Sustainable carbon capture via halophilic and alkaliphilic cyanobacteria: the role of light and bicarbonate

The two-step photosynthetic biogas upgrading process, which combines CO2 capture by carbonate solution and carbonate regeneration by using aquatic microbial oxygenic photoautotrophs (i.e., cyanobacteria, algae, and diatoms), may provide a potential alternative to the commercial routes used for gaseous biofuel upgrading. Such a process not only provides a green and low energy intensive biogas upgrading pathway but also converts CO2 in biogas into high value biomass. To improve the upgrading performance, the effects of light intensity and NaHCO3 concentration on the growth and the HCO3- transformation characteristics of halophilic and alkaliphilic Spirulina platensis were investigated in this study. Experimental results showed that the light attenuation of S. platensis culture was significant. Increasing light intensity up to 210 μmol m-2 s-1 effectively improved the S. platensis growth and photosynthetic pigment accumulation. S. platensis could grow in the range of 0.05 to 0.6 M NaHCO3, and a maximum biomass concentration of 1.46 g L-1 was achieved under an optimal growth condition of 0.1 M NaHCO3, which was 65.9% higher than at 0.05 M NaHCO3. Moreover, the bicarbonate utilization efficiency reached 42.0%. Finally, in a case study, a biogas stream at a flow rate of 800 m3 h-1 could generate biomass up to 344 kg h-1, corresponding an energy value of 5591 MJ h-1.

Conventional physical/chemical technologies for biogas upgrading such as water, organic solvent and chemical scrubbing, pressure swing adsorption, membrane separation, and cryogenic separation are currently available commercially, but these technologies have high requirements for operation conditions, complex equipment structure, and high energy consumption, which limit their economic and environmental sustainability for biogas upgrading (Khan et al., 2017). Energy requirements of these technologies range between 0.12 to 0.77 kWh m -3 of upgraded biogas (consuming 3%-6% of the energy of the produced biogas), while the capital investment costs are 0.10-0.40 € m -3 of biogas (Patterson et al., 2011;Xia et al., 2015;Gotz et al., 2016).
Alternatively, photosynthetic biogas upgrading can utilize aquatic microbial oxygenic photoautotrophs (AMOPs) to fix CO2 while obtaining microbial biomass that can be used for further production of high valueadded products (Dismukes et al., 2008;Mata et al., 2010;Kumar et al., 2016;Velazquez-Lucio et al., 2018). One of the major bottlenecks of direct biogas upgrading by using AMOPs is the release of photosynthetic oxygen, which leads to contamination of biomethane and a potential explosive hazard (Posadas et al., 2015;Franco-Morgado et al., 2018). In contrast, two-step photosynthetic biogas upgrading separates the two processes of AMOPs cultivation and biogas upgrading and is considered as an effective solution to reduce the oxygen content in upgraded gaseous biofuel (Toledo-Cervantes et al., 2017).
As shown in Figure 1, in the first step, the carbon dioxide content of biogas can be efficiently captured by carbonate solution forming bicarbonate. In the second step, bicarbonate is used as a carbon source for AMOPs cultivation. AMOPs can absorb and utilize bicarbonate through carbon concentration mechanism and produce organic carbon through photosynthesis. During their growth and metabolism, the pH of the medium increases gradually, leading to carbonate regeneration and realizing the carbonate/bicarbonate cycle (Xia et al., 2015).
To improve the efficiency of the photosynthetic biogas upgrading, improving AMOPs growth and the carbonate regeneration rate are very important. A number of studies have reported AMOPs cultivation using concentrated CO2 at various concentrations. However, very few studies

AMOPs
Aquatic microbial oxygenic photoautotrophs APC Allophycocyanin C Carbon content of cyanobacteria biomass (%) CaCl2 Calcium have focused on the impacts of environmental parameters on carbon utilization and AMOPs growth during the two-step photosynthetic biogas upgrading process. Unlike AMOPs grown in a CO2 environment, only bicarbonate is employed as carbon source in these systems. Moreover, gas spargers may not be required in such AMOPs systems, thereby leading to different growth and carbon fixation behaviours for AMOPs. The cyanobacteria (also known as blue green microalgae) Spirulina platensis is tolerant to high alkalinity and high salinity conditions and is considered as a candidate for photosynthetic biogas upgrading (Xia et al., 2015). The effect of light transmission on microalgae growth is a primary point of concern (Carvalho et al., 2011; Cheng et al., 2013). Sun et al. (2018b) found that light attenuation reduced the light utilization efficiency of Chlorella vulgaris. It should be noted that filamentous S. platensis has completely different light absorption and scattering characteristics from spherical C. vulgaris. Nevertheless, there is also a lack of relevant research.
The innovation of this study is that it comprehensively investigates the effect of environmental parameters on the growth and bicarbonate transformation characteristics of halophilic and alkaliphilic S. platensis for photosynthetic biogas upgrading. The objectives are to investigate the light transmission characteristics in S. platensis culture, to assess the effects of light intensity and NaHCO3 concentration on bicarbonate utilization and cyanobacteria growth, and to discuss the design and application of the photosynthetic biogas upgrading system.

Light attenuation characteristics of cyanobacteria
The Lambert Beer's law was used to describe the light attenuation characteristics in the suspension of cyanobacteria cells , as shown in Equation 1.
where I represents the output light intensity, I0 denotes the incident light intensity, and l refers to the distance the light penetrates through cyanobacteria suspension. k0 is the total extinction coefficient, including the extinction coefficient kw caused by the liquid medium and the extinction coefficient kb caused by the cyanobacteria biomass, as shown in Equation where Cb represents the cyanobacteria biomass concentration.
To determine the coefficients kb and kw, the incident light intensity in the internal surface of the photobioreactor was measured by a light quantum meter, and was set to 120 μmol m -2 s -1 by adjusting the luminous power of a surface light source. The output light intensity of 42 different combinations were simultaneously tested: 7 different levels of Cb, from 0 to 0.706 g L -1 , and 6 levels of l, from 1 to 6 cm. After obtaining the corresponding total extinction coefficient k0 at different biomass concentrations, linear fitting of k0 and Cb was performed according to Equation 2. It was assumed that the incident light intensity was 80 and 210 μmol m -2 s -1 , and the cyanobacteria biomass concentration was 0.05, 0.3 and 0.8 g L -1 . Subsequently, the corresponding cyanobacteria light attenuation curve under each working condition was obtained from Equation 1.

Cell growth with different light intensity
Modified medium was used for the cultures to adjust the concentrations of NaHCO3, NaNO3, and Na2CO3 to 25.2, 1.5, and 0 g L −1 , respectively. Three light intensities (80, 150, 210 μmol m -2 s -1 ) were set to measure the growth and carbon fixation characteristics of cyanobacteria in the culture cycle (Kebede and Ahlgren, 1996;Chi et al., 2013;Bahr et al., 2014;Chi et al., 2014). All other cultivation parameters were as described in Section 2.1.

Cell growth with different concentrations of NaHCO3
The NaHCO3, NaNO3, and Na2CO3 concentration in the medium were also regulated; NaNO3 and Na2CO3 were set to 1.5 and 0 g L −1 , respectively, and NaHCO3 was set to 0.05, 0.1, 0.3, 0.6, or 1.0 M to measure the growth and carbon fixation characteristics of cyanobacteria in the culture cycle. The batch was cultivated under a light intensity of 210 μmol m −2 s −1 . All other cultivation parameters were as described in Section 2.1.

Cyanobacteria biomass concentration and bicarbonate utilization efficiency
During cultivation, a 10 mL cell suspension sample was filtered daily with a filter membrane with an aperture of 5 μm. The cell pellet was then washed with 10 mL of deionized water 4 times. After washing, the sample was dried in an oven at 85 °C overnight to obtain the biomass concentration gravimetrically.
Bicarbonate utilization efficiency ƞ (%) was defined as the percentage of fixed carbon in total input carbon (Kim et al., 2017), as shown in Equation 3.
where C (%) is the carbon content of cyanobacteria biomass, DCW (g L -1 ) stands for the biomass concentration, and TIC (g L -1 ) denotes the input total inorganic carbon concentration in the culture system.
where Ve (L) represents the solvent volume and ME (g) is the dry biomass.

Nitrate and carbon concentration in the culture medium
For the measurement of residual nitrate and carbon concentration in culture medium, 1 mL of cell suspension was sampled daily from the photobioreactor, and then it was filtered to obtain the supernatant. Then, nitrate concentration was measured according to the method of Collos et al. (1999). The total inorganic carbon (TIC) concentration in the sample was measured by a total organic carbon (TOC) analyser (Multi N/C 3000 analyser, Analytikjena, Germany). Thereafter, the concentrations of CO2, HCO3and CO3 2in the supernatant were calculated according to the Roy's method (Roy et al., 1993).

Light attenuation characteristics of Spirulina platensis
The total extinction coefficient k0 was obtained by measuring the output light intensity under different cyanobacteria biomass concentrations and distances from light incident surface, as shown in Table 1 As shown in Figure 2, when light entered the cyanobacteria suspension, the light attenuated exponentially along the direction of light transmission, resulting in uneven distribution of light intensity in the photobioreactor, which was not conducive to cyanobacteria growth. When the incident light intensity was 210 μmol m -2 s -1 and the distance from light incident surface was 1 cm, as biomass concentration increased from 0.05 to 0.8 g L -1 , the light intensity significantly decreased by 29.2% to 93.9%. The increasing distance and decreasing incident light intensity also led to the reduced output light intensity.
The causes of light attenuation mainly include light absorption by photosynthetic pigments in microalgae cells, light scattering by microalgae cells, and the mutual shielding between microalgae cells (Sun et al., 2016). Therefore, the type and content of pigment, concentration of microalgae cells, light intensity, and distance from light incident surface can all affect the light transmission in the photobioreactor. Interestingly, the extinction coefficient kb of S. platensis was 2.8 times higher than that of Chlorella vulgaris (Sun et al., 2018b). In other words, the light attenuation degree of S. platensis was much higher than that of Chlorella vulgaris. This is because the S. platensis cells are large and filamentous (5-10 μm in width

Effect of light intensity on growth and carbon fixation
The biomass concentration and residual NO3concentration in cyanobacteria suspension under different light intensities changed with cultivation time. As light intensity increased from 80 to 210 μmol m -2 s -1 , the maximum biomass concentration significantly increased from 0.67 to 0.98 g L -1 (Fig. 3a). Moreover, the maximum specific growth rate of 1.64 d -1 was observed at a light intensity of 210 μmol m -2 s -1 , which was 1.42 times higher than the minimum light intensity condition. Liu et al. (2014) found that compared to low light intensity conditions, the maximum biomass concentration of Chlorella sp. and Scenedesmus obliquus sp. reached 1.4 g L -1 with a light intensity of 6800 lux. The NO3concentration in cyanobacteria suspension decreased with increasing light intensity. On the 10 th day, the residual concentration of NO3at the light intensity of 210 μmol m -2 s -1 decreased by 43.7% from the initial stage, which was the largest decrease recorded (Fig. 3b). Bicarbonate utilization efficiency under the light intensity of 210 μmol m -2 s -1 was 1.5 times higher than the minimum light intensity condition. These results suggested that high light intensity could facilitate the growth and carbon fixation of cyanobacteria. However, the biomass concentration or nutrient absorption efficiency at the light intensity of 210 μmol m -2 s -1 was not significantly different from that of the light intensity of 150 μmol m -2 s -1 . Therefore, methods to increase cyanobacteria biomass could include not only increasing light intensity but also suppressing light attenuation, through methods such as periodic harvesting of cyanobacteria or development of a novel photobioreactor that can decrease the distance from light incident surface or alter light/dark cycle frequency (Ye et al., 2019).
S. platensis biomass can be used to extract high value-added products, such as pigments, including Chl-a, carotenoid, and phycocyanin (Lima et al., 2018). As shown in Figure 4a, the contents of Chl-a and carotenoid in cyanobacteria cells increased with increasing light intensity. PC is a blue-coloured accessory photosynthetic pigment and mainly captures light energy by absorbing light over a range of wavelengths (such as orange-yellow light) that chlorophyll uses insufficiently (Eriksen, 2008a). It has multiple applications in antioxidants, cosmetics, medicines, and health products production (Kissoudi et al., 2018). The PC concentration at the light intensity of 210 μmol m -2 s -1 was 0.89 g L -1 , which was 14.9% of the dry biomass. (Fig. 4b). There was no significant difference in the contents of Chl-a, carotenoid and phycobiliprotein under different light intensities, possibly because light intensity was not the primarily limiting factor of pigment accumulation at this time. del Rio-Chanona et al. (2015) found that the Spirulina biomass concentration in outdoor raceway ponds hardly exceeded 0.8 g L -1 , and the PC content was only approximately 7% of the dry biomass. Göksan and Kılıç (2009) and de Jesus et al. (2018) also found that lower light intensity was not conducive to biomass and PC accumulation of Spirulina. Therefore, the optimal light intensity of 210 μmol m -2 s -1 was selected for further trials.

Effect of NaHCO3 concentration on growth and carbon fixation
The biomass concentration of S. platensis changed dramatically at different NaHCO3 concentrations, as shown in Figure 5a. The results showed that the maximum biomass concentration of 1.46 g L -1 was observed on the 8 th day under an optimal growth condition of 0.1 M NaHCO3, which was 65.9% higher than 0.05 M NaHCO3. Compared with the cell growth of 0.1 M NaHCO3, partial inhibition occurred in 0.05, 0.3, and 0.6 M, and severe inhibition occurred in 1.0 M. On the 9 th day, the residual concentration of NO3at 0.1 M NaHCO3 decreased by 48.7% from the initial stage, which was the largest decrease recorded (Fig. 5b). The bicarbonate utilization efficiency decreased with increasing NaHCO3  concentration from 0.1 M to 0.6 M, and maximum bicarbonate utilization efficiency of 42.0% was achieved at 0.1 M NaHCO3, which was 7.8 times higher than at 0.6 M NaHCO3. This was because the carbon content of 0.05 M NaHCO3 was too low that limited cyanobacteria growth. Moreover, salinity can seriously affect osmotic pressure, nutrient absorption rate, and suspension characteristics of microalgae cells (Pisal et al., 2005;Razzak et al., 2013). High salinity may lead to reduced activity of various enzymes or transporters in cells and consumption of more energy for maintaining the osmotic pressure balance (Ho et al., 2014). The effect of NaHCO3 concentration on salinity in cyanobacteria suspension during cultivation was also investigated. The results showed that the salinity of the medium increased with the increase of NaHCO3 concentration. During the cultivation, the maximum salinity of cyanobacteria solution ranged from 5.03‰ to 36.70‰ when NaHCO3 concentration ranged from 0.05 M to 1.0 M, respectively. Therefore, high NaHCO3 concentration led to a serious salt stress on cyanobacteria cells. Additionally, the pH of 0.1 M NaHCO3 was too high (close to 12) at the late stage of cyanobacteria cultivation, which inhibited the growth of cyanobacteria cells, resulting in an obvious reduction in biomass concentration compared with the maximum value.
The changes in TIC (including HCO3 -, CO3 2-, and CO2) concentration in cyanobacteria suspension with cultivation time could also indirectly explain the growth and carbon fixation characteristics of cyanobacteria. Compared to the initial stage, as NaHCO3 concentration increased from 0.1 to 1.0 M, the TIC concentration on the 9 th day significantly decreased by 62.0% to 19.7%. Additionally, the reduced TIC concentration at 0.1 M NaHCO3 was 1.3 times higher than that of 0.05 M NaHCO3 (Fig. 6). It could be observed that carbonate was effectively regenerated during the growth and metabolism of cyanobacteria, to reduce the energy consumption of absorbent regeneration and operating cost (Chi et al., 2011). The HCO3content on the 2 nd day was 29.5% and 57.1% of the remaining TIC content at 0.05 M and 0.1 M NaHCO3, respectively, while it was reduced to 2.2% and 1.1% on 5 th day. The results showed that the carbonate regeneration rate was high under 0.05 M and 0.1 M NaHCO3. Nevertheless, 0.05 M NaHCO3 was seriously insufficient for cyanobacteria growth.
As shown in Figure 7a, the maximum Chl-a concentration of 20.48 mg L -1 was achieved at 0.1 M NaHCO3, which was 4.4 times higher than that at 0.05 M NaHCO3. The carotenoid concentration at 0.1 M NaHCO3 was 1.6 times higher than at 0.05 M NaHCO3. Bicarbonate significantly changed the biomass and various biochemical components such as pigments in the microalgae cells. These findings were in line with those of the previous studies (Chi et    Moreover, the maximum PC concentration of 0.91 g L -1 was achieved at 0.3 M NaHCO3, which was 3.2 times higher than at 0.05 M NaHCO3. The PC content of 0.3 M NaHCO3 accounted for 13.7% of dry biomass, which was 2.6 times higher than that of 0.1 M NaHCO3 (Fig. 7b). Studies found that unfavourable environments lead to the decomposition of phycobiliprotein (especially PC) to provide nutrients needed for cyanobacteria growth, thus maintaining the normal metabolic function of cyanobacteria (Eriksen, 2008b). Compared with the maximum biomass concentration at 0.1 M NaHCO3, the value on the 9 th day decreased by 11.5% due to the high pH of the later growth stage. Therefore, the phycobiliprotein content of 0.1 M NaHCO3 was much lower. Compared to freshwater microalgae species, these results demonstrated that S. platensis had a favourable tolerance to high saline-alkaline environments while possessing promising capabilities to regenerate carbonate.

Design and application for photosynthetic biogas upgrading system
According to the experimental results of cyanobacteria cultivation, a case study was carried out to set up a photosynthetic biogas upgrading system. The biogas production of this project to treat cow manure and straw was set at 800 m 3 h -1 and S. platensis was selected as the working cyanobacteria species. Since after capturing the CO2 of biogas, the 0.05 M Na2CO3 absorbent was conducive to cyanobacteria growth, the average biomass productivity of S. platensis was recorded at 0.2 g L -1 d -1 under the condition of 0.1 M NaHCO3.
Moreover, producing 1 t of AMOPs biomass fixes approximately 1.83 tons of CO2 (Chisti, 2007). Accordingly, the amount of total biomass produced was 344 kg h -1 when the cyanobacteria fixed all the CO2 contained in the produced biogas. Therefore, the cultivation volume of cyanobacteria in the column photobioreactor was 41280 m 3 . The low calorific value of biogas with 60%-65% CH4 is 20-25 MJ m -3 (Angelidaki et al., 2018). According to Xia et al. (2014), the calorific value of S. platensis was calculated to be 19.15 kJ g -1 VS. Given the fact that the volatile solids content accounts for 85% of the microalgae total solids (Sun et al., 2018a), the total calorific value of microalgae could reach 5591 MJ h -1 , which is equivalent to 280 m 3 h -1 of biogas. This made up one third of the system's biogas production, indicating that the photosynthetic biogas upgrading had a promising development potential.

Conclusions
The halophilic and alkaliphilic S. platensis was demonstrated to be a potential candidate for sustainable carbon capture and biogas upgrading. Severe light attenuation was observed in filamentous S. platensis culture; the extinction coefficient in S. platensis was 2.8 times higher than that of spherical C. vulgaris. The maximum biomass concentration was 0.98 g L -1 at the light intensity of 210 μmol m -2 s -1 , which was 47.6% higher than that of 80 μmol m -2 s -1 . NaHCO3 concentration significantly affected cyanobacteria biomass and pigment production. The optimal biomass concentration of 1.46 g L -1 was achieved at an NaHCO3 concentration of 0.1 M, which was 65.9% higher than that of 0.05 M NaHCO3. Meanwhile, the bicarbonate utilization efficiency reached 42.0%. Cyanobacteria biomass of up to 344 kg h -1 could be generated with the biogas flow rate of 800 m 3 h -1 , corresponding to an energy value of 5591 MJ h -1 (i.e., that of one third of the biogas produced).