry and Food Engineering, Federal University of Rio Grande (FURG), PO Box 474, 96203–900, Rio Grande, RS, Brazil. 2 Laboratory of Microbiology and Biochemistry, College of Chemistry and Food Engineering, Federal University of Rio Grande (FURG), PO Box 474, 96203–900, Rio Grande, RS, Brazil. 3 Zero Emissions Research & Initiatives – ZERI, Brazil.


INTRODUCTION
The burning of fossil fuels, especially coal, in thermoelectric power plants is problematic due to the emission of the greenhouse gas carbon dioxide (CO 2 ) and the elevated temperature production of environmentally damaging sulfur and nitrogen oxides (SO x and NO x ).The sustainability of coal use is linked to the reduction of these pollutants.Processes need to be developed for CO 2 capture and sequestration, with one alternative being the use of photosynthetic and autotrophic microorganisms that are capable of removing CO 2 from gas streams; cyanophytes and chlorophytes (microalgae) can remove up to 50% of CO 2 and survive in the presence of toxic compounds, such as SO 2 and NO (Radmann et al., 2011).
Microalgae use CO 2 to multiply and produce biocompounds of interest, such as proteins, fatty acids and pigments.These microorganisms present antioxidant properties and are sources of carbohydrates, lipids, vitamins, essential amino acids and polyunsaturated fatty acids.The microalgae CO 2 biofixation mechanism is based on the ability of these microorganisms to perform photosynthesis but with higher fixation rates than those of higher plants.Another important feature is that gases that are emitted may be injected directly into microalgal cultivation through photobioreactor-coupled systems, while higher plants capture gases from the environment.Injection culture tanks may be used without the need to cool these gases because many microalgae have extremophile characteristics that allow them to withstand high temperatures (Pandey et al., 2014) .The flue gas emitted by Candiota II contained approximately 120 g L -1 CO 2 , 2 to 2.5 g L -1 SO X and 0.3 to 0.5 g L -1 NO X (Migliavacca, 2005;De Morais and Costa, 2008).
The objective of this study was to use Candiota II flue gas (hereafter simply called 'flue gas') for the cultivation of Spirulina sp.LEB 18 and Scenedesmus obliquus LEB 22 and to measure their kinetic characteristics and capacities for CO 2 fixation.
For the Spirulina sp.LEB 18 experiments, the Zarrouk medium carbon source (NaHCO3) was omitted and replaced with flue gas containing a mean CO2 concentration of 102 g L -1 .For the Scenedesmus obliquus LEB 22 trials, we used MC medium with the carbon source (NaHCO3) replaced with the same concentration of CO2 as in theSpirulinatrials.In both cases, the control experiments were performed using unmodified Zarrouk or MC medium as the controls.
We grew the organisms separately in acrylic raceway-type photobioreactors (length = 0.68 m, width = 0.18 m, and height = 0.07 m) with a working volume of 5 L, agitated by blades rotating at 18 rotations min -1 .The experimental conditions were maintained between 14 and 28°C for 960 h (40 days) at a light intensity of 32.5 µmol m -2 s -1 provided by 40 W daylight-type fluorescent lamps (white) and a 12 h photoperiod.The initial biomass concentration of the cultures was 0.15 g L -1 (Radmann and Costa, 2008).To maintain the original volume, evaporation was compensated for by adding sufficient distilled water each day.
The flue gas, which came from burning coal under normal operating conditions at Candiota II, contained 102 g L -1 CO2, 0.654 g L -1 carbon monoxide (CO) and 0.087 g L -1 NOX.The flue gas was collected daily during each experimental run and compressed and stored in industrial cylinders for aspiration into the cultures.Before the flue gas was aspirated into the photobioreactors, the gas stream passed through a column of 100 g L -1 hydrogen peroxide (H2O2) to remove sulfur dioxide (SO2) (Colle et al., 2005).The flue gas was aspirated into the culture media using sprinklers, which were uniformly distributed over the base of the photobioreactors, and aspiration was carried out for 15 min every 2 h during the 12 h photo period.
The biomass concentration (X, g L -1 ) was measured every 24 h by reading the optical density (O.D.) at 670 nm using a 700 Plus spectrophotometer (Femto) and a previously established biomass versus optical density calibration curve.The maximum and minimum temperatures and pH were measured at the same time as the O.D. values, and we determined the alkalinity of the culture medium every three days (American Public Health Association, 1998).These temperature and pH values were used to calculate the dissolved CO2, HCO3 -and CO3 2-concentrations (Carmouze, 1994).All of the samples were collected in triplicate.

Growth kinetics and statistical analysis
The kinetic characteristics of both organisms were evaluated by measuring the maximum specific growth rate (µmax, d -1 ) as calculated from the exponential regression of the logarithmic phase of the cell growth curve (Bailey, 1986).The generation time (tg, d) was calculated by tg = ln2/μmax and the productivity (g .L -1 .d -1 ) according to the methods of Borzani et al. (2008).Pmaxrepresents the maximum productivity obtained.
The efficiency of carbon biofixation (F) was calculated according to Equation 1: 1 Where F = biofixation efficiency (%, v/v); P = productivity (g L -1 d -1 ); XC/X = mass fraction of carbon in the cell (0.402 gC/gX); VBR = volume of the culture medium in the bioreactor (L); mCO2 = mass of carbon (as CO2) supplied in each experiment (g d -1 ); and MCO2 and MC = molar masses of CO2 and C (MCO2 = 44 g, MC = 12 g), respectively.The concentration of carbon in the final biomass (gC The fraction of dissolved carbon (C, % v/v) in the culture medium derived from the carbon supplied by the gas in each cycle was calculated by the ratio between the mass of inorganic carbon (CO2, HCO3 -and CO3 2-) and the flue gas in the culture medium using Equation 2: 2 Where Ct=15 = carbon concentration from dissolved CO2, HCO3 -or CO3 2-at the end of the flue gas injection (g L -1 ); Ct=0 = carbon concentration from dissolved CO2, HCO3 -or CO3 2-at the start of the flue gas injection (g L -1 ); VFBR = volume of the culture medium in the bioreactor (L); and 0.18 = carbon mass supplied in each injection cycle (g).
The kinetic data for both organisms were evaluated using Analysis of Variance (ANOVA) at p=0.05.

RESULTS AND DISCUSSION
Biomass (X max ), specific growth rate (µ max )and productivity (P max ) Several compounds that are present in flue gas cannot only affect the efficiency of carbon biofixation, but are also potentially toxic to the cultivation of Spirulina LEB 18 and Scenedesmus obliquus LEB 22.The CO 2 concentration in the flue gas ranged from 0.507 to 0.703 g L -1 , while that of NO X was 0.087 g L -1 and that of NO was 0.084 g L -1 . The presence of flue gas in the culture medium increased the biomass concentration of LEB 18 by 35.7% to a maximum of 0.78 g L -1 , which was significant at p > 0.05 (Figure 1).The production of up to 1.50 g L -1 of a Spirulina sp. has been reported, but this occurred when pure CO 2 , free from the other toxic components of flue gas, was used (Lodi et al., 2003).In our experiments, the μ max in the presence of flue gas was 0.026 d -1 for LEB 18 and 0.017 d -1 for the control; thus, the addition of flue gas increased the growth rate of LEB 18 by 34.6%.The μ max from our previous LEB 18 experiments, in which Spirulina was cultivated by injecting 120 g L -1 CO 2 , was 0.028 d -1 (De Morais and Costa, 2007).Other studies have shown that Spirulina exhibits high μ max values between 10 and 40°C (Richmond, 1990;Vonshak, 1997), as confirmed in our study, which showed high μ max values between 14 and 28°C.Although lower cultivation temperatures may decrease photosynthetic activity and thus impair CO 2 fixation (Torzillo and Vonshak, 1994), there have also been reports of higher concentrations of phenolic compounds and unsaturated fatty acids, such as ω-3 and ω-6, in S. platensis grown at temperatures below 35°C (Colla et al., 2007).An increased concentration of CO 2 leads to an increase in the specific growth rate, yield and biomass concentration (Maeda et al., 1995), and an increase in the maximum kinetic values was observed in the LEB 18 experiments with the addition of flue gas.The S. obliquus LEB 22 experiments showed no increase in the biomass concentration when flue gas was added to the medium (Figure 1).This result contrasts with a previous study by our team, which evaluated CO 2 fixation byLEB 22 in the presence of NO and SO 2 and showed that this organism produced a maximum biomass concentration of 0.81 g L -1 at 35°C and grew attemperatures between 20 and 38°C (Radmann and Costa, 2008) Brazil, may have affected growth in the presence of flue gas.Several studies by other groups have indicated that S. obliquus grows in the presence of combustion gases, containing up to 500 g L -1 CO 2 (Hanagata et al., 1992) and 0.300 g L -1 NO (Jin et al., 2008).Previous studies by our team have shown that the growth of LEB 22 was not impaired by the addition of flue gases and removed 27% of NO from a culture containing 60 g L for the control experiments, while previous experiments by our group had μ max values between 0.04 d -1 and 0.18 d -1 in culture medium that was supplemented with pure CO 2 , NO and SO 2 (Radmann and Costa, 2008).
The productivity of LEB 18 and LEB 22 (Figure 2a) showed that the media that were supplemented with flue gas LEB 18 and gave the highest yields, with P . For LEB 22, the values were much lower at P max = 0.2 g m -2 d -1 in flue gas-supplemented medium and 0.5 g m -2 d -1 for the unsupplemented control.The fact that LEB 22 productivity was lower in the supplemented medium may have been because the flue gas contained toxic components that inhibit growth.The final biomass of the LEB 18 grown with flue gas contained 46.8% protein and 4.8% lipid, while the LEB 22 biomass comprised 40.6% protein and 6.2% lipid.We have reported similar lipid values in two previous studies, one using media supplemented with 120 g L -1 CO 2 , which produced a final biomass with a lipid concentration of 5.2% for LEB 18, 3.3% for LEB 22 and 4.6% for the Chlorella vulgaris strain LEB 106 (De Morais and Costa, 2007), and another employing media supplemented with 120 g L -1 CO 2, 60 μL L -1 SO 2 and 100 μL L -1 NO, which resulted in lipid values of 5.97% for LEB 18, 6.18% for LEB 22, 5.21% for LEB 106 and 5% for Synechococcus nidulans LEB 25 (Radmann and Costa, 2008).
Other researchers have reported 57.61% protein and 8.16% lipid in Spirulina platensis biomass when grown in Zarrouk medium without the addition of flue gas (Colla et al., 2007), and the protein content of Spirulina has been reported to vary from 64 to 74% and the lipid content from 6 to 13 % (Vonshak, 1997).All the values are w/w on a dry-mass basis.Our experiments have shown that the cultivation of LEB 18 and LEB 22 in media that were supplemented with Candiota II flue gas produced a biomass that was rich in protein and lipid and has potential for use in, among other things, the production of fertilizers, biofuels and biopolymers.

Biofixation efficiency
The maximum biofixation efficiency (F max ) of CO 2 was 5.66 ± 0.88% for LEB 18 and 0.86 ± 0.56 % for LEB 22 (Figure 2b).The differences that were observed between the experimental and control trials appear to be related to the concentration of free CO 2 in the culture medium, especially for LEB 18, in which the CO 2 concentration in the medium was 25 mg L -1 higher than in the control experiments (Figure 3).For LEB 22, we found a significant difference (p=0.086) in the amount of CO 2 absorbed into the culture medium up to 360 h (day 15 d) of cultivation.
The addition of 150 g L-1 of CO2 to C. vulgaris cultures  resulted in a CO 2 fixation rate of 0.624 g L -1 d -1 (Yun et al., 1997).Membrane technology has been used to aid in CO 2 dispersion in experiments designed to remove this gas from the air.The resultant 10 g L -1 of CO 2 that is dispersed in the culture medium using filter modules increases pH of CO 2 biofixation efficiency from 2 to 20 % (Cheng et al., 2006).Other studies have reported that 4.4 kg of CO 2 from burning natural gas is required to produce 1 kg of Chlorella biomass.
We also discovered that during flue gas injection, the dissolution of CO 2 reduced the pH by altering the equilibria between HCO 3 -, CO 3 2-and CO 2 .In the last 75 min of the experiment, the pH value increased as a result of the CO 2 utilization by the microorganism in the media and the formation of bicarbonate and carbonate (Figure 4).The injection of flue gas can be carried out over a wide pH range without significantly affecting photosynthetic activity (Olaizola, 2003).We found that when flue gas was injected, the largest pH variations occurred at the beginning of cultivation (t < 10 d) and that LEB 18 showed less severe changes in pH than LEB 22.The differences in pH were associated with the alkalinity of the culture media, which averaged 53.1 ± 14.9 g L -1 CaCO 3 for LEB 18 and 16.3 g L -1 ± 6.  CaCO 3 for LEB 22.The higher the alkalinity, the greater the pH stability during cultivation (Vonshak,1997).
Because the rate of flue gas injection is associated with pH buffering capacity, drastic changes in pH to nonoptimal values must be avoided.An alkaline medium favors the formation of highly soluble chemicals, such as HCO 3 -and CO 3 2-, with some of the CO 2 in the injected flue gas converted to these chemical species and thus removed from the effluent gas.The concentrations of HCO 3 -, CO 3 2-and CO 2 derived from the flue gas dissolved in the culture medium are given in Table 1.In the LEB 18 culture, most of the CO 2 supplied by the flue gas remained dissolved in the form of HCO 3 -, which helped to maintain the pH, whereas in the LEB 22 culture, the largest carbon supply remained in the form of CO 2 .The percentage of CO 2 removed from the flue gas was 24.2 % for LEB 18 and 13.2 % for LEB 22.

Conclusions
At the end of the cultivation period, we found that Candiota II flue gas increased Spirulina sp.LEB 18 biomass production by 35%, with a 24% reduction in CO 2 from the flue gas and a 5.66% biofixation of CO 2 .When supplemented with flue gas, the final biomass of LEB 18 contained 46.8% protein and 4.8% lipid, while LEB 22 contained 40.6% protein and 6.2% lipid.Our results indicate that these microorganisms have the potential to be grown in power plants to biofix CO 2 from the flue gas that is produced from coal and can thus contribute to reducing global warming.
-1 CO 2 (De Morais and Costa, 2008).The LEB 22 experiments had μ max values of 0.013 d -1 in the presence of flue gas and 0.016 d -1

Figure 4 .
Figure 4.Changes in pH during the injection cycle and time of operation of reactors for Spirulina sp.LEB 18 (a) and Scenedesmus obliquusLEB 22 (b) when grown with flue gas.

Table 1 .
The (%) of inorganic carbon from the flue gas dissolved in the culture medium in the form of CO2, HCO3 -and CO3 2-, fixed by microalgae (F) and the total CO2 removed.