Introduction

Photosynthesis in biosphere involves a fundamental process of initial CO2 capture and fixation. The overall photosynthetic efficiency depends on both the efficiencies of solar-to-chemical energy conversion in light reactions, and the carbon fixation in dark reactions1,2,3,4. The insufficient supplementation and low CO2 affinity with ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) limit the process of carbon fixation in chloroplast5,6. Hence, besides improving the activity of Rubisco7,8, it’s also important to understand the carbon concentrating mechanism (CCM) in photosynthetic organisms to develop more efficient CO2 fixation methods9,10,11. Cyanobacteria and eukaryotic algae in the biosphere contribute to the majority of global carbon assimilation and oxygen evolution12,13,14,15. Different from CCM in terrestrial C4 plants which rely on atmospheric CO2 uptake and special Kranz anatomy derived from cell differentiation16, aquatic microalgae operate CCM in a unicellular manner. Between two major forms of dissolved inorganic carbons (DIC) in water, non-polar CO2 molecules tend to diffuse across biomembrane driven by concentration gradient, while HCO3 relies on ATP-dependent transportation of membrane-bound transporters and is enriched in microalgal cell17. It was recently reported that mitochondrial-produced ATP energizes HCO3 transporters on cytoplasm and chloroplast envelope in the green alga Chlamydomonas reinhardtii18. Carbonic anhydrase (CA) in periplasm and chloroplast plays a crucial role in the interconversion of two major forms of DIC, CO2, and HCO3, which strengthens the directed HCO3 active transportation to the thylakoid, then converts HCO3 to CO2 in the proximity of Rubisco at a low luminal pH generated by alternative photosynthesis pathways, and recaptures inevitably leaked CO2 to HCO3 in chloroplast stroma18,19. When ambient CO2 supplementation is not enough, the CCM in algae is commonly activated to compensate the constraint of insufficient carbon supply from ambient atmosphere20. Although CCM is prone to be manipulated for increasing carbon fixation, the effectiveness is difficult to compensate the constraint of insufficient carbon supply from ambient atmosphere. Therefore, intensifying capture of CO2 dissolved in the solution to increase the concentration of CO2 around the algal cell is an alternative approach to the transport of CO2 into the cell for improving photosynthetic efficiency.

Metal-organic frameworks (MOFs) composed of metal nodes and organic ligands are promising materials for CO2 capture and storage from ambient air21,22,23,24. Appropriate humidity is reported to be beneficial to CO2 adsorption in microporous and mesoporous MOF, as pre-adsorbed H2O molecules in mesopores can form microporous pockets to enhance the CO2 confinement at low pressures25,26,27,28. Moreover, amine-functionalized MOF exhibits enhanced CO2 uptake capacity in humid conditions benefits from the high affinity of amino groups to CO226. However, it is not reported that the assembly of such material on biological system would be favorable to CO2 capture and transportation to microalgal cells in aqueous solution.

Here, we demonstrate an artificial CO2 concentrating approach by self-assembly of MOFs on microalgae for enhancing algal photosynthetic efficiency. The chemical concentrating CO2 approach improves the apparent photo conversion efficiency to about 1.9 folds, which is up to 9.8% of the MOF/C. pyrenoidosa in ambient air. We find that the efficient conversion of the enriched CO2 to bicarbonate by the excreted carbonic anhydrase of C. pyrenoidosa intensifies the CO2 supply to microalgae and stimulates the over-expression of Rubisco, which results in the enhancement of biomass production.

Results

Self-assembly of MOF on C. pyrenoidosa

The as-prepared MOF material has an average size of 300 nm (Supplementary Fig. 1), good crystallinity (Supplementary Fig. 2), and high BET surface area of 1143 mg−1 (Supplementary Fig. 3, Supplementary Method 1). It is self-assembled with C. pyrenoidosa, a kind of well commercialized green alga. CO2 adsorption property of the MOF was investigated in both gas and liquid phases. Figure 1a shows the relationship between CO2 adsorption/desorption and relative pressure of CO2. The capacity of MOF for CO2 adsorption is estimated to be as high as 20 cmg−1. Figure 1b shows that the CO2 breakthrough curves of the MOF in ambient air (400 ppm CO2) under different relative humidity (RH). The adsorption capacity of CO2 is estimated to be 1.3 mg g−1 under 3% RH and 1.8 mg g−1 under 45% RH, indicating that water molecule has a stimulating effect on CO2 adsorption. Figure 1c displays the CO2 adsorption of the MOFs in aqueous solution. It can be seen that the adsorption amount of DIC (CO2 and HCO3) increases linearly with the MOF loading and complete adsorption is achieved when the amount of MOF is reached to 100 mg L−1. It is noted that CO2 can be also desorbed in aqueous solution (Supplementary Fig. 4). These results indicate that the MOF has capability of CO2 enrichment in microalgal culture/growth medium.

Fig. 1: CO2-capture performances of MOF.
figure 1

a The CO2 adsorption/desorption isotherms of MOF (NH2-MIL-101-Fe) at 298 K. b Competitive CO2 breakthrough curves measured at two relative humidity (RH) values (Temperature, 303 K; Flux, 10 mL/min; CO2 concentration, 400 ppm; MOF amount, 50 mg). c Plots of the remaining DIC concentration in air-saturated BG-11 medium versus the additional amount of MOF. (Temperature, 303 K; pH, 7.0). d Plot of the zeta potentials of MOF and C. pyrenoidosa versus the pH values of cultivation medium. Error bars in c, d represent the standard deviation of the results from three biologically independent experiments. Source data are provided as a Source Data file.

Matching of surface charge between MOF and algae is important for self-assembly. Figure 1d shows the Zeta potential changes at different pH values. It can be seen that the MOF possesses positive surface charge in a broad pH range from 6 to 9 in the culture conditions. It is known that the surface of C. pyrenoidosa cell is negatively charged, because the cell wall is formed by a microfibrillar layer composed of polysaccharides and proteoglycans29,30. Consequently, the opposite electrostatic potentials between MOF and C. pyrenoidosa tend to drive them self-assembly by van der Waals interaction as supported by flow cytometric results (Supplementary Fig. 5). We also noticed that the MOF on the surface of C. pyrenoidosa protects photosystem II from photodamage by strong light or UV stress rather than blocking the light harvesting properties of microalgal chloroplast (Supplementary Fig. 68).

The self-assembled structure of MOF on C. pyrenoidosa were examined by microscopic analysis. Figure 2a shows that the MOF material is in octahedral shape, while Fig. 2b shows that the C. pyrenoidosa cell has a spherical surface with an average diameter of 3 μm (Supplementary Fig. 9). Figure 2c is the SEM image of the MOF/C. pyrenoidosa hybrid sample, which clearly shows that the MOF particles are adsorbed on the surface of the C. pyrenoidosa cell. Interestingly, Fig. 2d shows the MOF particles adsorbed on the cell become relatively tight and uniformly distributed after culturing for two days (Supplementary Fig. 10). In addition, several small particles are dispersed on the surface of the MOFs in the culture supernatant of C. pyrenoidosa after cultivation (Supplementary Fig. 11). It indicates that the free enzymes are diffused out of the cell and adsorbed on the MOF. Figure 2e illustrates the entire self-assembled hybrid of MOF with microalga. The interfacial interactions between MOF and C. pyrenoidosa cell, including electrostatic attractions, Van der Waals forces and hydrogen bonding, are accountable for the self-assembly of MOF on C. pyrenoidosa cell.

Fig. 2: Self-assembly of MOF on the surface of C. pyrenoidosa.
figure 2

SEM images of a MOF in fresh BG-11 medium, b C. pyrenoidosa, c, d the hybrids of MOF and C. pyrenoidosa of c low and d high magnification, e Proposed self-assembly models of MOF/Algae hybrid.

Biomass growth in the MOF/C. pyrenoidosa

Cell growth was investigated and the dry cell weight was used to evaluate the performance of carbon fixation (Supplementary Fig. 12). Figure 3a shows the relations between microalgal biomass growth and the amount of MOF in the microalgal cultivation medium under ambient air. It can be seen that the biomass production rate of bare C. pyrenoidosa is at a rate of 0.13 g L−1 day−1. Interestingly, the biomass production rate is significantly increased when MOF was adsorbed on the C. pyrenoidosa algae. When 50 ppm MOF is used for the self-assembly with C. pyrenoidosa, the biomass growth rate reaches the maximum value with a rate of 0.25 g L−1 day−1, in which is about 1.9-fold enhancement of biomass growth. Further increasing the loading amount of MOF to 100 ppm doesn’t accelerate the biomass growth, indicating that the adsorbed MOF particles could be saturated for C. pyrenoidosa cells. Figure 3b shows the influence of pH value on the growth of MOF/C. pyrenoidosa hybrid system. Neutral and weak alkaline environments (pH 7, 8) are preferable for microalgal growth than weak acidic environment (pH 6). Under pH 7 condition, the MOF/C. pyrenoidosa hybrid considerably increases the biomass growth. Similar biomass growth promotion phenomenon was also observed on another green alga Chlamydomonas reinhardtii (Supplementary Fig. 13). As the pH value commonly dominates the form of CO2 in aqueous solution, and the biophysical reaction is closely related to the proton gradient, the pH drift level of the above different systems was examined. Figure 3c displays the pH drift level of the MOF/algae system during the culture process. It can be seen that the pH value drifts to alkaline direction for MOF/algae system, which is more pronounced than that of the control at pH range from 6 to 8. The pH drift to alkaline indicates that OH is released to the outside of the cell after converting HCO3 into CO2 for fixation. The distinct alkaline microdomain at interface between MOF and algae is prone to promote the conversion of the captured CO2 into HCO3. The above observation can serve as the sign of the HCO3 utilization capacity of microalgae, which positively correlates with the microalgal biomass growth31. Figure 3d exhibits the microalgal quantum yield of PS II during cultivation. The change on photosynthetic pigments after adding MOF was also measured (Supplementary Fig. 14). The result shows that the performance of the light reactions in the assembled MOF/algae system is the same as that of the bare algae even at different pH points. Hence, the biomass growth in hybrid system is contributed to the dark reaction of carbon fixation. According to the optimized results of biomass production, the apparent photo conversion efficiency of the MOF/algae system is calculated to be 9.8% (Supplementary Fig. 15).

Fig. 3: Biomass growth in the MOF/C. pyrenoidosa.
figure 3

a The growth curves of C. pyrenoidosa with the addition of different amounts of MOF at initial pH 7 and be aerated with ambient air. b The growth curves of C. pyrenoidosa and MOF/Algae during two-day cultivation in a photobioreactor containing BG-11 medium which was buffered by 20 mM HEPES with different initial pH of 6, 7 and 8. c, d pH-drifts of media c and effective PS II quantum yield, Y(II) of C. pyrenoidosa d corresponding to b, the white dash line in c marked the initial pH of media. Cultivation conditions: temperature, 26 °C; light intensity, 50 μE m−2 s−1; 20 mL min−1 ambient air flow (LC); 24 h continuous illumination. Error bars in ad represent the standard deviation of the results from three biologically independent experiments. Source data are provided as a Source Data file.

The chemo-stability of MOF during long-term cultivation of C. pyrenoidosa was tested to exclude the interference of organic carbon for algae biomass growth. About 13% organic ligand of MOF was released into the culture medium after two-day cultivation, but the free organic ligand and Fe ion hardly accelerated the growth of C. pyrenoidosa (Supplementary Fig. 16). In order to see if it was the porous nature of MOF functions as CO2 adsorber and concentrator contributing to algal photosynthesis, some other typical porous materials, MIL-101-Fe, MIL-101-Cr and a commercial zeolite 13X, were also tested in the same experimental conditions. All these materials show a similar promotion behavior for algal photosynthetic carbon fixation (Supplementary Fig. 17). These results reveal that the MOF or zeolite assembled on the surface of the algae cell enables CO2 enrichment and transportation for photosynthetic carbon fixation in algae.

CO2 transportation mechanism from MOF to C. pyrenoidosa

To understand how the CO2 molecule is adsorbed on the MOF and then transported into the cell, we first investigate the carbonic anhydrase (CA) activity of the C. pyrenoidosa using two inhibitors, acetazolamide (AZA) and ethoxazolamide (EZA), which can block the CO2 transport chain. The carbonic anhydrases in the periplasm (external CA, eCA) and the chloroplast (internal CA, iCA) are both responsible for the regulation of the balance between CO2 and HCO3 in total DIC32,33,34. It is known that the membrane-impermeable AZA inhibitor targets on the inhibition of periplasmic eCA35. While the membrane-permeable EZA inhibitor targets on both eCA and iCA as it is a kind of comprehensive inhibitor to CCM36.

Figure 4a shows the biomass growth of C. pyrenoidosa and MOF/C. pyrenoidosa cells grown at different pH under air in the presence or absence of 1 mM eCA inhibitor AZA. The addition of AZA shows slight effect on the biomass growth of both C. pyrenoidosa and MOF/C. pyrenoidosa cells at pH 6. But the addition of AZA causes considerable decrease of biomass growth rate of MOF/C. pyrenoidosa cells at pH 7 and 8. After treated with the EZA inhibitor, the biomass growth of C. pyrenoidosa was almost stopped when EZA was used (Supplementary Fig. 18). This indicates that AZA inhibits the extracellular hydration of CO2 to HCO3 catalyzed by eCA and EZA inhibits the intracellular dehydration of HCO3 to CO2 catalyzed by iCA for CBB cycle which is vital to microalgal cell growth. The free eCA can diffuse into the medium through the cell wall for hydration of the dissolved CO2 in culture medium37,38. The inhibition of eCA gives rise to the sharp decline of biomass growth of the MOF/algae system. On the other hand, the activity of eCA left in medium are assayed at different pH to understand the process of CCM (Fig. 4b). The activities are nearly the same at pH 7 and 8, but the activity of CA at pH 6 decreased by 42%. Due to the adsorption of CA on the surface of MOFs, the apparent activities of CA left in medium are reduced correspondingly. It also can be ascribed to the improved DIC uptake with MOF lowering the expression level of eCA serving as a low CO2 inducible enzyme. With pH increasing from 6 to 8, CA activity in the media of low-CO2 grown C. pyrenoidosa suspensions also increases due to the favorable environment for the hydration of CO2 to HCO3. (Supplementary Fig. 19). The adsorbed protein amount and CA activity on MOF particles were measured (Supplementary Fig. 20). After the activities normalized by protein amount, the specific activity of CA on MOFs is similar with that in free condition, indicating that eCA adsorbed on MOF enables efficient conversion of MOF-captured CO2 into HCO3. The result is consistent with the image in Fig. 2e where most of the eCAs released by cell were adsorbed on the MOF, resulting in the promotion of the DIC supplementation to C. pyrenoidosa and hence acceleration of biomass growth (Supplementary Fig. 21).

Fig. 4: The coupling of MOF and the intrinsic CCM of C. pyrenoidosa.
figure 4

a The average biomass growth rate of C. pyrenoidosa within two-day cultivation with the addition of 50 ppm MOF, 1 mM AZA (acetazolamide, inhibit external carbonic anhydrase, eCA) and both of them at different pH. b CA activity measurements of the supernatants of microalgal suspensions which cultivated with or without MOF at different pH under air. c The Michaelis constant K1/2 values as calculated from the Michaelis-Menten fit of the plot of the net O2 evolution rate versus the concentration of CO2 for C. pyrenoidosa (Algae) and MOF/C. pyrenoidosa (MOF/Algae) cells grown at different pH under air (LC, 0.04%), error bars in (c) represent the standard deviation of K1/2 values as fitted from three independent experimental results. Cell density of 1.0 × 107 cell mL–1 and the light source with a cutoff wavelength filter (under 50 μE m−2 s−1 irradiation, λ > 600 nm). d The linear fitting curves of the kinetic plots of the reaction that the hydration of CO2 into HCO3 in different environments (pH 7.0 20 mM HEPES buffer, the supernatant of microalgal suspension after two-day cultivation, and the addition of bare MOF or MOF pre-treated in supernatant), pH changes indicate the accumulation of products since proton generated simultaneously with HCO3 in this process. e The apparent activities of the Rubisco based on the dry cell weight of biomass and the total protein, respectively. f The protein, carbohydrate and lipid contents of C. pyrenoidosa cultured two days in the absence (Algae) and presence (MOF/Algae) of MOF. Cultivation conditions: temperature, 26 °C; light intensity, 50 μE m−2 s−1; 20 mL min−1 ambient air flow (LC); 24 h continuous illumination; 20 mM HEPES buffer (initial pH 7.0). Error bars in af represent the standard deviation of the results from three biologically independent experiments. Source data are provided as a Source Data file.

In order to support the mechanism that MOF acts as a CO2 concentrator to accelerate the process of photosynthetic carbon fixation, we measured the apparent affinity of net O2 evolution for CO2 concentration of C. pyrenoidosa and MOF/C. pyrenoidosa cells cultivated at different pH (Supplementary Fig. 22). Figure 4c shows half saturation constant (K1/2) of the system to indicate the apparent affinity for CO2. We observe that the K1/2 for CO2 of C. pyrenoidosa cells is 65 μM at pH 6, decreases to 31 μM at pH 7 and increases to 43 μM at pH 8. And under all the conditions from pH 6 to pH 8, the function of MOF enables the K1/2 for CO2 decreased comparatively. Especially, MOF particles enables the apparent affinity for CO2 enhanced by 82% at pH 7, indicates the synergetic effect between MOF capturing CO2 and the intrinsic CCM in C. pyrenoidosa cells. The effect of MOF obviously promotes the biomass growth increased to 1.9-fold of C. pyrenoidosa cells (harboring a CCM) grown at pH 7 under air (LC, 0.04%), but in the case of cells grown for high CO2 (2%) without harboring a CCM, no role of MOF is played in enhancing the biomass growth (Supplementary Fig. 23). These results verifies that MOF as a CO2 concentrator is favorable for accelerating CO2 hydration into HCO3 by eCA. Figure 4d shows the reaction kinetics of the hydration of CO2 into HCO3 by eCA. The eCAs adsorbed on the MOF plays a crucial role in the promotion of the DIC supplementation to C. pyrenoidosa and hence acceleration of biomass growth (Supplementary Fig. 24).

To understand the carboxylation of the Rubisco enzyme in the process of carbon fixation, activities of the Rubisco enzyme based on the amount of total biomass and protein were tested. Figure 4e shows that the apparent activity of the Rubisco increased from 9.1 to 16.5 nmol min−1 per mg biomass, and from 32.6 to 40.6 nmol min−1 per mg protein. The expression level of Rubisco qualitatively increased (Supplementary Fig. 25). Figure 4f shows the contents of proteins, carbohydrates and lipids in dry cell weight. Among them, the protein content increases from 28% to 41% after the addition of MOF, while the carbohydrate decreases from 30% to 24% and the lipid decreases from 32% to 27%, which were calculated by the analysis methods (Supplementary Fig. 26). These results reveal that such an increase in Rubisco content may result from an increased internal Ci concentration in MOF-treated algae, and may contribute to reaching high CO2 fixation rates and high biomass productivity. However, no increase in the maximal O2 evolution rate was observed under non-limiting CO2 (Supplementary Fig. S22, S27), indicating that the increased Rubisco content in MOF-treated algae mainly contributes to dark reactions.

Discussion

In this study, we find that the affinity for CO2 of MOF/C. pyrenoidosa cell is higher than that of bare C. pyrenoidosa cell (Supplementary Fig. 27). But the presence of MOF particles doesn’t remarkably have influence on the affinity of cells grown under high CO2 (Supplementary Fig. 27). Because C. pyrenoidosa cells grown in high CO2 is not harboring the intrinsic CCM20. The complete CCM is vital to transfer CO2 toward Rubisco enzyme for accelerating the rate of CO2 fixation. Moreover, we find that the expression level of Rubisco protein is upregulated in MOF/C. pyrenoidosa cells (Fig. 4e). It was reported that the Rubisco content in C. pyrenoidosa was affected by the CO2 concentration during cultivation and nearly full activity of Rubisco must be needed during photosynthesis in C. pyrenoidosa and other green algae39. It was observed that Rubisco protein is more likely to stay in the chloroplast stroma than to be close-packed in the pyrenoid when CO2 concentration is elevated. And the additional Rubisco was distributed in the chloroplast stroma rather than to be organized in the pyrenoid40. The kinetic properties of Rubisco vary among different photosynthetic organisms, but remain steady on a certain species13. It implies the biomass growth rate strong dependence on the Rubisco content in C. pyrenoidosa. Thus, the synergetic effect of the functional MOF and the intrinsic CCM in C. pyrenoidosa cells enables a high affinity for CO2 and the Rubisco content in C. pyrenoidosa for accelerating CO2 fixation.

Figure 5 illustrates the CO2 enrichment and transportation mechanisms of MOF/C. pyrenoidosa assembly under atmospheric CO2 conditions. In artificial part, MOF captures and concentrates CO2 from the culture medium balanced with the air. The eCAs adsorbed on MOF catalyzes the captured CO2 hydration to HCO3. The increasing amount of HCO3 at bio-inorganic interfacial microdomain intensifies the transportation of HCO3 into the cell. the artificial CO2-enrichment and transport pathways not only increase the CO2 concentration in the cell but also accelerate the kinetic rate of Rubisco for CO2 fixation. In natural part, HCO3 in periplasm is transported via the membrane-bound HCO3 transporters to chloroplast pyrenoid. Finally, HCO3 ions are converted to CO2 by iCA in pyrenoid to feed the carboxylation process of Rubisco (Supplementary Fig. 28). The reversible switch between HCO3 and CO2 catalyzed by eCA and iCA enzymes plays a crucial role in driving CO2 delivery to the Rubisco sites for CO2 fixation reaction. After HCO3 conversion into CO2 by iCA, the alkaline microdomain formed at the interface between MOF and alga promotes the captured CO2 conversion into HCO3 by eCA enzyme. Therefore, bio-organic interfacial microenvironmental proton gradient and related ATP production are attributed as the driving force for extracellular CO2 capture and conversion to HCO3 by the MOF/C. pyrenoidosa for carbon fixation to enhance biomass production.

Fig. 5: The proposed CO2-enrichment and transport mechanisms of the MOF/C. pyrenoidosa assembly.
figure 5

Scenario 1: natural carbon fixation of microalgae growth is limited by the CO2 concentration in the cell ([CO2]1) and enzymatic kinetic rate k1. Scenario 2: in this work, MOFs as a CO2 concentrating entrepot are applied to accelerate the transport of CO2. The captured CO2 in MOF is hydrated to HCO3 by the excreted eCA adsorbed on MOF. The higher concentration of HCO3 outside is transported into the cell by the transporter, and then in pyrenoid, HCO3 is converted to CO2 by iCA to feed the carboxylation process of Rubisco. Compared to the intrinsic carbon concentrating mechanisms that hydrate dissolved CO2 to HCO3 for inorganic carbon supplementation, the artificial CO2-enrichment and transport pathways not only increase the CO2 concentration ([CO2]2) in the cell but also accelerate the rate of Rubisco (k2) for CO2 fixation, which increases the apparent photo conversion efficiency to 1.9 folds, up to 9.8% in ambient air.

In brief, a MOF/algae self-assembled hybrid system is constructed through the adsorption of MOF particles on the surface of the Chlorella pyrenoidosa for efficient photosynthetic CO2 fixation. CO2 enrichment on MOF/C. pyrenoidosa assembled system increases the biomass growth of C. pyrenoidosa by 1.9-folds compare with that of bare C. pyrenoidosa, and elevates the expression level of Rubisco by 82% meantime. The apparent photo conversion efficiency reaches up to 9.8% with the aid of the artificial CCM. The mechanism study shows that CO2 is firstly captured by MOFs and then converted into HCO3 by the extracellular eCA on the surface of MOF for catalytic CO2 hydration at bio-inorganic interface between the artificial and natural counterparts. HCO3 is confirmed to be the main form of inorganic carbon source for transportation in C. pyrenoidosa, elucidating the well coupling between the MOF-based artificial CCM and the intrinsic CCM of C. pyrenoidosa. The artificial concentration of CO2 by the assembly of algae with MOF reported in this work provides an avenue for improving ambient CO2 supplementation and facilitating algal photosynthesis for more efficient CO2 fixation and conversion.

Methods

Algal culture

Chlorella pyrenoidosa (FACHB-9) was cultured with 40 mL BG-11 medium in an air-lift column photobioreactor (20 cm high and 2 cm in diameter) under 50 μE m−2 s−1 continuous one-side illumination from a white fluorescent lamp (temperature, 26 ± 2 °C). Ambient air (LC) or 2% CO2 enriched air (HC) was set as 20 mL min−1. About 0.1 g L−1 cells were inoculated and harvested after two days cultivation by centrifugation (2220 g, 5 min). For MOFs and other CO2-captured materials assisted algal cultivation, as well as the addition of CA inhibitors, a certain amount of their suspension/solution were injected to BG-11 medium before algal inoculation. The cell density was determined by the absorbance at 750 nm (OD750) and calibrated by a calibration curve of dry cell weight (DCW) versus OD750. 20 mM 4-(2-hydroxyethyl)1piperazineethanesulfonic acid (HEPES) buffer solution was supplemented to adjust the initial pH of BG-11 medium if necessary. Chlamydomonas reinhardtii (cc-137) was cultured in the same conditions except the medium was replaced by TAP medium. Three biological replicates of each growth condition were processed.

Synthesis of NH2-MIL-101-Fe

NH2-MIL-101-Fe was prepared by reaction of 2-aminoterephthalic acid (NH2-H2BDC) with FeCl3·6H2O in DMF at 110 °C for 36 h. After being cooled to room temperature, the resultant precipitates were separated by centrifugation and washed thoroughly with DMF and ethanol. Then, the as-prepared NH2-MIL-101-Fe was soaked in EtOH with reflux for 24 h in order to exchange of the guest solvent molecules and excess ligand molecules. Finally, the obtained product was dried by vacuum drying.

Flow cytometry

MOF, C. pyrenoidosa and MOF/C. pyrenoidosa were redispersed in BG-11 medium to be analyzed by a Sony SH800 flow cytometer. The cell density of C. pyrenoidosa was 0.2 × 107 cell mL–1, and the concentration of MOFs is 10 ppm. The sheath fluid is 10 mM phosphate buffer solution (pH 7.2). FSC (forward scatter) mode was used for MOF and C. pyrenoidosa to analyze the particle size distribution of those sample suspensions according to their FSC-area signals (analyzed in Cell Sorter Software version 2.1).

CO2 adsorption/desorption performances

CO2 adsorption isotherms of NH2-MIL−101-Fe (298 K) were plotted by ASAP2020 physical adsorption analyzer (Micromeritics, USA) to determine the surface area and pore volume of MOF. 100 mg MOF powder in tube was degassed 2 h under high vacuum and 100 °C, and then measured its CO2 adsorption isotherms at 298 K. Breakthrough experiments were performed in a column packed with 50 mg fully activated MOF powder. 10 mL min-1 airflow with different humidity flew through the column, and the CO2 concentration at the outlet of the column was monitored online with an IR detector. Temperature was controlled at 303 K. For the measurement of CO2 adsorption capacity in water, 50 mg MOF was dispersed in 20 mL fresh BG-11 medium (pH 7.0 20 mM HEPES buffer), 20 mL min−1 airflow was purged for 2 h, then separated MOF powder by centrifugation and subsequent dried in 60 °C for 12 h, and immediately thermal desorbed as breakthrough experiments at 363 K.

Electron microscopy measurements

NH2-MIL-101-Fe, C. pyrenoidosa and MOF/C. pyrenoidosa were centrifugated (2220 g, 5 min) and rinsed three times with deionized water, and subsequently dehydrated in anhydrous ethanol for 30 min. The cell suspensions in ethanol were dropped onto a copper grid (3 mm in diameter) and dried in the air. Scanning electron microscopic characterizations were conducted on a JSM-7900F FESEM (JEOL, Japan) after samples sputtered with 2 nm platinum foil.

For the observation of the cross-section of bare C. pyrenoidosa, the prefixed cells should be further fixed by 1% osmic acid, then dehydrated in a graded series of ethanol solutions, then embedded in epoxy resin. Finally, the samples were ultrathin sectioned at ~50 nm thickness and placed on a copper grid for transmission electron spectroscopic observation (HT7700, HITACHI, Japan).

CA activity measurements

The enzymatic activities of extracellular carbonic anhydrase (CA) that existed in the supernatant of microalgal suspensions were determined with a colorimetric CA activity assay kit (Solarbio, China). Briefly speaking, 0.7 mL of Tris buffer (50 mM, pH 7.5), 0.2 mL of p-nitrophenyl acetate (p-NPA) aqueous solution, and 0.1 mL of medium centrifuged from different microalgal suspensions (or the redispersed suspension of MOF pretreated in medium) were mixed in the quartz cuvette. Since the hydrolysis of p-NPA to p-nitrophenol (p-NP) can be catalyzed by CA, the increase rate in the concentration of p- (p-NP) indicates the enzymatic activity of CA. The absorbances at 405 nm of the mixture were recorded before and after 5 minutes’ incubation in water bath (37 °C) to calculate the CA activity. The self-dissociation of p-NPA was measured in the same conditions for correction. The concentration of p-NP was calibrated by p-NP standard solutions.

Reaction kinetics of the hydration of CO2 into HCO3

The reaction kinetics of the hydration of CO2 into HCO3 in the supernatant of microalgal suspension or the MOF suspension were monitored according to Rawat37. Briefly speaking, 2 mL of the samples (OD750 = 2.0 for microalgal suspensions, contain 20 mM HEPES, pH 8.0) were kept at 4 °C. The control group was set as 2 mL 20 mM HEPES buffer (pH 8.0). Adding 1 mL of ice-cold CO2-saturated water to initiate the reaction. Then record the time intervals (T) required for the pH drop from 7.7 to 6.3. To record the kinetics of CO2 hydration, pH was recorded per 20 s during the process.

Chlorophyll fluorescence measurements

For PSII quantum yield measurements, C. pyrenoidosa and C. pyrenoidosa/MOF were suspended in a quartz sample cuvette and recorded with Water-PAM fluorometer (Heinz Walz GmbH, Germany) with an actinic light of 226 μE m−2 s−1 after 10 min dark adaptation. For PSII electron transfer rate (ETR II) measurements, the actinic light intensity varies from 0 to 2405 μE m−2 s−1.

Rubisco identification and activity measurements

C. pyrenoidosa and C. pyrenoidosa/MOF were harvested by centrifugation (2220 g, 5 min) for three times, then resuspended in 1 mL Rubisco extraction buffer, followed by disruption using sonication. The supernatant was collected by centrifugation (12000 g, 10 min, 4 °C). Finally, Rubisco activity of were determined with a colorimetric Rubisco activity assay kit (Solarbio, China). Briefly speaking, ribulose-1,5-biphosphate (RuBP) interacted with CO2 to produce 3-phosphoglycerate (3-PGA) with the catalysis of Rubisco. 3-PGA could be reduced to 3-phosphoglyceraldehyde which was stoichiometrically fueled by NADH. By monitoring the absorbance at 340 nm in 5 min interval, the Rubisco activity was determined. For the measurement of Rubisco expression level, the protein samples above were separated on a 10% SDS/PAGE, then be stained by coomassie brilliant blue. Rubisco was identified by Western blot. The protein was separated by SDS-PAGE and then transferred onto PVDF membranes (Millipore Co., USA) in Tris-glycine buffer at 100 mA for 2 h. The membranes were blocked with 5% (w/v) nonfat milk in Tris-buffered saline containing 0.1% (v/v) Tween-20 (TBST) at 37 °C for 2 h. The membrane was then incubated with the appropriate primary antibodies (Catalog No. AG5359, Beyotime, China) at a dilution ratio of 1:10,000 at 4 °C overnight. After three washes with TBST, the membrane was incubated with horseradish-conjugated goat-anti-rabbit secondary antibody (Catalog No. A0208, Beyotime, China) at a dilution ratio of 1:3000 for 1 h at room temperature in TBST with 1% nonfat milk. After three additional washes with TBST, the membrane was developed with ECL reagent. The images of the Rubisco large subunit (55 kDa) band were collected with a ChemDoc XRS+ system (Bio-Rad, USA).

Apparent photo conversion efficiency calculations

The apparent photo conversion efficiency (APCE) values in this work were calculated according to Wagner et al.41. We define it as the conversion efficiency of the actual incident light to biomass:

$${{{{{\rm{APCE}}}}}}(\%)={{{{{{\rm{E}}}}}}}_{{{{{{\rm{B}}}}}}}/{{{{{{\rm{E}}}}}}}_{{{{{{\rm{I}}}}}}}$$
(1)

EB – energy fixed in biomass (J); EI – energy in actual incident light (J).

For the calculation of EB, the calorific value (HB) of 23.4 KJ g−1 for C. pyrenoidosa42 was taken to calculate the chemical energy stored in biomass. During two days cultivation, EB can be calculated as:

$${{{{{{\rm{E}}}}}}}_{{{{{{\rm{B}}}}}}}={v}_{{{{{{\rm{B}}}}}}}\times {{{{{{\rm{H}}}}}}}_{{{{{{\rm{B}}}}}}}\times {{{{{\rm{V}}}}}}\times {{{{{\rm{t}}}}}}$$
(2)

vB—biomass growth rate (g L−1 day−1); HB—calorific value of biomass (J g−1), here assumed to be 23.4 KJ g−1; V—working volume of photobioreactor, 20 mL; t—process time, 2 days.

For the calculation of EI, a 7IGF10 grating spectrograph (Saifan Optoelectronic Instrument Co., Ltd., China) was used to measure the spectrum of light source. Photosynthetically active radiation (PAR) range of 400–700 nm is considered for the calculation of the average quantum energy of photons (U):

$$U={\int }_{400{nm}}^{700{nm}}\frac{{{{{{\rm{\phi }}}}}}({{{{{\rm{\lambda }}}}}}){{{{{\rm{hc}}}}}}}{{{{{{\rm{\lambda }}}}}}}d{{{{{\rm{\lambda }}}}}}\div{\int }_{400{nm}}^{700{nm}}{{{{{\rm{\phi }}}}}}({{{{{\rm{\lambda }}}}}})d{{{{{\rm{\lambda }}}}}}$$
(3)

U – average quantum energy of photons (J); ϕ(λ) – photon flux at different wavelength (count per second, arbitrary unit); h – Planck’s constant (6.626 × 10−34 J s); c – speed of light (2.998 × 108 m s−1); λ – wavelength (nm); A QSL-2101 radiometers (Biospherical Instruments Inc., USA) with an integrating sphere was used to measure the actual incident light intensity (reflected and diffused light by photobioreactor and medium was excluded) in a photobioreactor filled with fresh BG-11 medium.

The total energy in actual incident light EI is:

$${{{{{{\rm{E}}}}}}}_{{{{{{\rm{I}}}}}}}={{{{{{\rm{N}}}}}}}_{{{{{{\rm{A}}}}}}}\times {{{{{\rm{U}}}}}}\times {{{{{\rm{I}}}}}}\times {{{{{\rm{A}}}}}}\times {{{{{\rm{t}}}}}}$$
(4)

EI—energy in actual incident light (J); NA—Avogadro’s constant (6.022 × 1023 mol−1); U—average quantum energy of photons (Joule per photon = J); I—actual incident light intensity, measured to be 48.8 μmol m−2 s−1; A—cross section area of algal suspension, measured to be 12 cm2; t—process time, 2 days; Values were averaged over three independent experiments. As a result, the apparent photo conversion efficiency (APCE) values are 5.1% for the control group (bare C. pyrenoidosa) and 9.8% for MOF/Algae hybrids.

Protein content measurements

C. pyrenoidosa and C. pyrenoidosa/MOF were harvested by centrifugation (2220 g, 5 min) for three times, then resuspended in protein extraction buffer for ultrasonic treatment for cell disruption. Then the homogenate was centrifugated at 12000 g for 10 min at 4 °C. For the determination of extracellular protein in medium, the supernatants of algal suspensions were directly centrifuged (12000 g, 5 min) and concentrated 10-fold by N2 purging treatment. The protein content in supernatant was determined using BCA protein assay kit (Meilunebio, China). A calibration curve was established in advance using bovine serum albumin (BSA) gradient concentration solutions by the same method.

Carbohydrate content measurements

C. pyrenoidosa and C. pyrenoidosa/MOF were washed and collected by centrifugation (2220 g, 5 min) for three times, then freeze dried and milled to powder. 5 mg algal powder was dispersed and disrupted in 2 mL deionized water by sonication, and adjusted the volume to 5 mL. In total 5 mL anthrone test solution (1 g L−1 anthrone/80% H2SO4) and 0.2 mL of the homogenate were added into 0.8 mL deionized water, the mixture was boiled in boiling water for 10 min. Subsequently, the absorbance of the mixture at 625 nm was detected, and the carbohydrate concentration was calculated according to a calibration curve mapped with glucose standard solutions using the same method.

Lipid content measurements

A total of 50 mg freeze-dried algal powder of C. pyrenoidosa and C. pyrenoidosa/MOF was weighted and put into a 5 mL plastic centrifuge tube, using 2 mL lipid extraction solvent of Chloroform/methanol (2:1) extracted for 30 min and repeated three times. The extraction supernatant was collected and combined, then completely evaporate the solvent in a nitrogen blowing concentrator. The content of lipid can be weighed and calculated.

Statistical analysis

Unless otherwise indicated, statistical analysis was performed in OriginPro version 9. The samples chosen for analysis were derived from at least three biologically independent experiments. Data are presented as graphs or in-text showing the mean values ± SD as appropriate.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.