English

• 1. Introduction

  In the past decades, with the increase of population and the extensive use of fossil fuels, CO₂ emissions have caused severe greenhouse effect [1-3], which has led to glacier melting, sea level rise, climate change, ocean storms and other environment problems [4,5]. It affects the normal life order and poses a great threat to human beings. In addition, the Paris Agreement points out that the global net zero emission of greenhouse gasses should be achieved by the end of this century [6]. Therefore, it is an urgent need to prevent global warming by capturing CO₂ and reduce the man-made carbon cycle. Carbon capture, utilization and storage (CCUS) by chemical method is considered as one of the most effective ways to reduce greenhouse gasses in the short term [7,8]. Based on the concept of CCUS, the conversion of CO₂ into high value-added products by electrochemistry has aroused widespread interest in industry and academia [9].

  Some non-electrochemical and relatively mature CO₂ capture methods have been applied in large-scale industrial applications [10], such as biological method [11,12], adsorption method [13,14], membrane method [15], ammonia absorption method [16,17] and low temperature capture [18]. However, these methods face challenges such as high energy consumption, limited environmental recycling sites, energy and materials loss of regenerated adsorbents [6,19,20].

  CO₂ capture and transformation by electrochemical method could be used as an efficient solution to overcome these problems. Electrochemical method is not limited to capture the CO₂ from point sources, but also capture and transform the carbonate from air, ocean, waste gas and liquid. The modular nature of the electrochemical process could offer multiple cell stacks, resulting in low-cost structural materials for interconnections, gas diffusion layers and catalytic layers [21]. In addition, such BPMs for CO₂ capture could realize a smaller floor area and be easily scale-up as needed [22]. The catalytic efficiency of CO₂ capture and transformation dependents on the product distribution and the electrochemical capture ratio of CO₂ source [23], which shows the unique advantages of electrochemical method.

  For the CO₂ capture and transformation of BPMs, the main driving force of the BPMs system is the electrochemical potential gradient. There is no need for external heat source or pressurization. Therefore, CO₂ capture could be carried out under isothermal condition by controlling the potential gradient [23,24], which is beneficial to reduce the demand for energy and cost. Usually, the acid-base capture of CO₂ by pH oscillation [25-27] and electrocatalytic reduction with transition metal (such as Zn [28], Cu [29,30] and Ni [31–33]) are two main approaches for electrochemical CCUS of CO₂ [23]. By the pH change of solution with the BPMs, CO₂ could be captured by local alkaline environment and finally converted into carbonate/bicarbonate, which CO₂ could be stored. As the reutilization of CO₂, the pH of the solution converted from high pH to low pH, the desorption of CO₂ could be realized under acid condition. Similar pH oscillation includes capacitive ion storage [34], proton coupled electron transfer (PCET) [35,36] and bipolar membrane electrodialysis (BPMED) [37,38].

  Bipolar membrane (BPM) is a new kind of ion exchange membrane which could dissociate water into OH⁻ and H⁺ by electrolysis [39]. BPM consists of anion exchange layer, cation exchange layer and catalytic layer (such as Al (OH)₃) between them [40,41]. When the BPMs is charged, the water molecules dissociated at the active sites of the catalytic layer into H⁺ and OH⁻. The high potential gradient would drive H⁺ to the cathode through the cation exchange layer and OH⁻ to the anode through the anion exchange layer. Hence, bipolar membrane electrochemical device based on the concept of pH oscillation could be used for CO₂ capture (such as bipolar membrane electrodialysis (BPMED)) [42,43]. In recent years, the electrocatalytic reduction of CO₂ into different high value-added products using renewable energy sources as electricity has received increasing attention from researchers. The conventional H-type electrolytic cell for electrocatalytic CO₂ reduction has a current density of less than 100 mA/cm² due to the lower mass transfer limitation. The membrane electrode electrolyzer (MEA) combined with membrane technology is considered as a promising component for industrial CO₂ reduction due to its ability to provide higher current density and higher selectivity. However, at the same time, the type of membrane has a key influence on the current density and selectivity of MEA electrolyzer, such as cation exchange membranes (CEM) membrane electrode electrolyzer which provides high current density but also brings serious hydrogen precipitation reaction, and anion exchange membranes (AEM) membrane electrode electrolyzer which reduces hydrogen precipitation reaction while allowing the cathode product to cross over to the anode for re-oxidation. Since the BPMs could prevent the crossover of H⁺ and OH⁻ and stabilize the pH value of electrolyte [44,45], thus improving ion transfer ability in electrolyte, electrochemical CO₂ reduction reaction (ECO₂RR) based on the BPMs could improve the electrochemical reduction performance of CO₂ [46]. However, commercial BPMs generally have non-negligible drawbacks such as high resistance, easy delamination and blistering. The thicker film layer is one of the reasons for the high resistance of BPMS compared to monopolar films [47]. Therefore, the fabrication of thinner anion exchange membrane (AEL) and cation exchange membrane (CEL) by improving BPMs thickness will result in high conductivity and selectivity of the overall BPMs, while decreasing delamination of BPMs requires improving AEL and CEL interfacial compatibility (e.g., developing a three-dimensional CEL-AEL interface). Thus, the active contact area and interfacial stability could be increased [48,49].

  In general, the BPMs have great potential for electrochemical capture and utilization for CO₂ (Fig. 1). In this review, the bipolar membrane electrochemical capture and more robust CO₂ utilization processes are introduced in detail. Different types of BPM membranes and devices design are investigated and compared. In addition, the challenges and opportunities are also discussed for the development of the BPMs structure design and reductive equipment assembling for the CO₂ capture and transformation.

  Figure 1

  

  Figure 1.  Schematic diagram of electrochemical capture of CO₂ based on bipolar membrane.

  DownLoad: Full-Size Img PowerPoint

  2. Basic reaction mechanism: pH oscillation and electrocatalytic reduction reaction

  Bipolar membrane electrodialysis (BPMED) uses the pH swing of working fluid to capture and separate CO₂ at ambient temperature and pressure, and to recover CO₂ by continuously changing the pH value of working fluid between alkaline and acidic pH to affect CO₂ balance [27]. As shown in the Figs. 2a and b, be reversible and the recycling of CO₂ could be realized by changing the carbonate balance. Electrocatalytic reduction of CO₂ is much more complex reduction process than acid-base CO₂ capture with BPMs, which contains multiple charge transfer reactions, usually involving 2~18 electron transfer pathways, many products, complex mechanism and high overpotential. The products include C₁ products (CO, CH₄, HCOOH, CH₃OH), C₂ products (C₂H₄, C₂H₅OH, C₂H₆, CH₃COOH, CH₃CHO) and C₃ products (C₂H₅CHO, C₃H₇OH), etc. Table 1 summarizes the main CO₂RR products at standard atmospheric pressure, 25 ℃ and the equilibrium potentials in aqueous solution with pH 7 (vs. SHE) [50]. CO₂ is first activated in aqueous solution into adsorbed CO₂(ads), and then CO₂(ads) is reduced by one electron to form CO₂ anion radicals (^·CO₂⁻) (Fig. 2c). ^·CO₂⁻ is formed at potential of −1.90 V (vs. SHE), which is a key step to determine the reaction rate. After that, the C—O bond of the formed anion radical is broken and combined with protons or carbon atoms to form C—H or C—C bond [51]. When the Hg, Pb, In, Sn and other metals are used as catalyst, formic acid could be selectively generated, the oxygen of the anion radical coordinates with metal active center, and the carbon atom combines with proton to form C—H bond and get two electrons to form formate [52]. On the contrary, when Zn, Au, Ag and other metal catalysts are used, CO is selectively generated. The carbon atom of the anion radical coordinates with the active center of the metal catalysts [53]. When the Cu is used as catalyst, CO₂(ads) could be desorbed or further reduced by combining four electrons in the presence of protons, thus obtaining an adsorbed: CH₂ species. The surface methylene (: CH₂) reacts with two other electrons and protons to form CH₄, or with another : CH₂ to form ethylene [51].

  Figure 2

  

  Figure 2.  (a) Relationship between carbon dioxide equilibrium and pH (for a closed system with a temperature of 25 ℃ and a salinity of 35 ppt). (b) Schematic diagram of pH fluctuations. Reproduced with permission [27]. Copyright 2020, Royal Society of Chemistry. (c) Reaction mechanism for the preparation of different products by electrochemical reduction of CO₂. Reproduced with permission [51]. Copyright 2007, Royal Society of Chemistry.

  DownLoad: Full-Size Img PowerPoint

  Table 1

  

  Table 1.  Reaction equations and electrode potentials of different reduction products [50].

  DownLoad: CSV

  3. Bipolar membrane electrodialysis (BPMED) method

  When bipolar membranes are used with CEM and AEM, they could form BPMED with different compartments, which different compartments could convert salts in corresponding aqueous solutions to acids and bases (Fig. 3) [54]. BPMED could be used in the environmental protection to capture CO₂ from seawater [55], flue gas [56], atmosphere [6] and the solutions containing carbonates. With electrochemical pH oscillation, Bandi et al. use the BPMED to regenerate absorbent and capture CO₂ from air [42]. This method uses KOH solution to absorb CO₂ and the generated K₂CO₃ is acidified by sulfuric acid. Then, K₂CO₃ solution enters the BPMED unit, which alkaline washing solution and sulfuric acid were finally obtained. In addition, BPMED uses stacked structure, which makes it possible to obtain high purity for CO₂ capture. BPMED mainly uses two compartments and three compartments. Two compartments compose of B-C-B (BPM-CEM-BPM), B-A-B (BPM-AEM-BPM) and three compartments compose of B-C-A-B (BPM-CEM-AEM-BPM). Different membrane stack configurations of BPMED have great differences in energy consumption, efficiency, yield and product purity. The main characteristics of the two compartments are low energy consumption and small volume of compartments for acid and base. The main drawbacks of the three compartments are more membrane layer, high membrane resistance and high energy consumption. However, since the AEM and CEM do not directly contact with acid-alkali solution, the current efficiency and the product purity could be increased [57,58]. The specific methods would be discussed in detail for CO₂ capture in the next section.

  Figure 3

  

  Figure 3.  Schematic diagram of B-C-A-B triple compartment bipolar membrane electrodialysis (BMED) unit for conversion of NaCl.

  DownLoad: Full-Size Img PowerPoint

  3.1 Extract carbon dioxide from seawater

  The ocean has absorbed about 40% of the anthropogenic carbon dioxide emitted since the industrial age [59,60], which its concentration is 2.1 mmol/kg [61], far exceeding the concentration (about 140 times) of carbon dioxide in the air [62]. This means that if CO₂ is captured from the air, a large volume of air would be processed, which brings significant cost challenges for carbon capture. However, the separation of carbon dioxide from the ocean by balancing total dissolved inorganic carbon (DIC) with the air would be an attractive way to capture carbon [63,64]. Carbon dioxide mainly exists in the form of HCO₃⁻ ions in the ocean. Electrodialysis (ED) has already proved that it could be used to recover carbon dioxide from electrodialysis of alkaline carbonate solution (Fig. 4a) [43,65]. However, this method with only ion exchange and no acidification to recover CO₂ has poor ion selectivity and low conductivity caused by diffusion and migration of carbonate/bicarbonate species, which are the main limitations of this method.

  Figure 4

  

  Figure 4.  (a) Removal of dissolved CO₂ and desalination of saline solution using electrodialysis cell. Reproduced with permission [65]. Copyright 2017, Elsevier. (b, c) BPMED schematics of CO₂ generation from seawater. (d) BPMED energy consumption for CO₂ generation from seawater. Reproduced with permission [66]. Copyright 2012, Royal Society of Chemistry.

  DownLoad: Full-Size Img PowerPoint

  BPMED is expected to solve the shortage of electrodialysis. Bicarbonate/carbonate could be converted into H₂CO₃ (pH oscillation) by using the acid chamber in the bipolar membrane devices containing BPMs, and then extracted into CO₂ by membrane contactor in vacuum [25,27]. Based on this idea, Eisaman et al. [66] proposed a new prototype design, the constructed BPMED could efficiently extract dissolved inorganic carbon from seawater. The absorption rate of CO₂ could reach 59% with the electrochemical energy consumption of 242 kJ/mol (CO₂), as shown in Figs. 4b–d. However, hydrogen evolution (HER) and oxygen evolution (OER) occur simultaneously in the cathode and anode during the electrolysis process, which would lead to extra electrochemical energy loss. Digdaya et al. [26] further proposed to replace the original HER and OER in cathode and anode with reversible redox reaction.

  Cyclically reversible redox coupling K₃/K₄[Fe(CN)₆)] solution was used as cathode and anode solution, thus significantly reduced polarization and electrochemical energy loss. The capture efficiency of CO₂ could reach 71% and the electrochemical energy consumption is 155.4 kJ/mol. However, the voltage drop with the BPMs is significant. The slow hydrodynamics causes the voltage loss.

  In addition, another method of using BPMED to capture CO₂ in seawater is mineral carbonization, and CO₂ mineralization is beneficial in thermodynamics [55], which determines that it could capture CO₂ with lower cost and energy consumption. CO₂ mineralization is the chemical combination of CO₂ with Ca²⁺/Mg²⁺ and other metal ions to form stable mineralized substances (such as CaCO₃, MgCO₃) [25,55]. In the process of capturing CO₂ and mineralizing metal ions, the system needs higher pH value (9.6–10) [67]. Compared with adding chemical alkali agents to the system, BPMED has shown the advantages of lower cost and no secondary pollution [55,68]. Zhao et al. [69] proposed a four-compartment bipolar membrane electrodialysis, as shown in Fig. 5a. Combined with calcium carbonate crystallizer, the problem of membrane fouling was solved, which decreased energy consumption during CO₂ carbonizing. Using cheap and easily available seawater, carbonate ions generated by CO₂ and BPMED alkali chamber migrate to salt chamber through cation exchange membrane, where they combine with seawater. The membrane pollutants are controlled to keep in ion state by adjusting CO₂ to enter alkali chamber instead of salt chamber. Meanwhile, hydroxide is consumed immediately and migrated to alkali chamber, which is one of the key factors to prevent membrane fouling. As shown in Figs. 5b and c, the CaCO₃ crystallizer and seed crystallization are also important to prevent membrane fouling. Experiments show that the fixation rate of CO₂ could reach 50.1%, and the decalcification rate could reach 73.4%. As shown in Fig. 5d, in addition, Zhao et al. [56] changed its gas-liquid contactor into a tower bubbling reactor by using the above system. The capture and utilization of acid gas (CO₂-SO₂) in flue gas were investigated. It was found that SO₂ in flue gas decreased carbon fixation rate, but increased decalcification rate. When current density was 12.5 A/m², gas velocity was 40 L/h and CO₂ in flue gas was 10%, carbon fixation rate was 31.4%, decalcification rate was 94.5%, and SO₂ in flue gas could be completely removed (100%).

  Figure 5

  

  Figure 5.  (a) Schematic diagram of four-chamber bipolar membrane electrodialysis. AEM membrane surface SEM of salt chamber side under (b) alkali chamber ventilation and (c) salt chamber ventilation. (d) Trend charts of decalcification rate, carbon sequestration rate and energy consumption under different experimental conditions. Reproduced with permission [69]. Copyright 2020, Elsevier.

  DownLoad: Full-Size Img PowerPoint

  Based on the above research results, to understand the operating mechanism of the process and to optimize the reaction conditions, Chen et al. [70] studied the competitive migration and reaction rate among OH⁻, HCO₃⁻ and CO₃²⁻, which experimentally determined the reaction rate of HCO₃⁻ and OH⁻ was greater than their migration rate. Therefore, the gas-liquid contact device could effectively improve the absorption rate of CO₂. By adding gas refining device, the aeration efficiency is greatly increased. Under the optimum conditions, the carbon fixation rate could reach 100% and the total energy consumption is 1.54 kWh/kg. Therefore, Sharifian et al. [71] directly pump seawater into the alkali chamber of BPEMD, instead of the previous capture mode with high purity alkali solution (high pH). 60% DIC and at least 16% Ca²⁺ were captured from seawater. The energy consumption for generating calcium carbonate was 0.88 kWh/kg. The irreversible overpotential of BPMs accounted for 55% of the theoretical value. However, membrane fouling is serious, which limits the large-scale use. To reduce the membrane fouling during the mineralization, development of bipolar membrane with optimized reaction kinetics and membrane thickness still remains a challenge.

  3.2 Carbon sequestration by high pressure BPMED

  When the bipolar membrane electrodialysis system captures CO₂ from seawater or carbonate solution, working under normal pressure and high current density would lead to CO₂ precipitation in the membrane stack, which would stay in the membrane stack and adhere to the membrane surface in the form of bubbles, thus affecting the actual effective membrane area of bipolar membrane electrodialysis. This would result in increased resistance, increased electrochemical energy consumption, damage to the membrane surface and decreased membrane life.

  In addition, the concentration also determines whether CO₂ is precipitated or not [72]. Low concentration could effectively prevent gas from precipitating in the membrane stack and it also directly leads to the current density being reduced to 10 mA/cm² or even below. In order to solve this serious problem, Eisaman [72] and others proposed a high-pressure bipolar membrane electrodialysis. The designed device controls the pressure of acid, alkali and electrode solution by adjusting needle valves in the flow path, and adjusts the pressure in the range of 1.5–10 atm (Fig. 6). Experiments show that increasing the pressure could effectively prevent CO₂ precipitation in bipolar membrane electrodialysis stack. At a high current density of 139 mA/cm², the desorption energy at high voltage operation (6 atm) is reduced by 29% compared with that at approximately normal pressure operation (1.5 atm).

  Figure 6

  

  Figure 6.  (a, b) Schematic diagram of the high-pressure BPMED experimental setup used to solve CO. (c, d) Energy (kJ/mol CO₂) and CO₂ flow rate (slpm) versus pressure. Reproduced with permission [72]. Copyright 2011, Royal Society of Chemistry.

  DownLoad: Full-Size Img PowerPoint

  Furthermore, in order to accurately evaluate the high-pressure bipolar membrane electrodialysis proposed by Eisaman et al. and Sabatino et al. [6] used MATLAB to model the BPMED unit, in order to evaluate and improve the process performance by means of sensitivity analysis and process optimization. The data show that the model could well reproduce the experimental results, in which Eqs. 1 and 2 accurately describe the relationship between bubble generation and conductivity. The analysis shows that the rapid CO₂ generation rate would lead to the generation of CO₂ bubbles, the non-conductive phase would increase the resistance of the electrolytic cell, and the composition would also affect the conductivity, thus affecting the voltage drop of the electrolytic cell.

  (1)

  (2)

  where A is the parameter of bubble shape, B is the parameter of its conductivity, φ_CO2 is the volume fraction of CO₂ in the acid solution, and φ_m is the maximum volume fraction attainable [6].

  3.3 Microbial fuel cell-bipolar membrane electrodialysis (BPMED-MFC)

  Microbial electrochemical technology is a potential green method for sustainable production of electricity or chemical fuels [73]. Its sustainable energy recovery provides a broad prospect for electrochemical conversion of CO₂ into high value-added products [74]. CO₂ is a thermodynamically stable molecule with short bond length and high bond energy, so it needs high energy input to transform CO₂ molecules, which requires external energy to be carbon neutral, thus realizing green recovery of CO₂ [75]. Microbial fuel cell (MFC) is a typical microbial electrochemical system (MES), which could convert wastewater into renewable electricity by microorganisms. Combining MFC with electrodialysis of BPMs shows the potential of high cost-effectiveness and environmental friendliness for CO₂ capture [74,76].

  Based on the concept of pH oscillation, CO₂ adsorption by alkali (NaOH) has the characteristics of low pressure and low energy consumption. Electrodialysis with BPMs could be integrated with MFC to produce alkali to capture CO₂ impurities in biogas, so as to purify biogas. Chen et al. [77] used MFC as power combined with BPMED to capture CO₂ in biogas in alkali production. As shown in Fig. 7, the bioconversion substrate on the anode of MFC is consumed by electricity-producing anaerobic microorganisms, which generate electrons and protons.

  Figure 7

  

  Figure 7.  Schematic diagram of the AEM-CEM-BPM configuration of the BPMED-MFC: (A) Desalination chamber and (B) alkali production chamber. Reproduced with permission [77]. Copyright 2013, Elsevier.

  DownLoad: Full-Size Img PowerPoint

  The electrons are transferred to the cathode electron acceptor through external resistance, and the protons are transferred to the cathode chamber through cation exchange membrane, thus providing power for BPMED. The research proves that the alkali solution prepared by BPMED-MFC has a good CO₂ capture effect, and the CH₄ concentration in biogas could reach 100%. However, the voltage required to use BPMs is higher than that generated by the device with biological cathode and anode. Therefore, it is necessary to provide additional energy from the outside [75], which has also been confirmed by Chen et al. [77]. Zhu et al. [75] designed a microbial reverse electrodialysis electrolysis and chemical production cell (MRECC), which uses reverse electrodialysis (RED) stack as power supply. This device could generate electricity through salinity gradient in wastewater [76,78], avoiding the usage for external voltage. This new MRECC system mineralizes CO₂ into carbonate products. Additional H₂ could be produced by combining the generated acids and bases with minerals, which would hopefully reduce the cost of capturing CO₂. The results show the cost of capturing CO₂ per ton is $25, which is lower than that of other mature methods.

  Further, to recycle HCO₃⁻/CO₃²⁻ ions formed by CO₂ absorption by alkali solution, Jin et al. [79] proposed a microbial electrolytic capture separation regeneration cell (MESC). The anode and cathode of the regeneration tank are separated by the BPMs and AEM, and the middle chamber is an acid chamber. BPMs keep the anode pH stable to prevent anode bacteria from growing. The generated H⁺ migrates to the acid chamber through the cation exchange layer of BPMs, and the alkaline solution produced by the cathode is used to capture CO₂ in biogas. The generated HCO₃⁻/CO₃²⁻ migrates to the acid chamber through AEM driven by potential difference to combine with H⁺ for CO₂ regeneration. This combination of systems has been proven to be useful for capturing, separating and regenerating CO₂ from biogas, as well as for wastewater utilization. In addition, the newly proposed separation and regeneration system has a lower voltage for water decomposition than that of the abiotic electrolysis system. The results show the power consumption for biogas is 0.09 kWh/Nm³. The H₂ generated at the cathode makes up about 27% of the cost.

  Most of the reported BPMED-MFC systems use NaCl as electrolyte, which increases its operating costs. Therefore, seawater and high-salt wastewater could be combined in actual operating process. In addition, wastewater could increase the conductivity of the working fluid, which could further reduce the operation cost. However, the selection and exploration of related materials have not been carried out, and its interference to other components in wastewater needs to be further explored, which still remains a challenge and an opportunity.

  4. Electrocatalytic reduction method

  Electrochemical CO₂ reduction reaction (CO₂RR) is another way to convert CO₂ into value-added fuels and high value-added chemicals by using renewable energy sources (wind energy and solar energy) [80,81], such as CO [82-84], formic acid [85], methanol [86], methane [87], ethylene [88,89], ethanol [90], acetic acid [91], propanol [92]. Energy storage in chemical products could reduce dependence on traditional fossil energy, thus reducing greenhouse gasses. It also helps to solve the problem of power misallocation due to seasonal and climatic factors [93]. In addition, CO₂RR system is modular structure. Its structure is relatively simple, which could be operated at normal temperature and pressure. Meanwhile, different reduction products can be changed by controlling the reduction potential.

  The CO₂RR process is normally realized by an electrochemical CO₂ electrolytic cell, which is composed of a working electrode, a reference electrode and a three-electrode/two-electrode system. An ion exchange membrane is also used for separating the cathode chamber and the anode chamber to prevent the product from being reoxidized. In recent years, there are many studies on catalytic materials for electrochemical CO₂ reduction reaction electrode (CO₂RR). For C₁ products (such as formic acid, CO, methanol), Lin et al. [94] prepared homogeneous intermetallic Pd₃Bi nanocrystals by solvothermal method. Nearly 100% formate selectivity could be obtained for reversible hydrogen electrode at −0.35 V in 0.1 mol/L KHCO₃ solution. The crystal sequence of Pd and Bi atoms in the catalyst could inhibit the side reactions of CO. Ag is usually used for the electrochemical reduction of CO₂ to CO, however, monometallic catalysts suffer from slow kinetics, low product selectivity and poor stability, etc. Bimetallic catalysts made by doping with other metals are significantly different from monometallic catalysts. The newly doped metal of bimetallic catalysts not only changes the electronic structure of the raw material, but also provides more active sites and improves the structural stability of the active ingredient [95,96]. Zhang et al. [97] proposed a "two ships in a bottle" Zn-Ag-O ternary catalyst design (Figs. 8a–f), for the reduction of CO₂ to CO, achieving an energy efficiency of 60.9% and achieving a Faraday electrolysis efficiency of 94.1% ± 4.0% for CO (Figs. 8g–i). By thermodynamic DFT analysis and in situ X-ray absorption, it was shown that the delocalization of electrons stabilized the intermediate product *COOH, increasing the intrinsic CO activity and increasing the energy barrier for the formation of HCOO* and *H, thus inhibiting the occurrence of formic acid and hydrogen side reactions, resulting in enhanced selectivity for CO. Ren et al. [98] reported a strategy of implanting non-precious metal-active Sn-ZnO heterogeneous surfaces into the nanopores of high surface area carbon nanospheres (Sn-ZnO@HC, Figs. 9a–d) to reduce CO₂ to CO with a Faraday electrolysis efficiency of up to 91% at −0.53 V for CO and stability up to 240 h (Figs. 9e and f). With the help of density functional theory (DFT) and finite element method (FEM) calculations, the electronic interactions between Sn and ZnO reduced the Zn-d bandwidth and increased the Sn-d bandwidth, which facilitated the fracture of the *COOH intermediate. Hence, this bimetallic oxide heterostructure further promotes the CO₂ reduction reaction kinetics due to the mutual synergistic effect between the two metals [95]. Wei et al. [99] synthesized a multi-stage porous Ag nano-foam (AgNF) modified by SCN on Ag surface. The maximum faraday electrolysis efficiency of AgNF for CO reach 97%, which commonly maintains above 90% in the potential range of −0.5 ~ −1.2 V (vs. RHE). The main reason is the SCN ligand could promote the formation of COOH* intermediates by changing the local electron density on Ag surface. For C₂ products (such as ethylene, acetaldehyde), Zahid et al. [100] prepared oxidized derived copper (ODCu) electrode by hydrothermal method. The electrode has high selectivity for C₂₊ products, and the Faraday electrolysis efficiency of C₂₊ products could reach 58% at −0.95 V. Zhu et al. [101] prepared three-dimensional dendritic copper-cuprous oxide composites by reductive electrodeposition. The contact resistance between the electrocatalyst and composite matrix is almost zero. The dendritic structure helps to expose abundant active sites of Cu. Therefore, the overpotential for acetic acid and ethanol is only 0.53 V and 0.48 V. The Faraday electrolysis efficiency for C₂ products could reach over 80% at 11.5 mA/cm².

  Figure 8

  

  Figure 8.  "Two Ships in a Bottle" design concept and CO₂RR performance. (a) Schematic diagram of the "two ships in a bottle" catalyst. (b) SEM image of ZnO-Ag@UC. (c) TEM image of ZnO-Ag@UC. (d) HRTEM image of ZnO-Ag@UC. (e) Ag(111) and (f) ZnO(101). (g) LSV curve of catalyst. (h) FE distribution of ZnO-Ag@UC on product. (i) FE of catalyst on CO. Reproduced with permission [97]. Copyright 2021, American Chemical Society.

  DownLoad: Full-Size Img PowerPoint

  Figure 9

  

  Figure 9.  (a) Schematic diagram of Sn-ZnO@HC catalyst preparation (blue, gold, red, white, and gray atoms represent Sn, Zn, O, H and C elements). (b) SEM image of Sn-ZnO@HC. (c) TEM image of Sn-ZnO@HC. (d) Histogram of size distribution. (e) FE distribution of Sn-ZnO@HC on reduction products. (f) FE of different catalysts on FE of the partial current density of CO. Reproduced with permission [98]. Copyright 2022, Wiley.

  DownLoad: Full-Size Img PowerPoint

  From above research, most research focused on cathodes and anodes preparation and performances for electrochemical reduction of CO₂. However, the assembling of electrocatalytic reduction of CO₂ system also need to be introduced, since the structure design of electrolytic cell is also a key factor which affects current density, product selectivity, Faraday electrolysis efficiency and so on [93,102]. The typical H-type electrolytic cell has simple structure and low cost, which could be widely used. The drawback of the H-type electrolytic cell is slow mass transfer, large electrode space and small electrode area. Hence, its current density is far from the current density required by industry (≥ 200 mA/cm²). Continuous flow electrolytic cell has the advantages of reducing the transfer path, and reducing the mass transfer limitation when used with gas diffusion layer (GDL) [45]. To increase the catalytic performance of continuous flow electrolytic cell, polymer electrolyte membrane plays an important role, which could be divided into cation exchange membranes/proton exchange membranes (CEMs) [103], AEMs [104,105] and BPMs [106], etc. These polymer electrolyte membranes could prevent cathodic reduction products (such as formic acid, ethanol) from diffusing and migrating to anode to be reoxidized, thus avoiding the problem of reducing electrolysis efficiency of electrolytic cell. Different membranes determine different transfer types, which would directly affect the selectivity of products and Faraday electrolysis efficiency. For example, anion exchange membrane AEM would transfer cathode reactants (CO₃²⁻/HCO₃⁻) to anode, thus reducing CO₂ conversion efficiency. Cation exchange membrane would make H⁺ migrate to the cathode, resulting in cathode acidification environment, which is beneficial to the HER side reaction [103] and greatly reduces CO₂ conversion efficiency. Bipolar membrane is expected to solve these problems and provide a stable cathode/anode pH environment. Therefore, in this section, the influence of membrane properties on the selectivity and reduction efficiency of electrolytic cells is discussed. The structure and mass transfer of electrolytic cells with different cells design are compared and analyzed. The development of potential BPM-based electrolytic cells is emphatically summarized and discussed.

  4.1 H-type electrolytic cell and microfluidic cell

  Microfluidic electrolytic cell is composed of cathode gas diffusion electrode loaded with CO₂ reduction catalyst and anode gas diffusion electrode loaded with oxidation catalyst [107-109]. The two diffusion electrodes are separated by an extremely thin electrolyte channel. The distance is usually less than 1 mm. During electrochemical reduction, CO₂ gas enters the cathode and reaches the surface of gas diffusion electrode (GDE) to be reduced by cathode catalyst. O₂ produced by anode oxygen evolution is emitted into the air. Whipple et al. [110] first proposed a microfluidic reactor for electrochemical reduction of CO₂. As shown in Fig. 10a, an Ag/AgCl reference electrode is inserted into the electrolyte between two gas diffusion electrodes to detect the electrode potential. This reactor is used to quickly evaluate the catalyst and explore the influence of electrolyte pH on the reduction efficiency of carbon dioxide to formic acid. Experiments show that working at acidic pH value could significantly improve the performance of this electrolytic cell. When Sn/C is used as the cathode catalytic material, the Faraday efficiency (FE) of formic acid prepared by CO₂ is 89% and the energy efficiency reach 45% (current density 100 mA/cm²). As shown in Fig. 10b, the flowing electrolyte flow is beneficial to change the pH and composition of electrolyte. The ultra-thin electrolyte channel also reduces the ohmic polarization loss of the cell. In addition, compared with the membrane-based flow cell, it could also reduce the permeation resistance of water molecules across the membrane and the concentration gradient caused by proton transfer across the membrane, thus avoiding the cathode flooding. Furthermore, Wu et al. [111] used microfluidic electrolyzer to reduce CO₂ to CO and established steady-state isothermal microfluidic cell mathematical model for microfluidic cell to reduce CO₂ to CO, which confirmed CO₂ reduction kinetics depends on CO₂ concentration at active sites to a great extent. This model is important to enhance CO₂ transfer rate to the catalytic layer (CL). In addition, CO₂ concentration in feed, gas feed flow rate, channel length and GDE porosity all have different effects on performance. Although microfluidic cell has great potential in electrochemical reduction of CO₂, it also has many drawbacks. Liquid reduction products produced in cathode would diffuse in electrolyte to anode and be reoxidized, which leads to the decrease of reduction efficiency.

  Figure 10

  

  Figure 10.  (a) Schematic diagram of the microfluidic reactor used for CO₂ conversion. (b) Efficiency of different catalysts used in the microfluidic reactor. Filled symbols indicate Faraday efficiency, while symbols indicate energy efficiency. Reproduced with permission [110]. Copyright 2010, Electrochemical Soc Inc.

  DownLoad: Full-Size Img PowerPoint

  In order to avoid the reduction products migrating to the anode and being reoxidized, the ion exchange separation membrane between the cathode and the anode is expected to solve this problem. Typical H-type electrolytic cell is still the most commonly used CO₂RR reactor in experiments. The cathode and anode chambers of H-type electrolytic cell are separated by an ion exchange membrane that only allows corresponding ions to pass through. The cathode chamber contains working electrodes and reference electrodes, and the anode chamber contains counter electrodes. The current density is usually less than 100 mA/cm² [112-114] due to the low solubility of CO₂ in water solution. In addition, the small electrode area and large electrode spacing also affect the overall electrolysis efficiency. Therefore, to overcome problems such as low mass transfer efficiency, it is necessary to adopt ultra-thin spacing and GDE of microfluidic cell. Ion exchange membrane also needs to be introduced to overcome the cross diffusion between cathode and anode products, which ensure continuous flow of electrolyte and reduce the limitation of CO₂ mass transfer, thus increasing the current density of the cells. The specific discussion would be presented in the next section.

  4.2 CO₂ capture by MEA solid electrolyte

  4.2.1   MEA configuration electrolytic cell based on CEM

  Membrane electrode assembly (MEA) electrolytic cell is a new type of electrolytic cell. MEA electrolytic cell is a CO₂ reduction reactor formed by pressing polymer electrolyte membrane together with cathode and anode [115-117]. As shown in Fig. 11a, the continuously humidified CO₂ stream is directly supplied to the cathode. CO₂ reduction occurs at the boundary between the membrane and the cathode electrode. Gas-phase membrane-based CO₂RR reactor provides low current density by constructing electrochemical cells composed of membrane electrode assemblies (MEA), which solves the problem of low solubility and slow diffusion of CO₂ in water. Compared with liquid-phase CO₂RR reduction, it shows higher product selectivity and energy efficiency [45,115,118]. The addition of polymer electrolyte membrane effectively reduces the cross between reactants and products between cathode and anode, thus reduces the risk of catalyst poisoning caused by impurities in cathode electrolyte. Polymer electrolyte membranes used in MEA CO₂ reduction electrolytic cell are mainly divided into AEM, CEM and BPM, and their ion transfer mechanism is shown in Figs. 11b–d [119]. Delacourt et al. [103] first proposed the use of cation exchange membrane (Nafion®117) and the construction of MEA electrochemical cells. The cathode humidified CO₂ and the anode used pure water as electrolyte. At the current density of 20 mA/cm², no CO₂ reduction product was detected at the gas outlet, but the current efficiency of HER was close to 100%. The main reason is mainly due to the migration of H⁺ from the anode to the cathode through the cation exchange membrane, which promoted the cathode selectivity to the side reaction of HER. In addition, it is also related to electrode poisoning and electrolyte property change. In order to alleviate the influence of acidic environment of cathode on CO₂RR, a 0.5 mol/L KHCO₃ buffer layer supported by glass fiber was introduced between cathode and cation exchange membrane [103]. At the same current density (20 mA/cm²), the CO selectivity was greatly improved and the FE of CO was about 82%. The addition of buffer layer slowed down the arrival of H⁺ to cathode, thus reducing the occurrence of HER. In addition, in the Nafion-based electrochemical cells, cations (such as K⁺) would migrate from the anode side to the cathode in the Nafion-based electrolytic cell at pH < 1, which cations would affect the catalytic kinetics of CO₂RR [120,121]. In the process of CO₂RR based on proton exchange membrane using high concentration phosphate buffer (1 mol/L H₃PO₄), the increase of K⁺ concentration would weaken the selectivity of HER. At the higher current density of 400 mA/cm², when local protons are exhausted (local pH > 7), the increase of K⁺ concentration could tune the selectivity of HER to CO₂RR and increase the selectivity of CO₂RR [122]. The mechanism is that cations trigger CO₂ activation in the environment where protons almost disappear, thus inhibiting HER of water reduction [122,123]. On the other hand, the migration of metal cations from anode to cathode would cause the transfer of cathode electrolyte, which would increase the purification and separation cost of downstream electrolyte. In addition, metal cations migrating to the cathode would react with cathode carbonate and bicarbonate to form salt deposition at the interface between the membrane and the catalytic layer, thus preventing CO₂ gas from diffusing into the catalyst [124,125]. Adding sufficient water to the cathode could decrease salt deposition, but this would cause product dilution and GDE flooding in the cathode, which would limit the mass transfer of CO₂ and increase the cost of concentrating downstream products [126-128]. Therefore, improving water management system is a promising solution to salt deposition in CO₂RR based on cation exchange membrane. When the selective product is formic acid or formate, CEMs could prevent the migration of formate anions to the anode and prevent the product from being further oxidized at the anode [129,130]. In addition, the thickness of cation exchange membrane is low, so its potential drop in CO₂RR process is lower than that of other polymer electrolyte membranes (such as BPM) [131,132]. As shown in Figs. 11e and f, the initial cell potential of MEA configuration electrolytic cell based on bipolar membrane is 4 V, while the cell potential based on Nafion is 2.7 V [124]. Temperature also has an important influence on the voltage of CO₂RR electrolytic cell with MEA configuration. When the operating temperature of the electrolytic cell rises, the electrolytic cell voltage decreases [133].

  Figure 11

  

  Figure 11.  (a–c) CEM, AEM and BPM ion transport mechanisms used for MEA configuration CO₂ reduction electrolytic cell. (d) Schematic diagram of the membrane electrode configuration electrolytic cell structure. Reproduced with permission [133]. Copyright 2019, Royal Society of Chemistry. (e) Comparison between the BPM-based and Nafion-based zero-gap electrolyzers in terms of timekeeping potential. (f) FE of analytical products of BPM-based and Nafion-based zero-gap electrolyzers. Reproduced with permission [124]. Copyright 2011, Wiley.

  DownLoad: Full-Size Img PowerPoint

  When the cell temperature rises from 30 ℃ to 80 ℃, the electrolytic cell voltage decreases from 2.5 V to 2.2 V, shown in Fig. 12a. The main reason is the kinetic effect of electrode reaction and ion conduction could be improved when the temperature increases. The reduction of high frequency resistance does not decrease the corresponding voltage potential (Fig. 12b). However, the increase of temperature would also adversely affect the membrane. Cation exchange membranes based on perfluorosulfonic acid (such as Nafion) have low proton conductivity at high temperature and could be decomposed [134,135]. In addition, fluorine-based membranes are complicated and expensive, hydrocarbon-based membranes could be an alternative design. The introduction of -SO₂H- group between carbon atom and sulfur atom of sulfonic acid group reduces the activation energy and increases the stability of the film. The main reason is that SO₂ is inserted between carbon atom and sulfur atom in neutral and anionic states, which leads to the withdrawal of electrons by SO₂ group and charge distribution, resulting in uniform and flat charge distribution [136]. Therefore, introducing highly electronegative groups between carbon atoms and sulfur atoms of sulfonic acid groups could be used as an important way to prepare fluorine-free hydrocarbon base membranes.

  Figure 12

  

  Figure 12.  (a) Effect of operating temperature on the reduction of MEA CO₂RR electrolytic cell to CO cell voltage and FE. (b) AC impedance spectrum recorded at circuit voltage, inset zooms in on the high frequency region to show the variation of ion conduction with temperature. Reproduced with permission [133]. Copyright 2019, Royal Society of Chemistry.

  DownLoad: Full-Size Img PowerPoint

  4.2.2   MEA configuration electrolytic cell based on AEM

  MEA electrolytic cell based on AEM is more widely used than MEA electrolytic cell based on cation exchange membrane. On the one hand, AEMs prevent H⁺ from migrating to the cathode catalyst surface through fixed cationic groups, thus avoiding undesired HER. On the other hand, AEMs transfer anions (such as OH⁻) from cathode to anode through fixed cationic groups, and CO₂RR occurs in alkaline environment due to the continuous formation of cathode hydroxide. CO₂ stream would react with OH⁻ to form HCO₃⁻ and CO₃²⁻. These anions would migrate to the anode through AEMs (called "cross"). Compared with OH⁻, their charge mobility is lower and the transfer rate is slower, which leads to the decrease of ionic conductivity of AEMs [104]. HCO₃⁻ and CO₃²⁻ generated are neutralized by H⁺ at the anode and converted back to CO₂ again. The neutralized CO₂ accounts for 50%−60% of the CO₂-consuming gas stream [105,137]. In addition, these anions crossed to the anode would be oxidized back to CO₂ at the anode [138,139] and mixed with the product O₂ of the anode, which makes this part of CO₂ unable to be directly reutilized and increases the cost of downstream gas separation. The standard potential at 298 K and relative to the standard hydrogen electrode (SHE) [E⁰] [139] is summarized in Table 2. When the cathode target product is formate, the produced formate would also reach the anode electrolyte through AEMs, which reduces the concentration of the product and makes it difficult to purify the product [140].

  Table 2

  

  Table 2.  Standard potentials for cathode and anode electrochemical reactions [139].

  DownLoad: CSV

  Excessive water accumulation (water flooding) on porous electrodes is undesirable in the CO₂RR. Cathode flooding occurs when the amount of water transferred from anode to cathode exceeds the required amount of CO₂RR. The characteristics of AEMs are related to the water management of MEA electrolytic cell, and then affect the occurrence of flooding [119]. The use of low water absorption, thin AEM membrane (≤ 40 µm) and hydrophobic cathode has been proved to reduce water flooding [141]. In addition, transmembrane transfer of water could promote the migration of cations (such as K⁺) in cationic electrolyte across AEMs to cathode. These cations combine with cathode HCO₃⁻ and CO₃²⁻ and precipitate as K₂CO₃ and KHCO₃. The solubility of K₂CO₃ and KHCO₃ is lower than that of KOH, which would deposit on the surface of porous gas diffusion electrode. These precipitates hinder the reactants from reaching the catalyst layer and limit the migration of reactants, which affect the CO₂RR performance [104,142]. But cations at the cathode promote CO₂RR through local electrostatic effect [143,144]. Therefore, ideal AEMs hope to have high ionic conductivity, high ionic mobility and low permeability of products [93], thus reducing the occurrence of side reactions and improving the selectivity of products. Gabardo et al. [145] first proposed a strategy to apply AEMs to MEA (Figs. 13a and b), and to efficiently reduce CO₂ to C₂₊while carbon nanoparticles and graphite powder sprayed on the cathode surface of the Cu nanoparticles were used to achieve efficient removal of the concentrated liquid products (ethanol and n-propanol) from the exit gas stream. The anode is made of titanium mesh loaded with IrO₂. This design strategy achieves 50% and 80% Faraday electrolysis efficiencies for C₂H₄ and C₂₊ products (Figs. 13c–e). For AEMs, developing AEMs with higher performance is in urgent need to improve CO₂ reduction efficiency. Zou et al. [146] successfully synthesized a series of BC-PDDA-OH⁻ (PDDA: diallyl dimethyl ammonium chloride) membranes with different mass ratios by impregnation, chemical crosslinking and ion exchange with bacterial cellulose (BC) membrane as skeleton and polyquaternary ammonium salt compound.

  Figure 13

  

  Figure 13.  Schematic diagram of the structure of AEM-MEA and its CO₂RR performance. (a) Schematic diagram of the local structure and cathode structure of AEM-MEA. (b) Schematic diagram of the overall AEM-MEA. (c) Product selectivity of 250 nm thick Cu catalyst layer. (d) Liquid product selectivity at different voltages. (e) Stability of the AEM-MEA electrolytic cell Reproduced with permission [145]. Copyright 2019, Elsevier.

  DownLoad: Full-Size Img PowerPoint

  The results showed that when the mass ratio of BC to PDDA was 1:0.5, the ionic conductivity was 23.87 mS/cm. The ionic conductivity of OH⁻ was even higher (28.5 mS/cm) after soaking in 0.5 mol/cm KHCO₃ for 720 h. When these membranes are applied to the electrochemical reduction of CO₂, they have high selectivity for formate. As shown in Fig. 14, FE could reach 50.8%. The electrochemical performance is better than that of Nafion and basic A901 membranes in 0.5 mol/L KHCO₃/0.5 mol/L KOH electrolyte. The reason is that BC-PDDA-OH⁻ membrane could promote the kinetics of CO₂RR, inhibit HER and reduce the cross of formate products to anode, thus improving the performance of electrolytic cell. Wang et al. [147] reported a series of polymer composite AEMs based on polyvinyl alcohol and guar hydroxypropyl triammonium chloride (PGG). The effects of three functional groups in PGG membrane on the efficiency and selectivity of CO₂ reduction to formate were investigated. The performance of PGG-GT (GA and TCA as binary crosslinking agents) membrane containing thiophene group for formate was better than that of PGG-GT membrane containing dimethyl octanaldehyde and hydroxybenzyl group, and surpassed that of commercial Nafion 212 membrane. Similar to AEMs reported by Zou et al. [146], the high efficiency and selectivity of PGG-GT membranes for formate is mainly due to the higher ionic conductivity and lower product cross-over rate of PGG-GT containing thiophene groups, which could promote the ion transfer kinetics and inhibit HER during CO₂ reduction. Therefore, these different types of new AEMs have made new contributions to the scale-up of CO₂ electrochemical reduction.

  Figure 14

  

  Figure 14.  (a) Material production process flow diagram and schematic diagram of the preparation process of BC-PDDA-OH⁻ membrane. (b, c) FE of different membranes under 0.5 mol/L KHCO₃ and 0.5 mol/L KOH electrolytes. Reproduced with permission [146]. Copyright 2021, Elsevier.

  DownLoad: Full-Size Img PowerPoint

  4.2.3   MEA electrolytic cell based on BPM

  Although AEM-based devices have a wide range of applications and excellent CO₂ electrochemical conversion performance, HCO₃⁻/CO₃²⁻ species crossover from cathode to anode via AEM causes severe carbon loss. This has been identified as a major limitation of AEM-based devices [118,148]. MEA electrolytic cell with BPMs could prevent ion crossing better than that of AEMs electrolytic cell [75,106]. This BPMs-based electrolytic cell could work in two modes of "forward bias" and "reverse bias". When BPMs is in forward bias, CEL side of BPMs is close to anode with high potential and AEL side is close to cathode with low potential (Fig. 15a). AEL of BPMs could provide OH⁻ to the cathode. Meanwhile, it could provide an alkaline environment for the cathode and inhibit the HER without buffer layer [88]. Blommaert et al. [149] found that when BPMs was placed in forward bias mode, the Ag cathodic potential decreased 3 V at 25 mA/cm² compared with reverse bias mode since there was no water dissociation in the interlayer (Fig. 15b).

  Figure 15

  

  Figure 15.  (a) Schematic of species migration within the MEA with forward-biased BPM. (b) Voltage stability of C₂H₄ and forward-biased BPM at 50 mA/cm². (c) Photograph of BPM with forward-biased mode after running at 50 mA/cm² current density for 2 h. The BPM membrane blistered at the AEM: CEM interface. Reproduced with permission [88]. Copyright 2021, American Chemical Society. (d) Schematic diagram of species migration within the MEA with reverse-biased BPM..

  DownLoad: Full-Size Img PowerPoint

  In addition, anions in the CEL are repelled by Donnan repulsion of fixed negative charge group in the layer (e.g., sulfonic acid anion) driven by the potential difference [150]. The cations (e.g., H⁺) are transferred to the hydrolysis CL. Similarly, the anions (such as HCO₃⁻/CO₃²⁻) are transported to the CL layer through the AEL. These ions combine with cations (H⁺) to form CO₂ and H₂O again, which leads to the stratification of AEL and CEL at the junction of the membrane, resulting in the decrease of stability (Fig. 15c). Hence, electrolysis efficiency and reactant CO₂ also decreased in the intermediate layer [88,118].

  When BPMs is in reverse bias, the AEL side of BPMs is close to the anode with high potential, while the CEL side is close to the cathode with low potential. As shown in Fig. 15d, reverse bias is traditionally used to promote the dissociation of water and stabilize the pH values of electrochemical cells. In the reverse bias mode, hydrated ions migrate from the interface layer (IL) to both sides of the cathode and anode, thus depleting the moving charge on the film. To provide the required ion current and to keep the charge neutral, the dissociation of water could be further triggered [47]. The dissociation of water makes OH⁻ get alkaline environment through AEL, which avoids using noble metal OER catalysts (such as Pt and IrO₂) [46,92]. Abundant oxygen evolution catalysts on earth could be used, such as Ni and FeNiO_x. But at the same time, this also reduces the concentration of moving ions in the interface layer and increases the BPMs resistance [150,151].

  Based on these characteristics of bipolar membranes, to break the limitation of 50% conversion of CO as a target product, Yang et al. [152] used Ag as cathode and nickel foam as anode with the help of BPMEA (Figs. 16a and b). A high concentration of KOH electrolyte was delivered at the anode to allow the anode cation (K⁺) to cross the BPM to reach the cathode catalyst surface. The results showed that this approach improved the CO₂ reduction selectivity of the BPMEA system by a factor of 3, and the CO Faraday electrolysis efficiency reached 68% at a current density of 50 mA/cm² (Fig. 16c). Sinton et al. [88] developed permeable and renewable layer (PCRL) combined with CEMs for the instability of BPMs (Fig. 16d). This PCRL layer is made by spraying an anion-exchange ionomer solution (Aemion AP1-CNN5–00-X) to a certain loading level, which provided an alkaline environment on the catalyst surface combined with the acidic environment provided by CEMs. MEA CO₂ reduction electrolytic cell could be assembled with IrO₂ as anode and Cu catalyst as gas diffusion electrode. Local regeneration of CO₂ is realized, which is similar to the bipolar membrane structure, but avoids the undesired stratification of BPMs. The CO₂ crossover was controlled within 15% of the conversion product (Figs. 16e and f). The single CO₂ conversion rate reached 85% at 100 mA/cm² combined with low flow rate and low crossover, as shown in Fig. 16g. To address the previous limitation of only 25% single utilization of C₂₊ in neutral electrolytic cells, Xie et al. [153] designed an extremely thin cathode layer model (SC layer, consisting of a commercial porous support, material can be polyvinylidene difluoride (PVDF), poly tetra fluoroethylene (PTFE), polycarbonate (PC)) with the help of BPMEA, as shown in Fig. 16h. A porous support layer (2 cm × 2 cm, with different thicknesses) was placed on the cathode surfaces to achieve thin cathode, which was sonicated for 15 min in a saturated ideal electrolyte to degas before use. It is possible to limit the diffusion path length of CO₂ to 10 µm, and this cathode layer could recover CO₂ from the carbonate produced by the reaction, thus overcoming the problem of reactant loss in the carbonate. The combination of CuNPs cathode and IrO₂ anode achieves a single CO₂ utilization rate of 78%, while the highest current density is up to 300 mA/cm² (Figs. 16i–k). Wang et al. [105] reported a strategy of porous solid electrolyte reactor, a MEA CO₂ cell was constructed by forming a permeable ionic conductive sulfonated polymer electrolyte between AEM and PEM as a buffer layer with Ag nanowire as cathode and IrO₂ as anode. The cross carbonate could be combined with protons to form CO₂ gas again. Generated CO₂ could be recovered by flushing porous solid electrolytes (PSE) layer with deionized water. At the current of 200 mA/cm², the recovery of cross CO₂ and FECO could reach 90%. Mandal et al. [81] assembled strongly acidic CEL on polyelectrolyte layer by layer (LBL) containing polyacrylic acid/polyallylamine hydrochloride (PAH/PAA) and formed a weakly acidic layer with the function of cation exchanger, thus forming a new type of BPM. The combination of LBL BPMs and weak acid CEL in gas-supplied CO₂ electrolyte could inhibit HER activity and increase FE. As shown in Fig. 17, the FECO of LBL BPMs is higher than that of Nafion BPM. Its current density increases significantly at high potential.

  Figure 16

  

  Figure 16.  Schematic diagram of different structures of BPM-MEA and their CO₂RR performance. (a) Schematic diagram of BPMEA structure with high concentration anode electrolyte. (b) Schematic diagram of ion transport in BPMEA system. (c) The relationship between different current densities and electrolytic cell voltages under 3 M KOH anode electrolyte and the FE for CO. Reproduced with permission [152]. Copyright 2021, American Chemical Society. (d) Schematic diagram of MEA with PCRL structure. (e) Comparison of CO₂ flux in PCRL anode with AEM and CEM. (f) CO₂ flux distribution at PCRL coating thickness of 2.25 mg/cm². (g) Conversion efficiency and FE distribution at 2.25 mg/cm² coating thickness. Reproduced with permission [88]. Copyright 2021, American Chemical Society. (h) Schematic diagram of SC-BPMEA structure and mass transfer. (i-k) FE distribution of products with different cathode liquid thickness. Reproduced with permission [153]. Copyright 2022, Springer Nature.

  DownLoad: Full-Size Img PowerPoint

  Figure 17

  

  Figure 17.  (a) Schematic diagram of the modified BPM structure by LBL assembly. (b) CO FE versus cathodic potential for the ionomer-free catalyst layer. (c) CO fraction current density versus cathodic potential. Here the LBL BPM contains 10 PAH/PAA bilayers. the decrease in CO current density of Nafion BPM at higher potentials may be caused by water immersion in the ionomer-free catalyst layer. Reproduced with permission [81]. Copyright 2021, Springer Nature.

  DownLoad: Full-Size Img PowerPoint

  To sum up, the design and synthesis of BPMs materials have an important impact on electrochemical reduction of CO₂. There is still plenty of room for designing more efficient BPMs. At the same time, to improve the performance of CO₂ reduction electrochemical cell, catalyst selectivity, electrolyte type and operating conditions should be considered as a whole.

  5. Conclusions and prospects

  Electrochemical CO₂ capture is more convenient, sustainable and efficient than non-electrochemical methods, which bipolar membranes could play an important role in affecting acid/base production, low ion crossover and resource recovery in the electrochemical CO₂ capture process. This paper provides a detailed review of the bipolar membrane electrochemical CO₂ capture and transformation processes design with comparative discussions of the types of membranes and device types. Table 3 summarizes the development and comparison of different CO₂RR reactor designs.

  Table 3

  

  Table 3.  Comparison of different structure CO₂RR electrolytic cell development and design. CO₂ reduction reactor types.

  DownLoad: CSV

  (1) Make CO₂ capture and electroreduction more efficient and selective. CO₂ process design are discussed in detail and the types of membranes and devices are also compared. BPMED changes the equilibrium absorption of CO₂ through pH oscillation. Combined with mineralized substances in seawater, it has broad application prospects for seawater desalination. However, membrane fouling could not be ignored. The generation of non-conductive phases such as CO₂ bubbles also leads to the increased overall resistance and external power consumption. Therefore, it is necessary to further optimize operating conditions and processes. Developing bipolar membranes with higher rate for water dissociation and optimal membranes thickness is also important to solve these problems.

  As for the CO₂ reduction, the preparation of electrocatalysts is undoubtedly extremely important for the selectivity of electrocatalytic reduction of CO₂ and the Faraday electrolysis efficiency of the products. While designing the catalysts, the structure of different catalysts as well as the type of hybridization have important effects on the performance of catalysts for CO₂ reduction. For example, bimetallic catalysts have higher active sites and structural stability compared to monometallic catalysts, while the design of heterostructures also stimulates the mutual synergistic effect between metals to promote the kinetics of the reduction reaction. Another important aspect is the design of the electrochemical reduction electrolyzer. The BPMEA has advantages such as low ion crossover rate, stable cathode pH compared to other membrane electrode (AEM/CEM). In terms of product selectivity, suppression of unwanted HER generation could improve product selectivity. In addition, the BPM has a lower cathodic potential using forward bias than reverse bias. Reverse bias could have better water dissociation rate and CO₂RR stability compared to forward bias.

  (2) Make CO₂ capture and electroreduction more facile and safer. The BPM-based reactor avoids the use of toxic or expensive chemicals for CO₂ capture by pH oscillation. Cheap, widely available, nontoxic solutions are used as feedstocks, such as sodium chloride and seawater. The modular stack reactor also increases the ease of CO₂ capture. The BPM-based CO₂RR reactor allows the use of inexpensive and widely available non-precious metal anodes, which avoids the use of expensive precious metal.

  Overall, BPMs have great potential for bipolar membrane electrodialysis and CO₂RR electrolytic cells. The combination of bipolar membrane electrodialysis for on-site independent CO₂ capture and BPMs-based MEA electrocatalytic conversion of CO₂ to high value-added products is highly desired to achieve the goals of greenhouse gas reduction and sustainable energy, which may be a direction of exploration to combine the two into a continuous capture and conversion flow device. However, further studies are needed to understand their overall function. The structural modification of the bipolar membrane, such as coating the cationic membrane with polyelectrolyte or trying to expel CO₂ from the catalytic interlayer, could decrease cathode and anode layer decomposition and swelling during the electrolysis of bipolar membrane and increase the BPMs stability. In addition, the development of BPMs with high water dissociation rates, low ion crossover, strong interfacial stability and long lifetime is crucial for future electrochemical CO₂ capture and transformation.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This research was funded by the National Natural Science Foundation of China (Nos. 52272303 and 52073212), the National Key Research and Development Program of China (No. 2018YFC1602400), the General Program of Municipal Natural Science Foundation of Tianjin (Nos. 17JCYBJC22700 and 17JCYBJC17000), and the State Scholarship Fund of China Scholarship Council (Nos. 201709345012 and 201706255009).

  
  
• 1. [1]

     W. Rong, H. Zou, W. Zang, et al., Angew. Chem. Int. Ed. 60 (2021) 466–472. doi: 10.1002/anie.202011836

  2. [2]

     N. Gaulin, P. Le Billon, Clim. Policy. 20 (2020) 888–901. doi: 10.1080/14693062.2020.1725409

  3. [3]

     S. Nanda, S.N. Reddy, S.K. Mitra, et al., Energy Environ. Sci. 4 (2016) 99–122. doi: 10.1002/ese3.117

  4. [4]

     J.H. Mercer, Nature 271 (1978) 321–325. doi: 10.1038/271321a0

  5. [5]

     S.H. Schneider, Science 244 (1989) 904-904. doi: 10.1126/science.244.4907.904.c

  6. [6]

     F. Sabatino, M. Mehta, A. Grimm, et al., Ind. Eng. Chem. Res. 59 (2020) 7007–7020. doi: 10.1021/acs.iecr.9b05641

  7. [7]

     F. Rubiera, C. Pevida, Greenh. Gases 8 (2018) 396–397. doi: 10.1002/ghg.1783

  8. [8]

     F. Sha, N. Zhu, Y. Bai, et al., ACS Sustain. Chem. Eng. 4 (2016) 3032–3044. doi: 10.1021/acssuschemeng.5b01793

  9. [9]

     O.S. Bushuyev, P. De Luna, C.T. Dinh, et al., Joule. 2 (2018) 825–832. doi: 10.1016/j.joule.2017.09.003

  10. [10]

      A.S. Reis Machado, M. Nunes da Ponte, Curr. Opin. Green Sustain. Chem. 11 (2018) 86–90. doi: 10.1016/j.cogsc.2018.05.009

  11. [11]

      A. Anand, S. Raghuvanshi, S. Gupta, Curr. Sustain. /Renew. Energy Rep. 7 (2020) 40–47. doi: 10.1007/s40518-020-00149-1

  12. [12]

      I. Matito-Martos, C. Sepúlveda, C. Gómez, et al., Chem. Eng. J. 417 (2020) 128020.

  13. [13]

      B. Feng, H. An, E. Tan, Energy Fuels. 21 (2007) 426–434. doi: 10.1021/ef0604036

  14. [14]

      R.S. Patil, D. Banerjee, C. Zhang, et al., Angew. Chem. Int. Ed. 55 (2016) 4523–4526. doi: 10.1002/anie.201600658

  15. [15]

      K. Xie, Q. Fu, G.G. Qiao, et al., J. Membr. Sci. 572 (2019) 38–60. doi: 10.1016/j.memsci.2018.10.049

  16. [16]

      L. Li, G. Puxty, M. Maeder, et al., Energy Fuels 33 (2019) 5377–5383. doi: 10.1021/acs.energyfuels.9b00394

  17. [17]

      S. Ma, G. Chen, S. Zhu, et al., Appl. Energy 162 (2016) 354–362. doi: 10.1016/j.apenergy.2015.10.089

  18. [18]

      S.S. Mostafa, J.D. Hedengren, K.M. Powell, Appl. Energy. 228 (2018) 577–592. doi: 10.1016/j.apenergy.2018.06.098

  19. [19]

      M. Fasihi, O. Efimova, C. Breyer, J. Clean. Prod. 224 (2019) 957–980. doi: 10.1016/j.jclepro.2019.03.086

  20. [20]

      D.W. Keith, G. Holmes, D. St. Angelo, et al., Joule 2 (2018) 1573–1594. doi: 10.1016/j.joule.2018.05.006

  21. [21]

      H.K. Ju, G. Kaur, A.P. Kulkarni, et al., J. CO₂ Util. 32 (2019) 178–186.

  22. [22]

      R.A. Shaw, T.A. Hatton, Int. J. Greenh Gas Control. 95 (2020) 102878. doi: 10.1016/j.ijggc.2019.102878

  23. [23]

      S.E. Renfrew, D.E. Starr, P. Strasser, ACS Catal. 10 (2020) 13058–13074. doi: 10.1021/acscatal.0c03639

  24. [24]

      B. Gurkan, X. Su, A. Klemm, et al., iScience 24 (2021) 103422. doi: 10.1016/j.isci.2021.103422

  25. [25]

      C.F. de Lannoy, M.D. Eisaman, A. Jose, et al., Int. J. Greenh. Gas Control. 70 (2018) 243–253. doi: 10.1016/j.ijggc.2017.10.007

  26. [26]

      I.A. Digdaya, I. Sullivan, M. Lin, et al., Nat. Commun. 11 (2020) 4412. doi: 10.1038/s41467-020-18232-y

  27. [27]

      R. Sharifian, R.M. Wagterveld, I.A. Digdaya, et al., Energy Environ. Sci. 14 (2021) 781–814. doi: 10.1039/d0ee03382k

  28. [28]

      J. Chen, Z. Li, X. Wang, et al., Angew. Chem. Int. Ed. 61 (2022) e202111683. doi: 10.1002/anie.202111683

  29. [29]

      T. Qin, Y. Qian, F. Zhang, et al., Chin. Chem. Lett. 30 (2019) 314–318. doi: 10.1016/j.cclet.2018.07.003

  30. [30]

      B. Wang, F. Yang, Y. Dong, et al., Chem. Eng. J. 396 (2020) 125255. doi: 10.1016/j.cej.2020.125255

  31. [31]

      Z.L. Wang, J. Choi, M. Xu, et al., ChemSusChem 13 (2020) 929–937. doi: 10.1002/cssc.201903427

  32. [32]

      X. Yang, J. Cheng, B. Fang, et al., Nanoscale 12 (2020) 18437–18445. doi: 10.1039/d0nr04391e

  33. [33]

      K. Xu, S. Zheng, Y. Li, et al., Chin. Chem. Lett. 33 (2022) 424–427. doi: 10.1016/j.cclet.2021.07.016

  34. [34]

      L. Legrand, Q. Shu, M. Tedesco, et al., J. Colloid Interface Sci. 564 (2020) 478–490. doi: 10.1016/j.jcis.2019.12.039

  35. [35]

      M. Rahimi, G. Catalini, M. Puccini, et al., RSC Adv. 10 (2020) 16832–16843. doi: 10.1039/d0ra02450c

  36. [36]

      H. Xie, W. Jiang, T. Liu, et al., Cell Rep. Phys. Sci. 1 (2020) 100046. doi: 10.1016/j.xcrp.2020.100046

  37. [37]

      M.D. Eisaman, L. Alvarado, D. Larner, et al., Energy Environ. Sci. 4 (2011) 1045–1528. doi: 10.1039/c1ee90009a

  38. [38]

      W. Ye, J. Huang, J. Lin, et al., Sep. Purif. Technol. 144 (2015) 206–214. doi: 10.1016/j.seppur.2015.02.031

  39. [39]

      Y. Wang, C. Huang, T. Xu, J. Membr. Sci. 374 (2011) 150–156. doi: 10.1016/j.memsci.2011.03.026

  40. [40]

      B. Bauer, F.J. Gerner, H. Strathmann, Desalination 68 (1988) 279–292. doi: 10.1016/0011-9164(88)80061-4

  41. [41]

      R. Pärnamäe, S. Mareev, V. Nikonenko, et al., J. Membr. Sci. 617 (2021). doi: 10.1016/j.memsci.2020.118538

  42. [42]

      A. Bandi, M. Specht, T. Weimer, et al., Energy Convers. Manag. 36 (1995) 899–902. doi: 10.1016/0196-8904(95)00148-7

  43. [43]

      H. Nagasawa, A. Yamasaki, A. Iizuka, et al., AIChE J. 55 (2009) 3286–3293. doi: 10.1002/aic.11907

  44. [44]

      Y.C. Li, D. Zhou, Z. Yan, et al., ACS Energy Lett. 1 (2016) 1149–1153. doi: 10.1021/acsenergylett.6b00475

  45. [45]

      D.A. Salvatore, D.M. Weekes, J. He, et al., ACS Energy Lett. 3 (2017) 149–154.

  46. [46]

      D.A. Vermaas, W.A. Smith, ACS Energy Lett. 1 (2016) 1143–1148. doi: 10.1021/acsenergylett.6b00557

  47. [47]

      M.A. Blommaert, D. Aili, R.A. Tufa, et al., ACS Energy Lett. 6 (2021) 2539–2548. doi: 10.1021/acsenergylett.1c00618

  48. [48]

      C. Shen, R. Wycisk, P.N. Pintauro, Energy Environ. Sci. 10 (2017) 1435–1442. doi: 10.1039/C7EE00345E

  49. [49]

      Y. Chen, J.A. Wrubel, W.E. Klein, et al., ACS Appl. Polym. Mater. 2 (2020) 4559–4569. doi: 10.1021/acsapm.0c00653

  50. [50]

      H. Cui, Y. Guo, L. Guo, et al., J. Mater. Chem. A 6 (2018) 18782–18793. doi: 10.1039/c8ta07430e

  51. [51]

      G. Centi, S. Perathoner, G. Win, et al., Green Chem. 9 (2007) 671. doi: 10.1039/b615275a

  52. [52]

      J.S. Yoo, R. Christensen, T. Vegge, et al., ChemSusChem 9 (2016) 358–363. doi: 10.1002/cssc.201501197

  53. [53]

      L.I. Bingyu, Q. Mao, J. Zhao, et al., Chem. Eng. Prog. 38 (2019)4901–4910.

  54. [54]

      A.J.B. Kemperman, Handbook On Bipolar Membrane Technology (Chinese version), Chemical Industry Press, Beijing, 2004.

  55. [55]

      H. Xie, B. Liang, H. Yue, et al., Chem 4 (2018) 24–26. doi: 10.2112/si82-003.1

  56. [56]

      Y. Zhao, L. Wang, Z. Ji, et al., Desalination 494 (2020) 114654. doi: 10.1016/j.desal.2020.114654

  57. [57]

      Y. Li, S. Shi, H. Cao, et al., Water Res. 89 (2016) 201–209. doi: 10.3847/0004-6256/152/6/201

  58. [58]

      Y. Wang, N. Zhang, C. Huang, et al., J. Membr. Sci. 385-386 (2011) 226–233. doi: 10.1016/j.memsci.2011.09.044

  59. [59]

      T. DeVries, M. Holzer, F. Primeau, Nature 542 (2017) 215–218. doi: 10.1038/nature21068

  60. [60]

      C.L. Sabine, T. Tanhua, Ann. Rev. Mar. Sci. 2 (2010) 175–198. doi: 10.1146/annurev-marine-120308-080947

  61. [61]

      N.R. Bates, Front. Mar. Sci. 5 (2018) 398. doi: 10.3389/fmars.2018.00398

  62. [62]

      J.N. Butler, Carbon Dioxide Equilibria Their Applications, Routledge, United States, 1982.

  63. [63]

      H.D. Willauer, F. DiMascio, D.R. Hardy, et al., Ind. Eng. Chem. Res. 51 (2012) 11254–11260. doi: 10.1021/ie301006y

  64. [64]

      H.D. Willauer, F. DiMascio, D.R. Hardy, et al., Ind. Eng. Chem. Res. 53 (2014) 12192–12200. doi: 10.1021/ie502128x

  65. [65]

      S. Dara, M. Lindstrom, J. English, et al., J. CO₂ Util. 19 (2017) 177–184. doi: 10.1016/j.jcou.2017.03.013

  66. [66]

      M.D. Eisaman, K. Parajuly, A. Tuganov, et al., Energy Environ. Sci. 5 (2012) 7346. doi: 10.1039/c2ee03393c

  67. [67]

      W. Wang, M. Hu, Y. Zheng, et al., Ind. Eng. Chem. Res. 50 (2011) 8333–8339. doi: 10.1021/ie1025419

  68. [68]

      E. Koivisto, R. Zevenhoven, Entropy 21 (2019) 395. doi: 10.3390/e21040395

  69. [69]

      Y. Zhao, J. Wang, Z. Ji, et al., Chem. Eng. J. 381 (2020) 122542. doi: 10.1016/j.cej.2019.122542

  70. [70]

      T. Chen, Y. Zhao, Y. Zhao, et al., ACS Sustain. Chem. Eng. 9 (2021) 8372–8382. doi: 10.1021/acssuschemeng.0c09232

  71. [71]

      R. Sharifian, L. Boer, R.M. Wagterveld, et al., Chem. Eng. J. 438 (2022) 135326. doi: 10.1016/j.cej.2022.135326

  72. [72]

      M.D. Eisaman, L. Alvarado, D. Larner, et al., Energy Environ. Sci. 4 (2011) 4031–4037. doi: 10.1039/c1ee01336j

  73. [73]

      Y. Zhang, I. Angelidaki, Water Res. 56 (2014) 11–25. doi: 10.1016/j.watres.2014.02.031

  74. [74]

      Y. Tian, D. Li, C. Li, et al., Chem. Eng. J. 414 (2021) 128671. doi: 10.1016/j.cej.2021.128671

  75. [75]

      X. Zhu, M.C. Hatzell, B.E. Logan, Environ Sci. Technol. Lett. 1 (2014) 231–235. doi: 10.1021/ez500073q

  76. [76]

      X. Li, I. Angelidaki, Y. Zhang, Water Res. 142 (2018) 396–404. doi: 10.1016/j.watres.2018.06.013

  77. [77]

      M. Chen, F. Zhang, Y. Zhang, et al., Appl. Energy 103 (2013) 428–434. doi: 10.1016/j.apenergy.2012.10.005

  78. [78]

      X. Zhu, M.C. Hatzell, R.D. Cusick, et al., Electrochem. Commun. 31 (2013) 52–55. doi: 10.1016/j.elecom.2013.03.010

  79. [79]

      X. Jin, Y. Zhang, X. Li, et al., Environ. Sci. Technol. 51 (2017) 9371–9378. doi: 10.1021/acs.est.7b01574

  80. [80]

      H. Shen, Y. Wang, T. Chakraborty, et al., ACS Catal. 12 (2022) 5275–5283. doi: 10.1021/acscatal.2c00646

  81. [81]

      Z. Yan, J.L. Hitt, Z. Zeng, et al., Nat. Chem. 13 (2021) 33–40. doi: 10.1038/s41557-020-00602-0

  82. [82]

      C.M. Gabardo, A. Seifitokaldani, J.P. Edwards, et al., Energy Environ. Sci. 11 (2018) 2531–2539. doi: 10.1039/c8ee01684d

  83. [83]

      S. Verma, Y. Hamasaki, C. Kim, et al., ACS Energy Lett. 3 (2017) 193–198.

  84. [84]

      S. Verma, X. Lu, S. Ma, et al., Phys. Chem. Chem. Phys. 18 (2016) 7075–7084. doi: 10.1039/C5CP05665A

  85. [85]

      B. De Mot, J. Hereijgers, N. Daems, et al., Chem. Eng. J. 428 (2022) 131170. doi: 10.1016/j.cej.2021.131170

  86. [86]

      N. Kumar, N. Seriani, R. Gebauer, Phys. Chem. Chem. Phys. 22 (2020) 10819–10827. doi: 10.1039/c9cp06453b

  87. [87]

      A.I. Tsiotsias, N.D. Charisiou, I.V. Yentekakis, et al., Nanomaterials 11 (2020) 28. doi: 10.3390/nano11010028

  88. [88]

      C.P. O'Brien, R.K. Miao, S. Liu, et al., ACS Energy Lett. 6 (2021) 2952–2959. doi: 10.1021/acsenergylett.1c01122

  89. [89]

      D. Ren, Y. Deng, A.D. Handoko, et al., ACS Catal. 5 (2015) 2814–2821. doi: 10.1021/cs502128q

  90. [90]

      J. Yuan, J.J. Zhang, M.P. Yang, et al., Catalysts 8 (2018) 171. doi: 10.3390/catal8040171

  91. [91]

      C. Genovese, C. Ampelli, S. Perathoner, et al., Green Chem. 19 (2017) 2406–2415. doi: 10.1039/C6GC03422E

  92. [92]

      M. Ma, K. Djanashvili, W.A. Smith, Angew. Chem. Int. Ed. 55 (2016) 6680–6684. doi: 10.1002/anie.201601282

  93. [93]

      D. Ma, T. Jin, K. Xie, et al., J. Mater. Chem. A 9 (2021) 20897–20918. doi: 10.1039/d1ta06101a

  94. [94]

      L. Jia, M. Sun, J. Xu, et al., Angew. Chem. Int. Ed. 60 (2021) 21741–21745. doi: 10.1002/anie.202109288

  95. [95]

      Q. Li, Y.C. Wang, J. Zeng, et al., Rare Met. 40 (2021) 3442–3453. doi: 10.1007/s12598-021-01772-7

  96. [96]

      Y. Jia, F. Li, K. Fan, et al., Adv. Powder Mater. 1 (2022) 100012. doi: 10.1016/j.apmate.2021.10.003

  97. [97]

      Z. Zhang, G. Wen, D. Luo, et al., J. Am. Chem. Soc. 143 (2021) 6855–6864. doi: 10.1021/jacs.0c12418

  98. [98]

      B. Ren, Z. Zhang, G. Wen, et al., Adv. Mater. 34 (2022) e2204637. doi: 10.1002/adma.202204637

  99. [99]

      L. Wei, H. Li, J. Chen, et al., ACS Catal. 10 (2019) 1444–1453.

  100. [100]

       A. Zahid, A. Shah, I. Shah, Nanomaterials 12 (2022) 1380. doi: 10.3390/nano12081380

  101. [101]

       Q. Zhu, X. Sun, D. Yang, et al., Nat. Commun. 10 (2019) 3851. doi: 10.1038/s41467-019-11599-7

  102. [102]

       J.J. Wang, X.P. Li, B.F. Cui, et al., Rare Met. 40 (2021) 3019–3037. doi: 10.1007/s12598-021-01736-x

  103. [103]

       C. Delacourt, P.L. Ridgway, J.B. Kerr, et al., J. Electrochem. Soc. 155 (2008) B42–B49. doi: 10.1149/1.2801871

  104. [104]

       N. Ziv, W.E. Mustain, D.R. Dekel, ChemSusChem 11 (2018) 1136–1150. doi: 10.1002/cssc.201702330

  105. [105]

       J.Y.T. Kim, P. Zhu, F.Y. Chen, et al., Nat. Catal. 5 (2022) 288–299. doi: 10.1038/s41929-022-00763-w

  106. [106]

       Y.C. Li, Z. Yan, J. Hitt, et al., Adv. Sustain. Syst. 2 (2018) 1700187. doi: 10.1002/adsu.201700187

  107. [107]

       B.A. Rosen, A. Salehi-Khojin, M.R. Thorson, et al., Science 334 (2011) 643–644. doi: 10.1126/science.1209786

  108. [108]

       M.R. Thorson, K.I. Siil, P.J.A. Kenis, J. Electrochem. Soc. 160 (2012) F69–F74.

  109. [109]

       C.E. Tornow, M.R. Thorson, S. Ma, et al., J. Am. Chem. Soc. 134 (2012) 19520–19523. doi: 10.1021/ja308217w

  110. [110]

       D.T. Whipple, E.C. Finke, P.J.A. Kenis, Electrochem. Solid-State Lett. 13 (2010) B109–B111. doi: 10.1149/1.3456590

  111. [111]

       K. Wu, E. Birgersson, B. Kim, et al., J. Electrochem. Soc. 162 (2015) F23–F32. doi: 10.1149/2.1021414jes

  112. [112]

       S. Garg, M. Li, A.Z. Weber, et al., J. Mater. Chem. A 8 (2020) 1511–1544. doi: 10.1039/c9ta13298h

  113. [113]

       J. Liu, P. Li, J. Bi, et al., Chemistry (Easton) 28 (2022) e202200242.

  114. [114]

       D.M. Weekes, D.A. Salvatore, A. Reyes, et al., Acc. Chem. Res. 51 (2018) 910–918. doi: 10.1021/acs.accounts.8b00010

  115. [115]

       F.P. Garcia de Arquer, C.T. Dinh, A. Ozden, et al., Sci. 367 (2020) 661–666. doi: 10.1126/science.aay4217

  116. [116]

       X. Wang, Z. Wang, F.P. García de Arquer, et al., Nat. Energy 5 (2020) 478–486. doi: 10.1038/s41560-020-0607-8

  117. [117]

       L.C. Weng, A.T. Bell, A.Z. Weber, Energy Environ. Sci. 12 (2019) 1950–1968. doi: 10.1039/c9ee00909d

  118. [118]

       A. Pătru, T. Binninger, B. Pribyl, et al., J. Electrochem. Soc. 166 (2019) F34–F43. doi: 10.1149/2.1221816jes

  119. [119]

       E.W. Lees, B.A.W. Mowbray, F.G.L. Parlane, et al., Nat. Rev. Mater. 7 (2021) 55–64. doi: 10.1038/s41578-021-00356-2

  120. [120]

       A. Murata, Y. Hori, Bull. Chem. Soc. Jpn. 64 (1991) 123–127. doi: 10.1246/bcsj.64.123

  121. [121]

       S. Ringe, E.L. Clark, J. Resasco, et al., Energy Environ. Sci. 12 (2019) 3001–3014. doi: 10.1039/c9ee01341e

  122. [122]

       J.E. Huang, F. Li, A. Ozden, et al., Science 372 (2021) 1074–1078. doi: 10.1126/science.abg6582

  123. [123]

       C.J. Bondue, M. Graf, A. Goyal, et al., J. Am. Chem. Soc. 143 (2021) 279–285. doi: 10.1021/jacs.0c10397

  124. [124]

       B. De Mot, M. Ramdin, J. Hereijgers, et al., ChemElectroChem 7 (2020) 3839–3843. doi: 10.1002/celc.202000961

  125. [125]

       M.A. Izquierdo-Gil, V.M. Barragán, J.P.G. Villaluenga, et al., Chem. Eng. Sci. 72 (2012) 1–9. doi: 10.1016/j.ces.2011.12.040

  126. [126]

       P. Jeanty, C. Scherer, E. Magori, et al., J. CO₂ Util. 24 (2018) 454–462. doi: 10.1016/j.jcou.2018.01.011

  127. [127]

       P. Tiwari, G. Tsekouras, G.F. Swiegers, et al., ACS Appl. Mater. Interfaces 10 (2018) 28176–28186. doi: 10.1021/acsami.8b05358

  128. [128]

       G. Díaz-Sainz, M. Alvarez-Guerra, J. Solla-Gullón, et al., J. CO₂ Util. 34 (2019) 12–19. doi: 10.1016/j.jcou.2019.05.035

  129. [129]

       L. Li, A. Ozden, S. Guo, et al., Nat. Commun. 12 (2021) 5223. doi: 10.1038/s41467-021-25573-9

  130. [130]

       W. Ma, S. Xie, X.G. Zhang, et al., Nat. Commun. 10 (2019) 892. doi: 10.1038/s41467-019-08805-x

  131. [131]

       T. Luo, S. Abdu, M. Wessling, J. Membr. Sci. 555 (2018) 429–454. doi: 10.1016/j.memsci.2018.03.051

  132. [132]

       H. Strathmann, A. Grabowski, G. Eigenberger, Ind. Eng. Chem. Res. 52 (2013) 10364–10379. doi: 10.1021/ie4002102

  133. [133]

       Z. Yin, H. Peng, X. Wei, et al., Energy Environ. Sci. 12 (2019) 2455–2462. doi: 10.1039/c9ee01204d

  134. [134]

       M. Rikukawa, K. Sanui, Prog. Polym. Sci. 25 (2000) 1463–1502. doi: 10.1016/S0079-6700(00)00032-0

  135. [135]

       A.L. Rusanov, D. Likhatchev, P.V. Kostoglodov, et al., Proton-exchanging electrolyte membranes based on aromatic condensation polymers, in: Sunny G. (eds. ), Inorganic Polymeric Nanocomposites and Membranes, Springer Berlin Heidelberg, Berlin, 2005, pp. 83–134.

  136. [136]

       F. Akinori, S. Hironori, T. Takashi, Comput. Theor. Chem. 1121 (2017) 44–48. doi: 10.1016/j.comptc.2017.10.008

  137. [137]

       G.O. Larrazabal, P. Strom-Hansen, J.P. Heli, et al., ACS Appl. Mater. Interface 11 (2019) 41281–41288. doi: 10.1021/acsami.9b13081

  138. [138]

       S. Fierro, L. Ouattara, E.H. Calderon, et al., Electrochim. Acta 54 (2009) 2053–2061. doi: 10.1016/j.electacta.2008.06.060

  139. [139]

       J. Herranz, A. Pătru, E. Fabbri, et al., Curr. Opin. Electrochem. 23 (2020) 89–95. doi: 10.1016/j.coelec.2020.05.004

  140. [140]

       M. Ma, E.L. Clark, K.T. Therkildsen, et al., Energy Environ. Sci. 13 (2020) 977–985. doi: 10.1039/d0ee00047g

  141. [141]

       A. Reyes, R.P. Jansonius, B.A.W. Mowbray, et al., ACS Energy Lett. 5 (2020) 1612–1618. doi: 10.1021/acsenergylett.0c00637

  142. [142]

       D.G. Wheeler, B.A.W. Mowbray, A. Reyes, et al., Energy Environ. Sci. 13 (2020) 5126–5134. doi: 10.1039/d0ee02219e

  143. [143]

       S. Ringe, E.L. Clark, J. Resasco, et al., Energy Environ. Sci. 12 (2019) 3609–3610. doi: 10.1039/c9ee90056j

  144. [144]

       B. Endrodi, A. Samu, E. Kecsenovity, et al., Nat. Energy 6 (2021) 439–448. doi: 10.1038/s41560-021-00813-w

  145. [145]

       C.M. Gabardo, C.P. O'Brien, J.P. Edwards, et al., Joule 3 (2019) 2777–2791. doi: 10.1016/j.joule.2019.07.021

  146. [146]

       Q. Zou, X. Guo, L. Gao, et al., Sep. Purif. Technol. 272 (2021) 118910. doi: 10.1016/j.seppur.2021.118910

  147. [147]

       M. Wang, N. Preston, N. Xu, et al., ACS Appl. Mater. Interfaces 11 (2019) 6881–6889. doi: 10.1021/acsami.8b11845

  148. [148]

       J.J. Kaczur, H. Yang, Z. Liu, et al., Front. Chem. 6 (2018) 263. doi: 10.3389/fchem.2018.00263

  149. [149]

       M.A. Blommaert, R. Sharifian, N.U. Shah, et al., J. Mater. Chem. A Mater. 9 (2021) 11179–11186. doi: 10.1039/d0ta12398f

  150. [150]

       J.C. Bui, I. Digdaya, C. Xiang, et al., ACS Appl. Mater. Interfaces 12 (2020) 52509–52526. doi: 10.1021/acsami.0c12686

  151. [151]

       S. Mafé, P. Ramírez, Acta Polym. 48 (1997) 234–250. doi: 10.1002/actp.1997.010480702

  152. [152]

       K. Yang, M. Li, S. Subramanian, et al., ACS Energy Lett. 6 (2021) 4291–4298. doi: 10.1021/acsenergylett.1c02058

  153. [153]

       K. Xie, R.K. Miao, A. Ozden, et al., Nat. Commun. 13 (2022) 3609. doi: 10.1038/s41467-022-31295-3

• 

  Figure 1  Schematic diagram of electrochemical capture of CO₂ based on bipolar membrane.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) Relationship between carbon dioxide equilibrium and pH (for a closed system with a temperature of 25 ℃ and a salinity of 35 ppt). (b) Schematic diagram of pH fluctuations. Reproduced with permission [27]. Copyright 2020, Royal Society of Chemistry. (c) Reaction mechanism for the preparation of different products by electrochemical reduction of CO₂. Reproduced with permission [51]. Copyright 2007, Royal Society of Chemistry.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Schematic diagram of B-C-A-B triple compartment bipolar membrane electrodialysis (BMED) unit for conversion of NaCl.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) Removal of dissolved CO₂ and desalination of saline solution using electrodialysis cell. Reproduced with permission [65]. Copyright 2017, Elsevier. (b, c) BPMED schematics of CO₂ generation from seawater. (d) BPMED energy consumption for CO₂ generation from seawater. Reproduced with permission [66]. Copyright 2012, Royal Society of Chemistry.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) Schematic diagram of four-chamber bipolar membrane electrodialysis. AEM membrane surface SEM of salt chamber side under (b) alkali chamber ventilation and (c) salt chamber ventilation. (d) Trend charts of decalcification rate, carbon sequestration rate and energy consumption under different experimental conditions. Reproduced with permission [69]. Copyright 2020, Elsevier.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a, b) Schematic diagram of the high-pressure BPMED experimental setup used to solve CO. (c, d) Energy (kJ/mol CO₂) and CO₂ flow rate (slpm) versus pressure. Reproduced with permission [72]. Copyright 2011, Royal Society of Chemistry.

  下载: 全尺寸图片 幻灯片
  

  Figure 7  Schematic diagram of the AEM-CEM-BPM configuration of the BPMED-MFC: (A) Desalination chamber and (B) alkali production chamber. Reproduced with permission [77]. Copyright 2013, Elsevier.

  下载: 全尺寸图片 幻灯片
  

  Figure 8  "Two Ships in a Bottle" design concept and CO₂RR performance. (a) Schematic diagram of the "two ships in a bottle" catalyst. (b) SEM image of ZnO-Ag@UC. (c) TEM image of ZnO-Ag@UC. (d) HRTEM image of ZnO-Ag@UC. (e) Ag(111) and (f) ZnO(101). (g) LSV curve of catalyst. (h) FE distribution of ZnO-Ag@UC on product. (i) FE of catalyst on CO. Reproduced with permission [97]. Copyright 2021, American Chemical Society.

  下载: 全尺寸图片 幻灯片
  

  Figure 9  (a) Schematic diagram of Sn-ZnO@HC catalyst preparation (blue, gold, red, white, and gray atoms represent Sn, Zn, O, H and C elements). (b) SEM image of Sn-ZnO@HC. (c) TEM image of Sn-ZnO@HC. (d) Histogram of size distribution. (e) FE distribution of Sn-ZnO@HC on reduction products. (f) FE of different catalysts on FE of the partial current density of CO. Reproduced with permission [98]. Copyright 2022, Wiley.

  下载: 全尺寸图片 幻灯片
  

  Figure 10  (a) Schematic diagram of the microfluidic reactor used for CO₂ conversion. (b) Efficiency of different catalysts used in the microfluidic reactor. Filled symbols indicate Faraday efficiency, while symbols indicate energy efficiency. Reproduced with permission [110]. Copyright 2010, Electrochemical Soc Inc.

  下载: 全尺寸图片 幻灯片
  

  Figure 11  (a–c) CEM, AEM and BPM ion transport mechanisms used for MEA configuration CO₂ reduction electrolytic cell. (d) Schematic diagram of the membrane electrode configuration electrolytic cell structure. Reproduced with permission [133]. Copyright 2019, Royal Society of Chemistry. (e) Comparison between the BPM-based and Nafion-based zero-gap electrolyzers in terms of timekeeping potential. (f) FE of analytical products of BPM-based and Nafion-based zero-gap electrolyzers. Reproduced with permission [124]. Copyright 2011, Wiley.

  下载: 全尺寸图片 幻灯片
  

  Figure 12  (a) Effect of operating temperature on the reduction of MEA CO₂RR electrolytic cell to CO cell voltage and FE. (b) AC impedance spectrum recorded at circuit voltage, inset zooms in on the high frequency region to show the variation of ion conduction with temperature. Reproduced with permission [133]. Copyright 2019, Royal Society of Chemistry.

  下载: 全尺寸图片 幻灯片
  

  Figure 13  Schematic diagram of the structure of AEM-MEA and its CO₂RR performance. (a) Schematic diagram of the local structure and cathode structure of AEM-MEA. (b) Schematic diagram of the overall AEM-MEA. (c) Product selectivity of 250 nm thick Cu catalyst layer. (d) Liquid product selectivity at different voltages. (e) Stability of the AEM-MEA electrolytic cell Reproduced with permission [145]. Copyright 2019, Elsevier.

  下载: 全尺寸图片 幻灯片
  

  Figure 14  (a) Material production process flow diagram and schematic diagram of the preparation process of BC-PDDA-OH⁻ membrane. (b, c) FE of different membranes under 0.5 mol/L KHCO₃ and 0.5 mol/L KOH electrolytes. Reproduced with permission [146]. Copyright 2021, Elsevier.

  下载: 全尺寸图片 幻灯片
  

  Figure 15  (a) Schematic of species migration within the MEA with forward-biased BPM. (b) Voltage stability of C₂H₄ and forward-biased BPM at 50 mA/cm². (c) Photograph of BPM with forward-biased mode after running at 50 mA/cm² current density for 2 h. The BPM membrane blistered at the AEM: CEM interface. Reproduced with permission [88]. Copyright 2021, American Chemical Society. (d) Schematic diagram of species migration within the MEA with reverse-biased BPM..

  下载: 全尺寸图片 幻灯片
  

  Figure 16  Schematic diagram of different structures of BPM-MEA and their CO₂RR performance. (a) Schematic diagram of BPMEA structure with high concentration anode electrolyte. (b) Schematic diagram of ion transport in BPMEA system. (c) The relationship between different current densities and electrolytic cell voltages under 3 M KOH anode electrolyte and the FE for CO. Reproduced with permission [152]. Copyright 2021, American Chemical Society. (d) Schematic diagram of MEA with PCRL structure. (e) Comparison of CO₂ flux in PCRL anode with AEM and CEM. (f) CO₂ flux distribution at PCRL coating thickness of 2.25 mg/cm². (g) Conversion efficiency and FE distribution at 2.25 mg/cm² coating thickness. Reproduced with permission [88]. Copyright 2021, American Chemical Society. (h) Schematic diagram of SC-BPMEA structure and mass transfer. (i-k) FE distribution of products with different cathode liquid thickness. Reproduced with permission [153]. Copyright 2022, Springer Nature.

  下载: 全尺寸图片 幻灯片
  

  Figure 17  (a) Schematic diagram of the modified BPM structure by LBL assembly. (b) CO FE versus cathodic potential for the ionomer-free catalyst layer. (c) CO fraction current density versus cathodic potential. Here the LBL BPM contains 10 PAH/PAA bilayers. the decrease in CO current density of Nafion BPM at higher potentials may be caused by water immersion in the ionomer-free catalyst layer. Reproduced with permission [81]. Copyright 2021, Springer Nature.

  下载: 全尺寸图片 幻灯片

  Table 1.  Reaction equations and electrode potentials of different reduction products [50].

  下载: 导出CSV

  Table 2.  Standard potentials for cathode and anode electrochemical reactions [139].

  下载: 导出CSV

  Table 3.  Comparison of different structure CO₂RR electrolytic cell development and design. CO₂ reduction reactor types.

  下载: 导出CSV
•