The Effect of Light Wavelength on CO2 Capture, Biomass Production and Nutrient Uptake by Green Microalgae: A Step Forward on Process Integration and Optimisation

Microalgae have drawn the attention of several researchers as an alternative to the traditional physicochemical CO2 capture methods, since they can convert CO2 and water into organic matter and release oxygen into the atmosphere. Microalgal growth can be improved by changing light supply, such as light intensity, wavelength, and photoperiod. In this study, the effect of different light wavelengths on CO2 capture, nutrient removal from a synthetic effluent and biomass production of Chlorella vulgaris, Tetradesmus obliquus and Neochloris oleoabundans was studied. The experiments were conducted with light-emitting diodes (LEDs) with different wavelengths: 380–750 nm (white), 620–750 nm (red) and 450–495 nm (blue). The maximum specific growth rate was obtained by N. oleoabundans with white LEDs (0.264 ± 0.005 d−1), whereas the maximum biomass productivity (14 ± 4 mgdw L−1 d−1) and CO2 fixation rate (11.4 mgCO2 L−1 d−1) were obtained by C. vulgaris (also with white LEDs). Nitrogen and phosphorus removal efficiencies obtained under white light conditions were also the highest for the three studied microalgae.


Introduction
Since the pre-industrial period, the emissions of carbon dioxide (CO 2 ) have been rapidly increasing due to anthropogenic activities, mainly the combustion of fossil fuels [1]. The increase of the atmospheric CO 2 concentration can lead to ocean acidification and the intensification of the greenhouse effect, resulting in various negative impacts, such as [2][3][4]: (i) the increase of the global average temperature; (ii) the melting of polar ice; and (iii) the rise of sea levels. To address this environmental problem, the scientific community has been exploring diverse options to effectively capture CO 2 from the atmosphere or directly from emission sources.
Currently, the most used CO 2 capture technologies include physical adsorption, chemical absorption, membrane separation and cryogenic fractionation [5]. Physical adsorption uses a solid adsorbent to separate and capture CO 2 from flue gases. The main limitations of this technology are: (i) the necessity of a pre-treatment for the flue gas; (ii) the low adsorption capacity of the adsorbents; sources differ from each other in the spectrum, wavelength distribution, energy consumption and cost. Based on these characteristics, the most commonly used light sources for microalgal production are fluorescent lamps and LEDs. The use of LEDs as light source allows better control of light (when compared with fluorescent lamps) and the use of different wavelengths, which can be favourable for biomass production [16]. The light wavelength used in microalgal growth can induce different effects. Red light can promote higher growth rates with smaller cells and low nutrient uptake; blue light affects gene expression and some metabolic pathways of microalgae, triggering a high nutrient uptake, but inducing lower growth rates with larger cells. Green microalgae cannot use yellow and green light effectively due to the lack of phycobilins [17]. Taking into account the important role of light source on microalgal biomass production, this study aimed to evaluate the effect of different light wavelengths (white, 380-750 nm; red, 620-750 nm; and blue, 450-495 nm) on the growth, biomass productivity, and therefore CO 2 capture, of three green microalgae: Chlorella vulgaris, Tetradesmus obliquus and Neochloris oleoabundans. Considering the ability of microalgae to remove nutrients, nitrogen and phosphorus removal efficiencies were also evaluated in the studied conditions, to assess the possibility of process integration (microalgal biomass production with CO 2 and nutrients uptake) and, consequently, production cost reduction.

Microorganisms and Culture Medium
The microalgae C. vulgaris CCAP 211/11B and T. obliquus CCAP 276/34 were obtained from the Culture Collection of Algae and Protozoa (CCAP, Scotland, UK) and N. oleoabundans UTEX 1185 was obtained from the University of Texas Culture Collection of Algae (UTEX, Texas, USA). The microalgae were inoculated in modified OECD (Organization for Economic Co-operation and Development) test medium, with the same composition as the one reported by Gonçalves, et al. [18]. Microalgal stock solutions were prepared in 100-mL Erlenmeyer flasks (50 mL working volume) using the above-referred culture medium under aseptic conditions. Erlenmeyer flasks were maintained at a constant temperature of 25 • C under continuous light supply with an intensity of approximately 6.50 µmol m −2 s −1 and agitation was promoted by an orbital shaker (Unimax 1010, Heidolph, Germany) set at 100 rpm (rotations per minute).

Experimental Setup and Cultivation Conditions
Batch experiments using the above-referred culture medium were performed for 12 days in 1000-mL flasks with a working volume of 900 mL. The average temperature during the experiences was 21 ± 2 • C. Agitation of the cultures was promoted by the injection of atmospheric air (previously filtered through 0.22-µm cellulose acetate membrane filters). Light was continuously supplied (24:0 light:dark ratio) with a light intensity of 8 ± 1 µmol m −2 s −1 . The experiments were carried out in an installation with LEDs (see Figure 1) set at different light wavelengths: (i) white (W, 380-750 nm); (ii) red (R, 620-750 nm); and (iii) blue (B, 450-495 nm). These wavelengths were selected because red light is related to increased biomass growth and blue light is commonly associated with higher nutrient consumption [16]. The surface where the cultures were located was at a distance of 49.0 cm from the light source. The initial biomass concentrations were: (i) 55 ± 3 mg dw L −1 for C. vulgaris; (ii) 33 ± 1 mg dw L −1 for T. obliquus; and (iii) 19 ± 1 mg dw L −1 for N. oleoabundans. The experiments were performed in duplicates.

Growth Monitoring and Kinetic Growth Parameters
Samples were collected daily to evaluate microalgal growth by optical density at 680 nm (OD680) measurements using a UV-6300PC Double Beam spectrophotometer (VWR, Amadora, Portugal). Biomass concentration was obtained indirectly from calibration curves that relate biomass concentration in terms of dry weight (dw) with OD680. For this, a known volume of microalgal suspension at different concentrations (5 mL for C. vulgaris and 10 mL for T. obliquus and N. oleoabundans) was vacuum filtered using 0.45-µm cellulose acetate membranes previously dried at 60 °C for 24 h. After filtration, the cell-containing membranes were dried again at 105 °C until they reached a constant mass. The biomass concentration in terms of dry weight (dw) was obtained by the difference between the membrane mass divided by the filtered volume. Each microalgal suspension was also evaluated in terms of absorbance at 680 nm (OD680). The calibration curves were then established through a linear regression between OD680 and biomass concentrations (in mgdw L −1 ) (see Table 1).
With biomass concentration values, kinetic growth parameters, such as the specific growth rate and average biomass productivity, were determined. Specific growth rate (μ in d −1 ) for each experiment was determined according to Equation (1): where and represent the biomass concentration (in mgdw L −1 ) at the final ( ) and initial moments ( ) of the exponential growth phase. Average biomass productivities ( in mgdw L −1 d −1 ) were calculated as shown in Equation (2): where corresponds to the biomass concentration (in mgdw L −1 ) at the end of the cultivation time ( , d) and corresponds to the initial biomass concentration (in mgdw L −1 ) at the beginning of the cultivation time ( , d).

Carbon Dioxide Fixation Rate
In the last day of the experiments, samples of 200 mL were collected, centrifuged and washed twice with distilled water to determine the elemental composition of the produced biomass. After this process, biomass was dried at 70 °C for 48 h. Elemental analysis of microalgal biomass was carried out based on the method described by Rocha, et al. [19]. The average carbon fixation rates (RC in mgCO2 L −1 d −1 ) were calculated according to Equation (3): (1)

Growth Monitoring and Kinetic Growth Parameters
Samples were collected daily to evaluate microalgal growth by optical density at 680 nm (OD 680 ) measurements using a UV-6300PC Double Beam spectrophotometer (VWR, Amadora, Portugal). Biomass concentration was obtained indirectly from calibration curves that relate biomass concentration in terms of dry weight (dw) with OD 680 . For this, a known volume of microalgal suspension at different concentrations (5 mL for C. vulgaris and 10 mL for T. obliquus and N. oleoabundans) was vacuum filtered using 0.45-µm cellulose acetate membranes previously dried at 60 • C for 24 h. After filtration, the cell-containing membranes were dried again at 105 • C until they reached a constant mass. The biomass concentration in terms of dry weight (dw) was obtained by the difference between the membrane mass divided by the filtered volume. Each microalgal suspension was also evaluated in terms of absorbance at 680 nm (OD 680 ). The calibration curves were then established through a linear regression between OD 680 and biomass concentrations (in mg dw L −1 ) (see Table 1). With biomass concentration values, kinetic growth parameters, such as the specific growth rate and average biomass productivity, were determined. Specific growth rate (µ in d −1 ) for each experiment was determined according to Equation (1): where X 1 and X 0 represent the biomass concentration (in mg dw L −1 ) at the final (t 1 ) and initial moments (t 0 ) of the exponential growth phase. Average biomass productivities (P in mg dw L −1 d −1 ) were calculated as shown in Equation (2): where X f corresponds to the biomass concentration (in mg dw L −1 ) at the end of the cultivation time (t f , d) and X i corresponds to the initial biomass concentration (in mg dw L −1 ) at the beginning of the cultivation time (t i , d).

Carbon Dioxide Fixation Rate
In the last day of the experiments, samples of 200 mL were collected, centrifuged and washed twice with distilled water to determine the elemental composition of the produced biomass. After this process, biomass was dried at 70 • C for 48 h. Elemental analysis of microalgal biomass was carried out based on the method described by Rocha, et al. [19]. The average carbon fixation rates (R C in mg CO2 L −1 d −1 ) were calculated according to Equation (3): where C C is the carbon mass fraction in biomass (in % wt.), P is the average biomass productivity (in mg dw L −1 d −1 ), M CO2 is the molecular weight of CO 2 (in g mol −1 ) and M C is the molecular weight of C (in g mol −1 ).

Nutrients Removal
To analyse nutrients removal, samples were collected on days 0, 1, 2, 4, 7, 9 and 11 of each experiment. The samples were centrifuged (Himac CT6E Centrifuge, VWR, Amadora, Portugal) at 4000 rpm for 10 min and then filtered with 0.45-µm nylon membrane syringe filters (Specanalitica, Cascais, Portugal). Nutrient concentrations (PO 4 3− , and NO 3 − ) in the filtered samples were analysed by ion chromatography (ICS-2100, Dionex, VWR, Amadora, Portugal) using an AS9-HC column and the ASRS ® 300 suppressor. Evaluation of nutrients concentrations within cultivation time allowed the determination of nutrient removal efficiencies (RE, %) and nutrient uptake rates by the selected microalgae under the studied light conditions. Nutrients removal efficiencies were determined through Equation (4): where S i and S f represent nutrients concentration (in mg L −1 ) in the beginning and at the end of the cultivation time, respectively. Nutrient uptake rates were obtained by fitting the modified Gompertz model [20] (Equation (5)) to the experimental data (corresponding to the time-course evolution of nutrients concentration): where S(t) is the time-course evolution of nutrients concentration, k is the removal rate (in d −1 ) and λ is the lag time (in d). The kinetic parameters were obtained by minimising the sum of squared residuals using the Solver supplement of Microsoft Excel 2016. The quality of the model fits was assessed by calculating the coefficient of determination (R 2 ) and the root mean squared error (RMSE).

Statistical Analysis
The average and standard deviation were calculated for each parameter. The statistical significance of the results was assessed using the Student's paired t-test to verify if the differences between the studied microalgae and conditions could be considered significant. Statistical tests were performed at a significance level of 0.05.

Microalgal Growth
Phototrophic growth is closely related to the light quality used, with light wavelength being a factor that has a great influence on microalgal growth. Through the monitoring of biomass concentration, it was possible to analyse and characterise microalgal growth kinetics. Figure 2 presents the growth curves obtained for C. vulgaris, T. obliquus and N. oleoabundans grown under different light conditions (white, red and blue LEDs). All the species grew under the different light conditions, except N. oleoabundans, which did not grow when cultivated under blue LED. This observation may be related to the low light intensity values used in this study and to the light supply in a narrower range of the spectrum, as well as to the different responses of different microalgae to light conditions. In this case, N. oleoabundans may be more susceptible to light limitation when growing under low light intensity and in this range of the light spectrum. Higher growth was obtained when microalgae were cultured under white LED. Comparing red and blue LEDs, it was found that the growth of C. vulgaris and N. oleoabundans was favoured under red LED conditions. In the case of T. obliquus, it was observed that its growth was slightly higher in blue LED assays, being similar to the one obtained with white LED. From the growth curves present in Figure 2, it was also possible to infer the adaptation phase of each microalga to the different light conditions. This analysis shows that N. oleoabundans presented the longest adaptation phase (≈2 d) compared to the other studied species (≈0 d), for all tested light conditions. Table 2 presents the kinetic growth parameters determined for the selected species in each of the studied light conditions. The values of specific growth rates ranged between 0.0657 ± 0.0008 d −1 (for C. vulgaris with blue LED) and 0.264 ± 0.005 d −1 (for N. oleoabundans with white LED). C. vulgaris obtained the highest specific growth rate with red LED (0.090 ± 0.007 d −1 ) and the lowest with blue LED (0.0657 ± 0.0008 d −1 ). Results in the same order of magnitude were obtained by Mohsenpour, et al. [21]: when cultivating C. vulgaris under red light at a light intensity of 250 µmol m −2 s −1 , the authors reported a specific growth rate of 0.07 d −1 . Regarding the effect of different light wavelengths on microalgal growth, Li, et al. [22] also concluded that blue LED was not favourable for C. vulgaris growth. The higher value obtained with red light can be related to the fact that this wavelength can promote an acceleration of the cell cycle, resulting in higher growth rates and smaller cells. This occurs because the main photosynthetic pigment absorbs light at approximately 600-700 nm, thus improving photosynthetic efficiency under this range of wavelengths (white and red LEDs). On the other hand, cultivation under blue light can lead to slower growth with the formation of larger cells [23]. In the case of T. obliquus, the highest specific growth rate was obtained with white LED (0.12 ± 0.02 d −1 ), whereas the lowest value was obtained with red LED (0.08 ± 0.01 d −1 ). The values obtained are in line with the range of results presented by Gonçalves, et al. [24] for white, red and blue LEDs (0.07, 0.08 and 0.17 d −1 , respectively). However, the specific growth rate values obtained in the present study followed a different trend: white LED > blue LED > red LED. The specific growth rate obtained by N. oleoabundans with white LED (0.264 ± 0.005 d −1 ) was the highest. This value was statistically higher (p < 0.05) than the specific growth rates obtained for the other studied microalgae grown in the same light conditions (white LED). Using red LED, the growth did not exceed 0.17 ± 0.03 d −1 . Under blue LED, this microalga was unable to grow, and it was not possible to determine the specific growth rate in these conditions. In the study performed by Zhao, et al. [25], this microalga followed the same trend as the present study, though the specific growth rate value was lower under white LED (0.216 d −1 ) and similar for red LED (0.179 d −1 ). However, in this study N. oleoabundans was able to grow under blue LED, with a specific growth rate of 0.134 d −1 , which may be related to the lower light intensities provided to the cultures in the present work (8 ± 1 µmol m −2 s −1 , compared with the 200 µmol m −2 s −1 reported in the above-mentioned study). Regarding the maximum biomass concentration, the highest values were obtained with white LED for all microalgal species, ranging from 79 ± 9 mg dw L −1 (for T. obliquus) to 128 ± 5 mg dw L −1 (for C. vulgaris). Concerning the red LED assays, the highest value of maximum biomass concentration was obtained by C. vulgaris (116 ± 2 mg dw L −1 ), being this value statistically higher (p < 0.05) than those obtained by the other studied microalgae (66 ± 2 and 73 ± 8 mg dw L −1 for T. obliquus and N. oleoabundans, respectively). For blue LED conditions, C. vulgaris was the microalga that obtained the highest value of maximum biomass concentration (111 ± 2 mg dw L −1 ) and N. oleoabundans the lowest (34 ± 4 mg dw L −1 ). In terms of maximum biomass productivity values, achieved values are in agreement with the values of the specific growth rate. For all species, maximum biomass productivity values followed the trend: white LED > red LED > blue LED. These results are in accordance with previous studies that reported a similar behaviour of these microalgae in this range of light wavelengths [25]. The highest maximum biomass productivity values were obtained by C. vulgaris with white LED (14 ± 4 mg dw L −1 d −1 ) and the lowest was obtained by T. obliquus with blue LED (5.9 ± 0.2 mg dw L −1 d −1 ). The values obtained are below the values reported by Gonçalves, et al. [24] for the cultivation of Tetradesmus sp. with light intensity of 13 µmol m −2 s −1 using white LED (70 mg dw L −1 d −1 ), red LED (90 mg dw L −1 d −1 ) and blue LED (270 mg dw L −1 d −1 ). In the case of average biomass productivity values, the same behaviour as the other kinetic growth parameters can be observed, namely the values obtained for the assays performed under white LED were higher than the ones achieved in the other light conditions. Values obtained in these light conditions ranged from 4 ± 1 mg dw L −1 d −1 (for T. obliquus) to 6.8 ± 0.5 mg dw L −1 d −1 (for C. vulgaris). Assunção et al. [26] obtained an average productivity of 140 mg dw L −1 d −1 with T. obliquus, using a light intensity of 74 µmol m −2 s −1 . These values are significantly higher than those obtained in the present study, possibly due to the higher light intensity used in their study. According to Table 2, white LED promotes increased growth rates and productivities of C. vulgaris and N. oleoabundans, while the use of blue LED results in the lowest microalgal growth. These results might be related to the emission spectra of the studied light sources and the absorption spectra of the main pigments present in the studied microalgae. In the case of green microalgae, the main photosynthetic pigments are chlorophylls, which are the most important players in light-harvesting. Considering that these pigments absorb light at approximately 600-700 nm, the light sources that promote improved light-harvesting and, hence, improved photosynthetic efficiency, are those emitting in this range of wavelengths, namely white and red LEDs [27,28]. T. obliquus achieved the highest values for all kinetic growth parameters with white LED and the lowest values with red LED. These results show that the effects of light quality on the growth of microalgae are different, depending on the studied microalgal species. values, the same behaviour as the other kinetic growth parameters can be observed, namely the values obtained for the assays performed under white LED were higher than the ones achieved in the other light conditions. Values obtained in these light conditions ranged from 4 ± 1 mgdw L −1 d −1 (for T. obliquus) to 6.8 ± 0.5 mgdw L −1 d −1 (for C. vulgaris). Assunção et al. [26] obtained an average productivity of 140 mgdw L −1 d −1 with T. obliquus, using a light intensity of 74 µmol m −2 s −1 . These values are significantly higher than those obtained in the present study, possibly due to the higher light intensity used in their study. According to Table 2, white LED promotes increased growth rates and productivities of C. vulgaris and N. oleoabundans, while the use of blue LED results in the lowest microalgal growth. These results might be related to the emission spectra of the studied light sources and the absorption spectra of the main pigments present in the studied microalgae. In the case of green microalgae, the main photosynthetic pigments are chlorophylls, which are the most important players in light-harvesting. Considering that these pigments absorb light at approximately 600-700 nm, the light sources that promote improved light-harvesting and, hence, improved photosynthetic efficiency, are those emitting in this range of wavelengths, namely white and red LEDs [27,28]. T. obliquus achieved the highest values for all kinetic growth parameters with white LED and the lowest values with red LED. These results show that the effects of light quality on the growth of microalgae are different, depending on the studied microalgal species.  ; ND-value not determined; X max -maximum biomass concentration (mg dw L −1 ); P max -maximum biomass productivity (mg dw L −1 d −1 ); P-average biomass productivity (mg dw L −1 d −1 ).

Carbon Dioxide Fixation Rates
Assuming that all CO 2 captured is converted into biomass, CO 2 fixation rates were determined based on the carbon content in biomass and the average productivities. Figure 3 presents the CO 2 fixation rates obtained for C. vulgaris, T. obliquus and N. oleoabundans cultivated under different light conditions. The values of CO 2 fixation rate ranged from 5.1 ± 0.2 (for T. obliquus with red LED) to 11.4 ± 0.5 mg CO2 L −1 d −1 (for C. vulgaris with white LED). As expected, the CO 2 fixation rates were higher in the experiments with white LED, since high biomass productivities were achieved. For C. vulgaris and N. oleoabundans, the highest CO 2 fixation rates were obtained with white LED (11.4 ± 0.5 and 9.2 ± 0.5 mg CO2 L −1 d −1 , respectively) and the lowest with blue LED (7.11 ± 0.07 mg CO2 L −1 d −1 for C. vulgaris). Similarly, for T. obliquus the highest value was obtained with white LED (7.4 ± 1.2 mg CO2 L −1 d −1 ). However, the lowest value was obtained with red LED (5.1 ± 0.2 mg CO2 L −1 d −1 ). When comparing the results obtained for the studied microalgal species, it is possible to observe that T. obliquus presented the lowest values in all LED conditions. At the end of the experiments, the highest values of CO 2 fixed by all species were achieved with white LED: 126 mg CO2 for C. vulgaris, 81.4 mg CO2 for T. obliquus and 101 mg CO2 for N. oleoabundans. Ho, et al. [29] reported CO 2 fixation rates between 374 and 745 mg CO2 L −1 d −1 using different T. obliquus strains. These results are significantly higher than the ones obtained in the present study due to the higher light intensity used in the above-mentioned study (140 µmol m −2 s −1 ) and also to the aeration of microalgal cultures with a CO 2 -enriched stream (2.5% v/v). Chaudhary, et al. [30] also reported a CO 2 fixation rate of 141 mg CO2 L −1 d −1 for C. vulgaris and 130 mg CO2 L −1 d −1 for T. obliquus, both species fed with a 5% (v/v) CO 2 stream and grown under cool white fluorescent light. The values obtained by N. oleoabundans were also below those reported by Razzak [31]: 80 mg CO2 L −1 d −1 with a light intensity of 65 µmol m −2 s −1 . Besides the determination of carbon content in microalgal cells, the elemental analysis of the produced biomass allowed the evaluation of the effect of light wavelength on the chemical composition of microalgae and also the assessment of the different nutritional needs in each of the studied light conditions. Mass percentages of carbon (C), hydrogen (H), nitrogen (N) and sulphur (S) determined in the studied microalgae in the evaluated light conditions are presented in Table 3. Regarding C. vulgaris, determined C and H contents were higher in biomass produced with white LED (46 ± 2 wt.% and 6.8 ± 0.3 wt.%) and lower in biomass resulting from cultures grown under blue LED (43 ± 1 wt.% and 6.5 ± 0.2 wt.%). In the case of biomass produced with red LED, C and H mass fractions were not statistically different (p > 0.05) from those obtained with white LED. On the other hand, N and S contents present in microalgal biomass grown under blue LED (8.0 ± 0.3 wt.% and 0.6 ± 0.1 wt.%, respectively) were higher than those obtained under red LED conditions (7.5 ± 0.1 wt.% and 0.4 ± 0.1 wt.%, respectively), these values being statistically different (p < 0.05). The lower C content obtained by C. vulgaris with blue LED can be related to the fact that blue light releases high energy photons that induce endogenous respiration, which is responsible for the breakdown of carbohydrates reserves and their conversion into CO2 that is further released to the atmosphere. Accordingly, this phenomenon results in a great loss of carbohydrate reserves, which is reflected in the low carbon contents determined in biomass resulting from blue LED experiments [32]. At the same time, this microalga obtained the highest values of N content with the same light condition (blue light), due to a light stress caused by this light wavelength that triggers the accumulation of photo-protective pigments and induces an increase in protein content (which is closely linked with N content) [17,33]. Phukan, et al. [34] reported C, H and N contents in C. vulgaris sp. of 47.5, 7.1 and 6.7 wt.%, respectively, whose values are very similar to those obtained in the present study with Besides the determination of carbon content in microalgal cells, the elemental analysis of the produced biomass allowed the evaluation of the effect of light wavelength on the chemical composition of microalgae and also the assessment of the different nutritional needs in each of the studied light conditions. Mass percentages of carbon (C), hydrogen (H), nitrogen (N) and sulphur (S) determined in the studied microalgae in the evaluated light conditions are presented in Table 3. Regarding C. vulgaris, determined C and H contents were higher in biomass produced with white LED (46 ± 2 wt.% and 6.8 ± 0.3 wt.%) and lower in biomass resulting from cultures grown under blue LED (43 ± 1 wt.% and 6.5 ± 0.2 wt.%). In the case of biomass produced with red LED, C and H mass fractions were not statistically different (p > 0.05) from those obtained with white LED. On the other hand, N and S contents present in microalgal biomass grown under blue LED (8.0 ± 0.3 wt.% and 0.6 ± 0.1 wt.%, respectively) were higher than those obtained under red LED conditions (7.5 ± 0.1 wt.% and 0.4 ± 0.1 wt.%, respectively), these values being statistically different (p < 0.05). The lower C content obtained by C. vulgaris with blue LED can be related to the fact that blue light releases high energy photons that induce endogenous respiration, which is responsible for the breakdown of carbohydrates reserves and their conversion into CO 2 that is further released to the atmosphere. Accordingly, this phenomenon results in a great loss of carbohydrate reserves, which is reflected in the low carbon contents determined in biomass resulting from blue LED experiments [32]. At the same time, this microalga obtained the highest values of N content with the same light condition (blue light), due to a light stress caused by this light wavelength that triggers the accumulation of photo-protective pigments and induces an increase in protein content (which is closely linked with N content) [17,33]. Phukan, et al. [34] reported C, H and N contents in C. vulgaris sp. of 47.5, 7.1 and 6.7 wt.%, respectively, whose values are very similar to those obtained in the present study with white LED. C, H, N and S contents obtained by T. obliquus followed the trend: white LED > red LED > blue LED. In the case of N. oleoabundans, the highest C, H and N contents were obtained in biomass grown with white LED (46.0 ± 0.6, 6.7 ± 0.1, 7.9 ± 0.1 wt.%, respectively), being statistically higher (p < 0.05) than those obtained with red LED conditions. Tibbetts, et al. [35] reported that the proximate nitrogen content of N. oleoabundans was 6.3 wt.%, being this value lower than the one obtained in the present study. Taking into account the mass fractions of each analysed element, the molecular formula (CO a H b N c P d S e ) of the biomass for each microalga was determined (a and d were not determined):

Nutrient Removal
The European Union (EU) has set limits for the concentration of nutrients (nitrogen and phosphorus) in effluents, as well as minimum percentages of load reduction [36,37]. According to these Directives, the limits for effluent discharge are the following: (i) 10-15 mg N L −1 for nitrogen, with a minimum reduction percentage of 70%-80%; and (ii) 1-2 mg P L −1 for phosphorus, with a minimum reduction of 80%.
To evaluate the nutrient (nitrogen and phosphorus) removal capacity of the studied microalgae under the range of light wavelengths evaluated, nitrogen and phosphorus removal efficiencies and kinetics were determined. The results are presented in Table 4. Values of nitrogen removal efficiency ranged between 17 ± 1% (for N. oleoabundans grown with red LED) and 36 ± 1% (for C. vulgaris grown in the same light conditions). On the other hand, phosphorus removal efficiency values ranged between 7 ± 5% (for C. vulgaris grown with blue LED) and 20 ± 9% (for C. vulgaris grown with white LED). Regarding the potential of the studied microalgae for the uptake of these nutrients, it was observed that C. vulgaris was the microalga presenting the highest nitrogen and phosphorus removal efficiencies with values statistically higher (p < 0.05) than those obtained for the other studied microalgae. On the other hand, N. oleoabundans was the microalga achieving the lowest efficiency results. In terms of light wavelength, it was observed that the lowest values of nitrogen and phosphorus removal efficiency were obtained in cultures performed with blue LED. At the end of the experiments, the values of nitrogen concentration ranged from 23.0 to 36.4 mg N L −1 and the values of phosphorus concentration ranged between 8.9 and 12.7 mg P L −1 , values above the limits defined by EU legislation. The low biomass productivity and specific growth rates and, consequently, the low nutrients removal efficiencies obtained in this study can be attributed to light limitation, as the light intensity supplied to the cultures did not exceed 8 ± 1 µmol m −2 s −1 . Kim, et al. [38] cultivated T. obliquus under white, red and blue LEDs and the amount of nitrogen and phosphorus removed varied depending on the light used for microalgal growth, wherein the highest amounts of nitrogen and phosphorus removed were obtained with blue LED, followed by red LED and, finally, white LED. The same behaviour was obtained in the present study for nitrogen removal. Microalgae consumed more nitrates with blue light than with red light, because when growing with blue light, microalgae are more dependent on the use of nutrients for their growth. At this wavelength, the activation of nitrate and phosphorus reductases occurs, leading to an increase in the absorption of these nutrients [16,38]. The modified Gompertz model was used to determine the lag time (λ) and the uptake rate (k) for nitrogen and phosphorus. Analysing the performance parameters R 2 and RMSE, obtained coefficients of determination are close to one (R 2 ≥ 0.934) and RMSE values are low, showing the suitability of the modified Gompertz model to describe the experimental data. The highest values of nitrogen and phosphorus removal rate were obtained by T. obliquus with blue (0.397 d −1 ) and white (0.517 d −1 ) LEDs, respectively. Based on these results, it can be concluded that T. obliquus was able to uptake nitrogen and phosphorus faster than the other studied microalgae. Regarding nitrogen kinetic parameters, the lag time was higher when the cultures grew under blue LED.

Integrated Evaluation of the Effect of Light Wavelength on Microalgal Growth and Carbon and Nutrient Uptake
The results obtained in this study showed that light wavelength influenced microalgal growth, CO 2 and nutrient uptake and elemental composition of biomass in different ways. In addition, changes in the light wavelength resulted in different responses from the studied microalgae, which demonstrates a species-specific behaviour. Table 5 summarises how light wavelength influenced microalgal growth rates, biomass productivities, CO 2 uptake rates, nitrogen and phosphorus uptake and C, H, N and S contents of the studied microalgae. In terms of microalgal growth and CO 2 uptake rates, in general, higher values were obtained in cultures performed with white LED, rather than with red and blue LEDs. Considering nitrogen and phosphorus uptake, higher efficiencies were also obtained for cultures grown with white LED, except in the case of phosphorus uptake by T. obliquus, which was higher in red LED conditions. Finally, C, H, N and S contents varied according to the light wavelength used, but also according to the microalgal species. These results are very important, as they give important information on how to select the best operational conditions (in terms of light wavelength), depending on the microalgal species in use and on the purpose of microalgal cultivation. Table 5. Summary of the influence of light wavelengths on studied microalgae.

Microalgae
Light Effects

C. vulgaris White
High biomass productivities; High C, H and N contents; High CO 2 uptake rates; High N and P uptake.
Red High specific growth rates; High C and H contents.

Blue
High N and S contents.

T. obliquus White
High specific growth rates and biomass productivities; High C, H, N and S contents; High CO 2 uptake rates; High N uptake.
Red High P uptake.

N. oleoabundans White
High specific growth rates and biomass productivities; High C, H, N and S contents; High CO 2 uptake rates; High N and P uptake.

Conclusions
Microalgal kinetic growth parameters, such as the specific growth rate and biomass productivity, were evaluated using different light wavelengths. The results showed white LED to be the light source promoting higher growth, except for C. vulgaris, which presented higher growth with red LED. This observation may be a result of an acceleration of the cell cycle in these light conditions, which can lead to an increase in specific growth rates. Regarding CO 2 fixation rate values, these were higher in the experiments performed under white LED, with values ranging from 7.4 ± 1.2 mg CO2 L −1 d −1 (for T. obliquus) to 11.4 ± 0.5 mg CO2 L −1 d −1 (for C. vulgaris). The removal of nutrients from the culture medium was evaluated in terms of removal efficiencies and uptake kinetics. Nitrogen removal efficiencies obtained under white LED were the highest for the three microalgae, ranging from 21 ± 2% to 36 ± 1%. At this light wavelength range, the highest value was obtained by C. vulgaris and the lowest by N. oleoabundans. In the case of phosphorus, the white LED assays showed the highest values of removal efficiency (up to 20%) and the blue LED assays presented the lowest values of this parameter (up to 13%). The results from this study present relevant insights on the influence of light wavelength on biomass production, CO 2 and nutrients uptake, and biochemical composition of biomass, pointing out some criteria for selection of the light wavelength to use depending on the intended application for microalgae and on the microalgal species used. Regarding CO 2 uptake rates determined in this study, although these values were lower than other values already reported in the literature, it is important to note that higher light intensities could be used to improve photosynthetic efficiency and, hence, CO 2 uptake rates.