Enhancement of acetate production in hydrogen-mediated microbial electrosynthesis reactors by addition of silica nanoparticles

Microbial electrosynthesis (MES) is a promising technology for CO2 fixation and electrical energy storage. Currently, the low current density of MES limits its practical application. The H2-mediated and non-biofilm-driven MES could work under higher current density, but it is difficult to achieve high coulombic efficiency (CE) due to low H2 solubility and poor mass transfer. Here, we proposed to enhance the hydrogen mass transfer by adding silica nanoparticles to the reactor. At pH 7, 35 ℃ and 39 A·m− 2 current density, with the addition of 0.3wt% silica nanoparticles, the volumetric mass transfer coefficient (kLa) of H2 in the reactor increased by 32.4% (from 0.37 h− 1 to 0.49 h− 1), thereby increasing the acetate production rate and CE of the reactor by 69.8% and 69.2%, respectively. The titer of acetate in the reactor with silica nanoparticles (18.5 g·L− 1) was 56.9% higher than that of the reactor without silica nanoparticles (11.8 g·L− 1). Moreover, the average acetate production rate of the reactor with silica nanoparticles was up to 2.14 g·L− 1·d− 1 in the stable increment phase, which was much higher than the other reported reactors. These results demonstrated that the addition of silica nanoparticles is an effective approach to enhancing the performance of H2-mediated MES reactors. Graphical Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s40643-023-00627-6.


Introduction
Microbial electrosynthesis (MES) is a process that utilizes microorganisms as the cathodic catalysts for electrochemical CO 2 reduction (Nevin et al. 2011). Compared to abiotic catalysts, microorganisms have the advantages of high selectivity, self-regenerating ability, and capability of producing multi-carbon organics (Salehizadeh et al. 2020). Consequently, MES has drawn increasing attention in the fields of CO 2 valorization and renewable energy stargate (Logan & Rabaey 2012). The concept of MES emerged around 2010 after some methanogens (Cheng & Logan 2007) and acetogens were found to be able to uptake electrons from the cathode for the production of methane and acetate, respectively. After that, a lot of work has been done to modify electrode materials, discover effective microorganisms, understand extracellular electron transfer (EET) mechanisms, and develop novel reactor configurations (Aryal et al. 2017;Krieg et al. 2014). The current density (i.e., productivity) of MES has improved a lot, but it is too low to bring this technology onto the market (Jourdin & Burdyny 2021;Prevoteau et al. 2020). Figure 1 shows the working principle of MES. In this system, renewable electricity is used to oxidize H 2 O to form O 2 , H + , and e − in the anodic chamber, and then H + and e − are transported to the cathode chamber in which they are used by autotrophic microbes to fix CO 2 for organic matter production. The conventional MES was driven by the biofilm on the cathode surface, because biofilm enabled high coulombic efficiency (CE), low energy input, and good biomass retention (Fruehauf et al. 2020). However, the current density of the biofilmdriven MES was limited by the activity of the biomass and the mass transfer efficiency. Currently, the current density of cathodic biofilms has reached the maximum values of electro-active biofilms (i.e., 10-100 A·m − 2 ) (Claassens et al. 2019). Thus, it is very challenging to further increase the current density of biofilm-driven MES, especially for its large-scale applications. Therefore, nonbiofilm-driven MES has been proposed recently. This non-biofilm-driven MES used suspended biomass and hydrogen produced from the cathode for CO 2 conversion, which was also named H 2 -mediated MES. Without the limitation of biofilm, high current density could be applied to the reactor for hydrogen production. However, the low solubility and poor mass transfer of H 2 make it difficult to achieve high CE under high current density for this type of H 2 -mediated MES (Liu et al. 2016).
Previously, it has been reported that the enhancement of H 2 mass transfer is an effective method to improve the performance of H 2 -mediated MES reactors. For example, Rodrigues et al. reported that adding perfluorocarbon nanoemulsion (a hydrogen carrier) into H 2 -mediated MES increased the productivity of acetate by 190% and obtained a CE of nearly 100% (Rodrigues et al. 2019). Our group also found that adding porous polyurethane (PPU) particles could promote the hydrogen uptake efficiency and acetate production rate in an H 2 -mediated MES reactor (Xue et al. 2022). Moreover, nanoparticles have also been applied to syngas fermentation and methane fermentation to improve the gas-liquid mass transfer (Kim et al. 2014;Zhu et al. 2008). Therefore, we hypothesized that introducing nanoparticles to H 2 -mediated MES may also enhance the hydrogen mass transfer, CE and acetate production rate. Among all reported nanoparticles, silica nanoparticles (SiO 2 NPS) are biocompatible, dispersible in water, and commercially available at a relatively cheap price. Compared to perfluorocarbon nanoemulsion, SiO 2 NPS can be easily separated from the fermentation broth for reuse. The PPU particles are on a millimeter scale, so that they can be easily maintained in the reactor, but the specific surface area of PPU is rather limited when compared to SiO 2 NPS. Adding SiO 2 NPS may further improve the hydrogen mass transfer in H 2 -mediated reactors. Thus, the goal of this study was to investigate whether the addition of SiO 2 NPS could enhance the hydrogen mass transfer and the performance of an H 2 -mediated MES.
Our experimental results showed that the addition of SiO 2 NPS in the MES system could significantly increase the H 2 mass transfer rate and H 2 solubility in water, thereby increasing the acetate production rate by 69.8% and the acetate titer by 56.9%. The average acetate production rate of the reactor with silica nanoparticles (2.14 g·L − 1 ·d − 1 ) was higher than those of the reported MES reactors. These results demonstrated adding SiO 2 NPS is an efficient way to enhance the performance of H 2 -mediated MES reactors. The method reported here dramatically increases the current density and coulombic efficiency of H 2 -mediated MES reactors. It can not only be applied in H 2 -mediated MES reactors for acetate and methane production but also in other H 2 -dependent reactions, such as biologically catalyzed N 2 fixation, CH 4 functionalization and microbial protein production.

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 cm 2 (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, CO 2 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 IrO 2 -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 Na 2 SO 4 solution, and its pH was adjusted to 2 with H 2 SO 4 . The catholyte was a modified M9 medium containing 6 g·L −1 Na 2 HPO 4 , 3 g·L −1 KH 2 PO 4 , 0.5 g·L −1 NH 4 Cl, 0.5 g·L −1 NaCl, 0.1 g·L −1 MgSO 4 ·7H 2 O, 0.0146 g·L −1 CaCl 2 , 4 g·L −1 NaHCO 3 , 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 CO 2 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 SiO 2 NPS used in this study were commercial hydrophilic SiO 2 NPS that was produced by the sol-gel method (average diameter 200 nm, Hebei Juli Metal Material Co., Ltd., China). The SiO 2 NPS could easily disperse in water so they were used as purchased without any pretreatment.
In the control reactor, no SiO 2 NPS were added, while in the experimental reactor 0.3wt% SiO 2 NPS were added (1.8 g of SiO 2 NPS were added into 600 mL of catholyte). The dosage of the SiO 2 NPS (0.3wt%) was chosen based on literature regarding syngas fermentation enhancement by adding SiO 2 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 N 2 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 CO 2 flow rate was controlled at 1.96 and 3.92 mL·min − 1 , respectively, which resulted in an H 2 / CO 2 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% SiO 2 NPS were added to the experimental reactor.

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 H 2 and CO 2 , 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

Hydrogen mass transfer coefficient measurement
The volumetric mass transfer coefficient (K La , 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 H 2 concentrations. The K La of H 2 in the reactor with SiO 2 NPS and that without SiO 2 NPS were measured under 0.5 A in abiotic conditions. For each experiment, the reactor was degassed with N 2 to remove H 2 before electrolysis. The values of the dissolved hydrogen sensor were recorded once a minute until reaching saturation.

Calculations
The k La was calculated according to the adsorbing equation (Myung et al. 2016): Here, C * is the saturated concentration of dissolved hydrogen (mg H 2 L − 1 ), K L is the mass transfer coefficient (cm·h − 1 ), and a is the gas/liquid interfacial area per volume of liquid (cm 2 ·cm − 3 ).
The CE was calculated as described earlier (Liu et al. 2015), to reflect the productivity of the system: (1) Here, ΔC HAc (mol·L -1 ) is the change in the concentration of acetate during the experiment, V solution 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.

Results and discussion k La determination and dissolved H 2 concentration
The K La was measured under different conditions to determine the mass transfer rate of dissolved H 2 at the gas-liquid interface (Fig. 3B). At 0.5 A, the K La of the reactor added with SiO 2 NPS reached 0.49 h − 1 , which was 32.4% higher than that reached without SiO 2 NPS (0.37 h − 1 ). The relation between the dissolved H 2 concentration and the electrolysis time is shown in Fig. 3A. With the addition of SiO 2 NPS, the saturated dissolved H 2 concentration increased from 1.04 mg·L − 1 to 1.11 mg·L − 1 at 0.5 A. All these results confirmed that adding SiO 2 NPS to the catholyte could improve the mass transfer of the dissolved H 2 in the gas-liquid interface and increase the saturated dissolved H 2 concentration of the reactor.
The kinetics of H 2 oxidation by hydrogenase is ratedependent for H 2 -induced CO 2 reduction. As H 2 has a limited solubility of 0.79 mM (Liu et al. 2016) in water at ambient conditions. Based on the information obtained from our experiments, the H 2 K La of the reactors could be increased by introducing SiO 2 NPS. The improvement of the mass transfer coefficient by the interaction between nanoparticles and the gas-liquid interface could be Overall charge (C) × 100% Fig. 3 The dissolved H 2 concentration curves (A) and the K La of H 2 (B) of reactors at 0.5 A under abiotic conditions Pan et al. Bioresources and Bioprocessing (2023) 10:3 explained by three theories: a shuttling or grazing effect, hydrodynamic influences at the gas-liquid interface, and changes in the specific gas-liquid interfacial area (Kim et al. 2014). The diameter of the nanoparticles was much smaller than that of the gas-liquid boundary layer, which is usually 5-25 μm. Therefore, it was likely that shuttling or grazing effects played an important role in improving gas-liquid mass transfer (Kim et al. 2014).

Acetate production
As shown in Fig. 4B, the acetate production rate was lower than 0.50 g·L − 1 ·d − 1 in the first 2 days, but it increased sharply in the coming days. When SiO 2 NPS were present in the reactor, the acetate concentration increased from 0.29 g·L − 1 to 13.1 g·L − 1 (2-8 days), and the average acetate production rate reached 2.14 g·L − 1 ·d − 1 in batch 1. In contrast, the acetate production rate and the final acetate concentration of the experiment without SiO 2 NPS were only 1.16 g·L − 1 ·d − 1 and 7.22 g·L − 1 in batch 1, respectively. The second batch displayed a comparable production profile, but the acetate titer outperformed the first batch. The maximum acetate concentration and average acetate production rate (10-18 days) of the experiments in batch 2 with (without) particles were 18.5 (11.8) g·L − 1 and 2.14 (1.26) g·L − 1 ·d − 1 , respectively. The performance of the second batch was superior to the first batch in terms of acetate titer due to the enrichment of more acetate-tolerating microorganisms during the long-term operation. In batch 2, the acetate production rate with nanoparticles was 1.7 times higher than that without nanoparticles at the stable increment stage (10-18 days) of acetate, and the titer of acetate increased by 56.9%. All these results suggested that the addition of SiO 2 NPS could improve both the acetate production rate and titer of acetate by enhancing the mass transfer of the dissolved H 2 at the gas-liquid interface and increasing the saturated dissolved H 2 concentration in the reactor. This could be attributed to the fact that enhanced gasliquid mass transfer would normally increase the amount of biomass in the reactor (Xue et al. 2022).

Electron balance
Based on the VFAs concentration and the H 2 left in the reactor, the electron balance was calculated. Most of the electrons from the cathode ended up in acetate and H 2 , while the rest could be attributed to other VFAs and biomass (Fig. 5). In general, the percentage of electrons that went to acetate increased at first and then decreased over time, whereas the percentage of electrons that ended in H 2 decreased at first and then increased over time. In batch 1, the average percentages of electrons that went to the acetate of the reactor with and without SiO 2 NPS were 36% and 19%, respectively. In the stable increment phase (10-18 days) of batch 2, the average percentage of electrons went to acetate, other VFAs and biomass, and H 2 were 44%, 41%, and 15% with SiO 2 NPS, while those without SiO 2 NPS were 26%, 28%, and 46%, respectively. Therefore, the addition of SiO 2 NPS could make more electrons go to acetate instead of H 2 .
In the stable increment phase, the reactor without SiO 2 NPS reached its throughput bottleneck, because 50% of the electrical H 2 was not utilized by the culture due to the limitation of H 2 solubility in water. In contrast, the addition of SiO 2 NPS can alleviate the throughput bottleneck Fig. 4 The acetate concentration (A) and acetate production rate (B) of the MES reactor in two batches Pan et al. Bioresources and Bioprocessing (2023) 10:3 by reducing the limitation and making almost all H 2 be utilized by the culture.

CO 2 uptake efficiency
The CO 2 uptake efficiency (Fig. 6) was consistent with electron balance. To facilitate the acclimation of the microbial community, the reactor was operated under 0.25 A in the first two days. And the CO 2 uptake efficiency of the reactor with SiO 2 NPS reached 100%.
Then, the applied current and the flow rate of CO 2 were doubled. As shown in Fig. 6, the CO 2 uptake efficiency with SiO 2 NPS was over 60% except for some slight fluctuations, while the CO 2 uptake efficiency without SiO 2 NPS was between 30% and 70%. After that, the rapidly decreased CO 2 uptake efficiency might be caused by product inhibition during the last two days of each batch. All these results indicated that the addition of SiO 2 NPS could increase the CO 2 uptake efficiency of the reactor. In a broader context, SiO 2 NPS may improve gas solubilities in water for a series of small non-polar gases, such as N 2 and CH 4 , whose solubilities in an aqueous medium are also rate-limiting. We propose that SiO 2 NPS are also potentially applicable for biologically catalyzed N 2 fixation and CH 4 functionalization (Rodrigues et al. 2019), two challenging processes in small molecule activation.

Comparison to other MES reactors
The performance of this H 2 -mediated MES reactor was compared to other reported MES reactors in Table 1. Most studies were devoted to improving biofilm formation by developing effective cathodes. Apparently, the current density and acetate production rate of 3D porous cathodes were much higher than those of 2D cathodes in most reported MES reactors. It was simply because the 3D porous cathode could provide a higher specific surface area for the microorganisms and formed a thicker biofilm on the electrode. To our best knowledge, the acetate production rate of all reported MES reactors was less than 1.06 g·L − 1 ·d − 1 , and the maximum acetate concentration was around 10 g·L − 1 . By contrast, the acetate production rate and acetate titer of this reactor containing SiO 2 NPS could reach 2.14 g·L − 1 ·d − 1 and 18.5 g·L − 1 , respectively, which were about twice those of the reported reactors.
These results could be partly attributed to the better mass transfer to suspended cells in the reactor. Compared to biofilm, the suspended cells had lower limitations At 0.25 A and 0.5 A, the CO 2 flow rates were 1.96 mL·min − 1 and 3.92 mL·min − 1 , respectively Pan et al. Bioresources and Bioprocessing (2023) 10:3 caused by the substrate and product diffusion and electrode surface area, which could significantly improve the production rate of the bacteria. Furthermore, the addition of SiO 2 NPS further increased the mass transfer rate of dissolved H 2 . However, nanoparticles could aggregate and gravitationally settle down in the stagnant flow region (dead zone) in the reactor. Moreover, the behavior of nanoparticles penetrating the cell cytoplasm may be toxic to the cell (Hwang et al. 2008). To obtain the best performance of acetate production, the optimum type and concentration of nanoparticles need to be further studied. In addition, many strategies, such as developing 3D structures and modifying functional groups on the surface of the nanoparticle could be utilized to increase the mass transfer rate and recyclability of nanoparticles.
In the future, we could centrifuge the effluent to test whether it is possible to recover the silica nanoparticles for reuse, and we could also modify the silica nanoparticles into magnetic silica nanoparticles to reuse them.

Conclusion
In summary, this work developed a novel strategy, i.e., adding SiO 2 NPS, to improve the H 2 solubility in the H 2 -mediated MES reactor. We found that the addition of SiO 2 NPS in the MES system could significantly increase the H 2 mass transfer rate and maximum saturated H 2 solubility in water to accelerate the production rate and titer of acetate. With the addition of 0.3wt% SiO 2 NPS, the H 2 K La of the reactor increased by 32.4% at 0.5 A. The titer of acetate in batch 2 of the reactor with SiO 2 NPS (18.5 g·L − 1 ) was 56.9% higher than the reactor without SiO 2 NPS (11.8 g·L − 1 ). In the stable increment phase, the average acetate production rate of the reactor with SiO 2 NPS reached 2.14 g·L − 1 ·d − 1 , which was much higher than other reports. This pioneering effort made with the addition of SiO 2 NPS provided insights for the development of effective microbial electrosynthesis applications. In a broader context, the nanoparticle can also be utilized in other H 2 -dependent reactions, such as biologically catalyzed N 2 fixation, CH 4 functionalization and microbial protein production.