Reactor design and construction
An electrochemical continuous stirred-tank reactor (E-CSTR) was used for this study. The design of the E-CSTR is shown in Fig. 2. The reactor consisted of a tubular anode membrane assembly and a jacketed mixing vessel (outer diameter: 12.4 cm, inner diameter: 10.7 cm, height: 14.5 cm, maximum volume: 1 L). The anode membrane assembly consisted of a lid and a tubular membrane cell (diameter: 3.5 cm, height: 12.5 cm) that was placed in the middle of the lid. The bottom of the anode chamber was sealed by a PVC cap, while the top was plugged with a rubber stopper. The rubber stopper held the anode and an anode gas outlet tube. The size of the cation exchange membrane (CMI-7000, Membranes International Inc., USA) window on the anode membrane assembly was 42 cm2 (6 cm × 7 cm). There were six openings on the lid of the reactor, and they were used for the Stainless steel bar cathode current collector, CO2 inlet tube, base dosing tube, pH probe, cathode gas outlet tube, and liquid sample port. A stainless-steel mesh cylinder (length: 16 cm, height: 8 cm, thickness: 0.3 mm, Anguo Chengli Metal Co., Ltd., China) was used as the cathode. An IrO2-coated titanium mesh plate (width: 3.2 cm, height: 8 cm, Baoji Zhiming Special Metal Co., Ltd., China) was used as the anode. The reactor lid and the flange of the mixing vessel were sealed by an O-ring and a clamp.
The working volume of the anolyte and catholyte was 100 mL and 600 mL, respectively. The anolyte was 0.2 mol·L− 1 Na2SO4 solution, and its pH was adjusted to 2 with H2SO4. The catholyte was a modified M9 medium containing 6 g·L−1 Na2HPO4, 3 g·L−1 KH2PO4, 0.5 g·L−1 NH4Cl, 0.5 g·L−1 NaCl, 0.1 g·L−1 MgSO4·7H2O, 0.0146 g·L−1 CaCl2, 4 g·L−1 NaHCO3, 1 ml·L−1 trace element solution and 1 ml·L−1 vitamin solution. The compositions of the trace element solution and vitamin solution were provided in Table S1. The jacketed mixing vessel was connected to a recirculating water bath (DC-1006, Ningbo Scientz Biotechnology Co., Ltd., China) to control the reactor temperature. A magnetic stirrer was placed at the bottom of the vessel to control the agitation intensity of the catholyte. The anode and cathode were connected to the positive and negative terminals of the DC power supply (KA3005P, Korad Technology Co., Ltd., China), respectively. The pH of the catholyte was controlled with a pH controller (pH3.0-NI2-AC, Huzhou Tianze Biotechnology Co., Ltd., China) by automatically dosing 5 M NaOH. The reactor was equipped with a mass flow controller (D07-7, Beijing Sevenstar Flow Co., Ltd, China) to control the CO2 flow rate. A water displacement column (filled with 1200 mL 1 mM HCl) was connected to the reactor to collect the offgas for gas flow rate and composition measurement.
Reactor startup and operational procedure
Two reactors were run in parallel, one as the control and the other one as the experimental reactor. The SiO2 NPS used in this study were commercial hydrophilic SiO2 NPS that was produced by the sol–gel method (average diameter 200 nm, Hebei Juli Metal Material Co., Ltd., China). The SiO2 NPS could easily disperse in water so they were used as purchased without any pretreatment.
In the control reactor, no SiO2 NPS were added, while in the experimental reactor 0.3wt% SiO2 NPS were added (1.8 g of SiO2 NPS were added into 600 mL of catholyte). The dosage of the SiO2 NPS (0.3wt%) was chosen based on literature regarding syngas fermentation enhancement by adding SiO2 NPS (Kim et al. 2014).
The reactor temperature was controlled at 35 ± 0.5 ℃ by the water bath, and the stirring rate of the magnetic stirrer was set at 650 rpm. The catholyte of the reactor was first flushed with N2 for 30 min to remove the dissolved oxygen. Then, the cathode chamber was inoculated with an enriched acetate-producing mixed culture that was dominated by Acetobacterium. The reactor was operated in a galvanostatic mode with a starting current of 0.25 A and then switched to 0.5 A. At 0.25 and 0.5A, the CO2 flow rate was controlled at 1.96 and 3.92 mL·min− 1, respectively, which resulted in an H2/CO2 molar ratio of 2:1. The pH of the catholyte was controlled at 7 using the pH controller.
Each day, 2 mL of catholyte was taken to measure the pH and VFAs, and the same amount of fresh catholyte was added. Occasionally, fresh anolyte was added to the anode chamber to compensate for the water loss. The reactors were operated in fed-batch mode. When the acetate concentration stopped increasing, 90% of the catholyte was replaced with a fresh medium to start a new batch. In the second batch, 0.3wt% SiO2 NPS were added to the experimental reactor.
Analytical methods
Gas analysis
Gas samples were taken from the water displacement column to analyze the volume and composition of the unused gas. The gas composition, i.e., the concentrations of H2 and CO2, was analyzed by a compact gas chromatograph (GC, 7890B, Agilent). Details of the method could be found in our previous publication (Cai et al. 2022).
VFAs analysis
Liquid samples were taken from the reactor to analyze the concentration of acetate. The liquid samples were first extracted by diethyl ether and then injected into gas chromatography (GC-2010 Pro, AOC-201, Shimadzu, Japan) with an FID detector to analyze. Details of these methods could be found in our previous publication (Cai et al. 2022).
Hydrogen mass transfer coefficient measurement
The volumetric mass transfer coefficient (KLa, h− 1) which describes the transfer resistance at the gas–liquid interface was tested by the dynamic-gasing method described in a previous article (Beckers et al. 2015). Briefly, the reactor was filled with M9 medium without inoculum. The temperature was kept at 35 ℃. The reactor was equipped with a dissolved hydrogen sensor (Clean, DH200, China) at the cover to measure and record dissolved H2 concentrations. The KLa of H2 in the reactor with SiO2 NPS and that without SiO2 NPS were measured under 0.5 A in abiotic conditions. For each experiment, the reactor was degassed with N2 to remove H2 before electrolysis. The values of the dissolved hydrogen sensor were recorded once a minute until reaching saturation.
Calculations
The kLa was calculated according to the adsorbing equation (Myung et al. 2016):
$$\frac{dC}{dt}={K}_{L}a\left({C}^{*}-C\right)$$
(1)
Here, \({C}^{*}\) is the saturated concentration of dissolved hydrogen (mg H2 L− 1), KL is the mass transfer coefficient (cm·h− 1), and a is the gas/liquid interfacial area per volume of liquid (cm2·cm− 3).
The CE was calculated as described earlier (Liu et al. 2015), to reflect the productivity of the system:
$$CE = \frac{{\Delta {\text{C}}_{\text{HAc}} {\text{(mol L}}^{{ - 1}}) \times {\text{ V}}_{\text{solution}} {\text{(L) }} \times 8 \times {\text{ F (C mol}}^{{ - 1}} {)}}}{{\text{Overall charge (C)}}}{ } \times { } 100 \%$$
(2)
Here, ΔCHAc (mol·L–1) is the change in the concentration of acetate during the experiment, Vsolution is the total volume of the solution in the cathode chamber, and ‘Overall charge’ is the total electric charge passing through the cathode chamber and F is the Faraday’s constant.