Light-driven biohybrid system utilizes N2 for photochemical CO2 reduction

ABSTRACT Attempting to couple photochemical CO2 reduction with N2 fixation is usually difficult, because the reaction conditions for these two processes are typically incompatible. Here, we report that a light-driven biohybrid system can utilize abundant, atmospheric N2 to produce electron donors via biological nitrogen fixation, to achieve effective photochemical CO2 reduction. This biohybrid system is constructed by incorporating molecular cobalt-based photocatalysts into N2-fixing bacteria. It is found that N2-fixing bacteria can convert N2 into reductive organic nitrogen and create a localized anaerobic environment, which allows the incorporated photocatalysts to continuously perform photocatalytic CO2 reduction under aerobic conditions. Specifically, the light-driven biohybrid system displays a high formic acid production rate of over 1.41 × 10−14 mol h−1 cell−1 under visible light irradiation, and the organic nitrogen content undergoes an over-3-fold increase within 48 hours. This work offers a useful strategy for coupling CO2 conversion with N2 fixation under mild and environmentally benign conditions.

The widely distributed P. azotofixans could tolerate extreme environments and interact with a variety of plants. After the N 2 -fixing bacteria P. azotofixans were recovered and purified, their 16S rDNA were extracted for strain identification. The sequence is as follows:

Section 2 Supplementary methods
Cultivation and inoculation of N 2 -fixing bacteria. In this work, the used N 2 -fixing bacterium was Paenibacillus azotofixans (formerly Bacillus azotofixans, ATCC 35681), which can be cryopreserved in a -80 °C freezer with 25% glycerin as a cryoprotectant. To obtain a large number of N 2 -fixing bacteria, Paenibacillus azotofixans was initially cultured in the heterotrophic medium. Specifically, 500.0 μL of thawed cryopreserved N 2 -fixing bacteria stock was added into 25.0 mL of heterotrophic medium, and incubated in a shaker (200 rpm) at 30 °C. After the growth for 2 days, the N 2 -fixing bacteria were collected by centrifugation at 6000 rpm for 5 min. The resulting bacteria were washed with PBS for three times, and re-suspended in 2.0 mL of PBS to obtain N 2 -fixing bacteria stock solution. 100.0 μL of the bacteria stock solution was dispersed into 15.0 mL of nitrogen fixation medium. The N 2 -fixing bacteria were cultured in nitrogen fixation medium at 30°C for another 3 days. All operations were carried out under a sterile environment.
Incorporation of Co-TTAP into N 2 -fixing bacteria. 100.0 μL of the bacteria stock solution was dispersed into 15.0 mL of nitrogen fixation medium. The N 2 -fixing bacteria were cultured in nitrogen fixation medium at 30 °C for 3 days. 500.0 μL of nitrogen fixation medium solution containing Co-TTAP (20 μg/mL) was added to the plate, and co-cultured with N 2 -fixing bacteria for 2 hours. The N 2 -fixing bacteria incorporating Co-TTAP were collected by centrifugation at 6000 rpm for 5 min, and re-dispersed into nitrogen fixation medium for photocatalytic testing. Each N 2 -fixing bacterium consists of approximately 1.26 × 10 7 of Co-TAPP molecules.
Laser scanning confocal microscope. The fluorescence imaging of biohybrid system was performed with confocal microscope (Leica TCS SP8 STED), equipped with air objectives. The N 2 -fixing bacteria were cultured in nitrogen fixation medium at 30 °C S7 for 3 days. 500.0 μL of nitrogen fixation medium solution containing photosensitizer (20 μg/mL) was added to the plate, and co-cultured with N 2 -fixing bacteria for 2 hours.
The N 2 -fixing bacteria incorporating photosensitizer were collected by centrifugation at 6000 rpm for 5 min, and re-dispersed into 500 μL of PBS. 5 μL of bacteria suspension was dropped onto a clean glass slide, then a cover glass was placed on the drop. The bacteria were observed via a laser scanning confocal microscope.
Scanning electron microscopy (SEM). Bare N 2 -fixing bacteria and Co-bacteria hybrid were collected and purified by centrifugation. Then, bare N 2 -fixing bacteria and Co-bacteria were fixed by soaking in glutaraldehyde solution (2.5 vol.%) at 4 °C overnight. After centrifugation, bare N 2 -fixing bacteria and Co-bacteria were subjected to ethanol dehydration by placing them in 35, 50, 70, 95 and 100% ethanol in PBS for 10 min each. The 100% ethanol was changed three times, and the samples were dried with N 2 purge. Finally, free N 2 -fixing bacteria and Co-bacteria were adhered to SEM posts with carbon film tape and imaged with a SEM at 5 kV (Sigma).

Transmission electron microscopy (TEM).
After the N 2 -fixing bacteria were co-cultured with the Co-TTAP (20 μg/mL) for 4 hours, the obtained Co-bacteria were fixed with glutaraldehyde solution (2.5 vol.%) at 4 °C overnight. The fixed Co-bacteria were collected by centrifugation and re-dispersed into PBS. Samples of fixed Co-bacteria were prepared for TEM and EDS mapping by dropping the Co-bacteria suspension onto Cu TEM grids and settling for 2 hours. After the samples were washed briefly in deionized water, the grids were dried in air overnight. TEM imaging and EDS mapping was performed with a JEM-F200 transmission electron microscope at 200 kV.
Colony forming unit assay. The optical density of N 2 -fixing bacteria suspension at S8 was diluted 10 4 , 10 5 and 10 6 times respectively to prepare the bacteria dilutions. 100 μL of the bacteria dilution was added to the plate loaded with nitrogen fixation medium. 500.0 μL of medium solution containing Co-TTAP (20 μg/mL) was added to the medium. The plate was sealed with parafilm, placed upside down and incubated in an incubator at 30 °C for 3 days. After the growth for 3 days, the white circular colonies were counted to determine the CFU mL -1 as a measure of bacteria number and viability. All operations were carried out under a sterile environment.
Determination of nitrogenase activity. The acetylene (C 2 H 2 ) method was used to determine the nitrogenase activity of N 2 -fixing bacteria. Briefly, 40.0 mL of nitrogen fixation medium was added to the glass bottle (100 mL). N 2 -fixing bacteria were inoculated in nitrogen fixation medium and cultured at 30 °C for 3 days. 1.0 mL of Co-TTAP (20 μg/mL) medium solution was added to the medium containing N 2 -fixing bacteria. After co-cultivation for 2 hours, N 2 -fixing bacteria incorporating Co-TTAP were collected by centrifugation at 6000 rpm for 5 min. The obtained N 2 -fixing bacteria containing Co-TTAP (Co-bacteria) were re-dispersed into 40.0 mL of nitrogen fixation medium in 100 mL-glass bottle. Control groups included Co-TAPP (containing the equal amount of Co-TAPP as Co-bacteria solution) and bare N 2 -fixing bacteria. The glass bottles were sealed with rubber stoppers, and then 10.0 mL of acetylene was injected. After the growth for 3 days, the gas phase composition was analyzed by gas chromatography (Fast refinery gas analysis system based on the Agilent7890B GC system). The bacteria were collected for concentration (6000 rpm, 5 min). The protein content of bacteria was measured according to Bradford with bovine serum albumin as the standard. The activity of nitrogenase is calculated according to the following equation (1): Here, m refers to the quality of bacterial protein and T is the reaction time.
The nitrogenase activity determination of Co-bacteria after photocatalytic CO 2 reduction tests. The acetylene (C 2 H 2 ) method was used to determine the nitrogenase activity. The Co-bacteria were re-dispersed into 40.0 mL of nitrogen fixation medium in 100 mL-glass bottle. Control groups included Co-TTAP and light-irradiated bare N 2 -fixing bacteria. The glass bottles were sealed with rubber stoppers, and then 10.0 mL of acetylene was injected. After the growth for 3 days, the gas phase composition was analyzed by gas chromatography (Fast refinery gas analysis system based on the Agilent7890B GC system). The bacteria were collected for concentration (6000 rpm, 5 min). The protein content of bacteria was measured according to Bradford with bovine serum albumin as the standard. The activity of nitrogenase is calculated according to the following equation (1).
Photocatalytic CO 2 reduction tests. Prior to photocatalytic experiments, (2.08 ± 0.09) × 10 8 of Co-bacteria (containing ~1.26 × 10 7 Co-TAPP/cell) were inoculated in nitrogen fixation medium on the plate with a diameter of 4 cm. The plate was transferred into a standard gas-liquid-solid photo-reactor. Then, the photo-reactor was purged with a mixture of gases (N 2 /CO 2 /O 2 , 6:3:1). The reactor was illuminated with a 300 W Xe lamp with 420-cut filter (removing light with wavelengths less than 420 nm) to obtain visible light (light intensity, 7.5 mW cm -2 ). Liquid phase solution was centrifuged to remove Co-bacteria. The liquid products of photocatalytic process were analyzed by an ion chromatography (the sample is diluted 100 times). The composition of the gas atmosphere was monitored by gas chromatography (Fast refinery gas analysis system based on the Agilent7890B GC system). A three channel system on the Agilent7980B GC system was used for the determination of the composition of gas atmosphere in this photocatalytic system. Five valves and eight S10 columns are used with the system. Channel 1, using a Flame Ionization Detector (FID1), was used for hydrocarbons from methane to C6+. Permanent gases (i.e. O 2 , N 2 , CO 2 and CO) and hydrogen sulfide were measured on Channel 2, using a Thermal Conductivity Detector (TCD2). Hydrogen (H 2 ) was measured on the Channel 3, using a Thermal Conductivity Detector (TCD3), where N 2 was used for carrier. Control groups included: (1) free Co-TAPP in nitrogen fixation medium + light, (2) bare N 2 -fixing bacteria + light, and (3) Co-bacteria in dark. Each group includes 3 parallel samples.
Trace the C source using 13 CO 2 . In order to trace the source of C in photocatalytic products, we performed photocatalytic reactions of Co-bacteria by using a mixture of gases containing 13 CO 2 (N 2 / 13 CO 2 /O 2 , 6:3:1). Except that the 12 CO 2 was replaced with 13 CO 2 , H 2 O was replaced with D 2 O, and the other photocatalytic treatments were the same as before. After 24 h, the reaction products were collected and analyzed by carbon spectroscopy on a Bruker Avance III HD 400 MHz.
Gas chromatography analysis. The composition of the gas atmosphere was monitored by gas chromatography (Fast refinery gas analysis system based on the Agilent7890B GC system). A three channel system on the Agilent7980B GC system was used for the determination of the composition of gas atmosphere in this photocatalytic system. Five valves and eight columns are used with the system. Channel 1, using a Flame Ionization Detector (FID1), was used for hydrocarbons from methane to C6+. Permanent gases (i.e. O 2 , N 2 , CO 2 and CO) and hydrogen sulfide were measured on Channel 2, using a Thermal Conductivity Detector (TCD2).
Hydrogen (H 2 ) was measured on the Channel 3, using a Thermal Conductivity Detector (TCD3), where N 2 was used for carrier.
Materials and reagents of full analysis: carrier gas, N 2 (purity > 99.999%), H 2 S11 (purity > 99.995%); fuel gas, H 2 (purity > 99.995%); oxidant gas, air; Auxiliary gas (septal purge and tail purge), N 2 (purity > 99.995%  Photocatalytic CO 2 reduction tests of Co-TAPP. We investigated the photocatalytic activity of Co-TAPP by using different amino acids as sacrificial electron donors under full spectrum irradiation (λ ≥ 420 nm). Co-TAPP was dispersed in deionized water for photocatalytic measurements. A standard gas-liquid-solid reactor was used for the assessment of photocatalytic CO 2 reduction performance. In a typical S12 procedure, photocatalysts were dispersed in deionized water containing amino acid (50 mM) solution. The concentration of Co-TAPP photocatalyst was determined to 6.0 µg/mL (containing the equal amount of Co-TAPP as Co-bacteria solution). The mixture was pre-degassed with CO 2 for removing the dissolved O 2 , and then injected into the gas-liquid-solid reactor. The reactor was illuminated using a 300 W Xe lamp (Purchased from Beijing China Education Au Light Technology Co., Ltd.) with UVIRCUT420 nm filter to obtain full spectrum irradiation light source (light intensity, 7.5 mW cm -2 ). The irradiation time is 24 hours. The composition of the gas atmosphere was monitored by gas chromatography (Fast refinery gas analysis system based on the Agilent7890B GC system). The liquid products of photocatalytic process were analyzed by an ion chromatography (DIONEX ICS 2500).
Total organic nitrogen content analysis. (2.08 ± 0.09) × 10 8 of Co-bacteria (containing ~1.26 × 10 7 Co-TAPP/cell) were inoculated in nitrogen fixation medium (without N) on the plate. The plate was transferred into a standard gas-liquid-solid photo-reactor. Then, the photo-reactor was purged with a mixture of gases (N 2 /CO 2 /O 2 , 6:3:1). The reactor was illuminated with a 300 W Xe lamp with 420-cut filter to obtain visible light (light intensity, 7.5 mW cm -2 ). After the samples were irradiated for 48 hours, the medium containing bacteria and in the plate were freeze-dried for total organic nitrogen analysis, and the mass of the freeze-dried mixture was determined to be 245.8 ± 17.4 mg. Analyzes of total organic nitrogen content were performed on a FOSS automatic kjeldahl nitrogen analyzer. Control groups included: (1) free Co-TAPP in nitrogen fixation medium + light, (2) bare N 2 -fixing bacteria + light, and (3) Co-bacteria in dark. Each group includes 3 parallel samples. Before irradiation, the culture medium containing Co-bacteria was freeze-dried for total organic nitrogen content analysis. The mass of the freeze-dried mixture was S13 determined to be 238.6 ± 11.2 mg.
Total organic nitrogen content analysis without light irradiation. (2.08 ± 0.09) × 10 8 of Co-bacteria (containing ~1.26 × 10 7 Co-TAPP/cell) were inoculated in nitrogen fixation medium (without N) on the plate. After the sample was cultured for 48 hours, the mixture containing bacteria and medium in the plate was collected for total organic nitrogen analysis. Analyzes of total organic nitrogen content were performed on a FOSS automatic kjeldahl nitrogen analyzer. Control groups included: (1) free Co-TAPP in nitrogen fixation medium, and (2) bare N 2 -fixing bacteria. Each group includes 3 parallel samples.
Apparent quantum efficiency (AQE) study. The AQE was measured in identical experimental setup and under the same condition to the CO 2 photo-reduction test except for the incident light resource. Co-bacteria (containing ~1.26 × 10 7 Co-TAPP/cell) were inoculated in nitrogen fixation medium on the plates. The photo-reactor was purged with a mixture of gases (N 2 /CO 2 /O 2 , 6:3:1). The typical irradiation area of photocatalytic system was 12.56 cm 2 . The distance between photocatalytic solution surface and the light resource fixed at 10.0 cm throughout the entire catalytic experiment. The intensity of visible light was calibrated to be 7.5 mW cm -2 (averaged wavelength, 546 nm). The sample was irradiated for 48 hours and the photo-catalytic products were quantitatively detected by ion chromatography. The AQE value was calculated as the number of electrons consumed to generate the product divided by the total number of incident photos (Equation 2). This method has been widely adopted in the field of CO 2 photoreduction. Here, 2 electrons were used to convert one CO 2 molecule to HCOOH (Equation 3). Each AQE measurement was repeated at least three times in parallel to obtain a reproducible and reliable value with the error bar displayed.
Microbial metabonomics study. (2.08 ± 0.09) × 10 8 of Co-bacteria (containing ~1.26 × 10 7 Co-TAPP/cell) were inoculated in nitrogen fixation medium on the plates. These plates were randomly divided into dark group and light group (each group includes 3 parallel samples). The plate was transferred into a standard gas-liquid-solid photo-reactor. Then, the photo-reactor was purged with a mixture of gases (N 2 /CO 2 /O 2 , 6:3:1). The light group was illuminated with a 300 W Xe lamp with 420-cut filter (light intensity, 7.5 mW cm -2 ). The control group (dark group) was placed in a dark environment. After treating for 24 hours, Co-bacteria were collected and preserved at -80 °C.
50.0 mg of Co-bacteria for each sample was accurately weighed into a 2-mL centrifugation tube. Then, a grinding bead with a diameter of 6 mm and 400 μL of extraction solution containing the internal standard of L-2-Chlorophenylalanine (0.02 mg/mL) were added to the centrifuge tube. The frozen tissue grinder was used for grinding samples at a frequency of 50 Hz at -10 ℃. The sample was put into the S15 low-temperature ultrasound (40 kHz, 5 ℃) for 30 min to extract metabolites fully, and then placed at -20 ℃ for 30 min. After the sample is centrifuged, the supernatant is transferred to an injection vial with an inner cannula for analysis on the UHPLC-Q Exactive HF-X system (Thermo Fisher was cultured with Co-bacteria for sensitive, one-step fluorometric to detect intracellular ROS in live cells within 1-hour incubation. The analysis was performed 24 hours after the illuminated in nitrogen fixation medium. The Co-bacteria in the dark was as the control. The samples were observed via a laser scanning confocal microscope (Leica TCS SP8 STED, λ ex = 488/λ em = 530 nm).

Intracellular oxygen analysis.
Luminescent oxygen sensor tris(4,7-diphenyl-1,10-phenanthroline)ruthenium(II) dichloride complex is a luminescent hypoxic probe and its fluorescence intensity increases as the S16 concentration of oxygen decreases. This hypoxic probe was cultured with N 2 -fixing bacteria for detecting intracellular oxygen within 1-hour incubation. The samples were observed via a laser scanning confocal microscope (Leica TCS SP8 STED, λ ex = 450/λ em = 620 nm). The dead N 2 -fixing bacteria was as the control. S52 Figure S36 Orthogonal partial least squares discriminant analysis (OPLS-DA) of differential metabolites are used to assess the differences between dark groups and light groups.

Figure S37
Histogram of functional pathway analysis for the differential metabolites between dark groups and light groups.
S54 Figure S38 Bubble chart of the KEGG pathway enrichment analysis for the differential metabolites between dark groups and light groups. S64 Figure S48 The effects of different temperatures on the cell viability of biohybid system.  Table S5. Co-TAPP were used for photocatalytic CO 2 reduction tests. The gas chromatography analysis showed the composition of the gas atmosphere in Co-TTAP photocatalytic system (glutamic acid was used as sacrificial electron donor).  Table S6. Comparison of the external sacrificial reagent of previously reported biohybrid systems and our biohybrid system in this work.

Section 4 Supplementary Tables
Life-unit organism Photocatalyst Substrate Products