A scalable membrane electrode assembly architecture for efficient electrochemical conversion of CO2 to formic acid

The electrochemical reduction of carbon dioxide to formic acid is a promising pathway to improve CO2 utilization and has potential applications as a hydrogen storage medium. In this work, a zero-gap membrane electrode assembly architecture is developed for the direct electrochemical synthesis of formic acid from carbon dioxide. The key technological advancement is a perforated cation exchange membrane, which, when utilized in a forward bias bipolar membrane configuration, allows formic acid generated at the membrane interface to exit through the anode flow field at concentrations up to 0.25 M. Having no additional interlayer components between the anode and cathode this concept is positioned to leverage currently available materials and stack designs ubiquitous in fuel cell and H2 electrolysis, enabling a more rapid transition to scale and commercialization. The perforated cation exchange membrane configuration can achieve >75% Faradaic efficiency to formic acid at <2 V and 300 mA/cm2 in a 25 cm2 cell. More critically, a 55-hour stability test at 200 mA/cm2 shows stable Faradaic efficiency and cell voltage. Technoeconomic analysis is utilized to illustrate a path towards achieving cost parity with current formic acid production methods.

5) It is crucial to provide clarity on how the authors ensure a relative humidity of 100% in the cathode gas stream.Controlling the amount of water in the CO2 input stream is one potential approach to achieve this, but further details are required.To enhance the understanding of the experimental setup and procedures, it is recommended to include a schematic diagram of the setup in the manuscript.
6) It is essential to discuss how the authors estimate the oxidation of formic acid, considering that it is collected in both compartments of the MEA structure with an anion exchange membrane (AEM) and a perforated cation exchange membrane (CEM).Providing insights into the methodology used to estimate the oxidation of formic acid will enhance the understanding of the experimental process and the reliability of the results.
7) The authors are suggested to provide the cathode and anode potential in order to compare with others' work.
8) The calculus of figures of merit (e.g.rate, energy efficiency) should be included in the methodology section.9) Can you explain the advantages of using an anion exchange membrane (at the beginning of the work) instead of a cation exchange membrane?
10) The loading of the catalyst in both the anode and cathode is an important parameter that can significantly impact the performance of the CO2 electroreduction process.The catalyst loading refers to the amount of catalyst material deposited on the electrode surface.It plays a crucial role in determining the catalytic activity, selectivity, and overall efficiency of the electrochemical reactions.
11) The authors are encouraged to clearly identify the uniqueness of their manuscript in comparison to other works and adjust their claims accordingly.
12) The structure of the results section in the study could be improved with more sub-sections to enhance readability.Additionally, providing a more comprehensive explanation of the different phenomena occurring throughout the process would help readers better understand the results.13) While the manuscript focuses on the energy consumption aspect of the CO2 electroreduction process, it is important to also address the formic acid/formate concentration achieved in comparison to the existing literature.The concentration of formic acid or formate is a critical parameter as it directly relates to the efficiency and practicality of the process for potential applications.Therefore, it is recommended to include a discussion regarding the formic acid/formate concentrations obtained in this study and their comparison to those reported in the literature.By highlighting the concentration values achieved in the current work and comparing them to the reported concentrations in previous studies, the authors can provide insights into the performance and competitiveness of their proposed method.
14) It is recommended to provide additional details about the perforation process of the cation exchange membrane in order to enhance the readers' understanding.It would be beneficial to include illustrative figures or diagrams that depict the steps involved in the perforation process.
In this work, the authors developed a zero-gap membrane electrode assembly (MEA) cell configuration using a perforated cation exchange membrane (PCEM) for the conversion of CO2 to formic acid.By incorporating the PCEM layer, they successfully addressed the issues associated with the forward bias operation of a bipolar membrane for CO2 reduction.The authors utilized Bi-based catalysts for CO2 reduction and Pt/C catalysts for hydrogen oxidation at the anode, resulting in a formic Faradaic efficiency of over 80% at a current density of 300 mA/cm2.Their system produces formic acid, which is collected at the anode side, and remains stable for 55 hours at a current density of 200 mA/cm2.When H2 oxidation is employed at the anode, the cell voltage decreases to below 2 V at a current density of 200-300 mA/cm2.Overall, this work presents an interesting approach to membrane modification for CO2 reduction to formic acid and potentially has a significant impact.I have some comments and questions below: 1. Details about the "pore" structure of the PCEM layer should be provided.How do the pore size and density affect the diffusion of formic products and the overall performance?2. The quantification and discussion of CO2 crossover through the system should be included.It is unclear if CO2 crossover has been included in the TEA model.

3.
It is interesting that water and formic acid do not cause flooding in the carbon layer of the anode.Why did the author use carbon paper with 5% PTFE content? 4. Could the authors explain why the perforation was employed only on CEM and not on both the carbon paper and CEM?The latter configuration may enhance the diffusion of formic acid from the CEM layer and prevent further oxidation.
Reviewer #3 (Remarks to the Author): In this manuscript, a novel CO2RR electrolyzer was designed by introduced a perforated cation exchange membrane (PECM) to facilitate the mass transport of formic acid generated at the membrane interface.The PECM cell exhibited good stability and high formic acid selectivity under high current.Authors also demonstrated the value of this technique through a techno-economic analysis.Though this novel structure of CO2RR electrolyzer is of interest for the electrocatalytic community, the reviewer thinks the study and the presented data are not comprehensive.Some questions and concerns need to be addressed before the manuscript can be assessed for publication.
1， The control device using a cation exchange membrane should be constructed.And the issue of this kind of device also should be discussed.
2， Detailed information about the PCEM perforated cation exchange membrane, such as optical pictures and its preparation method, should be provided.The author should also provide physical pictures of each component of the equipment, and mark the function of each component in the actual working equipment.
3，Authors should conduct a systematic investigation on the pore size and distribution of cation exchange membrane on the performance of PCEM.
4. Figure S3 shows the durability of the system with an 80 mm AEM and perforated CEM at 60oC, What is the reason for the periodic fluctuation of the data in the figure ? 5. How is formic acid effectively collected in actual devices 。 6.Is it possible to quantify the proportion of formic acid oxidation current in the total anodic current.

Response to reviewer comments
We would like to extend our thanks to the reviewers for putting their valuable time into this review.Their comments and suggestions have been valuable in improving the manuscript.We have made changes in response to the reviewers' comments.Our responses are in red and text changes are highlighted.

Reviewer #1 (Remarks to the Author):
The manuscript titled "A novel, scalable membrane electrode assembly architecture for efficient electrochemical conversion of CO2 to formic acid" presents a significant advancement in the field by introducing an innovative architecture that enables the direct synthesis of formic acid from carbon dioxide through the perforation of a cationic exchange membrane.The obtained results are highly promising and intriguing, demonstrating the ability to achieve formic acid concentrations of up to 0.25 M. Additionally, the manuscript includes stability tests conducted over a 55-hour period, which exhibit consistent Faradaic Efficiencies values and stable cell voltage.Moreover, the authors have taken a commendable step by performing a techno-economic analysis to provide insights into the potential cost competitiveness of their approach compared to existing formic acid production methods.This analysis enhances the manuscript's value by illustrating a viable pathway towards achieving cost parity in formic acid production.The novelty of the proposed architecture is evident, and the manuscript holds significant potential for publication in Nature Communications.However, to ensure the manuscript's quality, it is essential to address the following comments and suggestions: Comment 1.I would recommend including an explanation of the state-of-the-art in the field of CO2 electroreduction to obtain formic acid or formate (as well the Table S1 of  Response: Thanks for your suggestions.We've updated our introduction section and included more stateof-the-art CO2 electroreduction devices.For recommended references, only reference (i) contains the energy consumption info, ref (ii) and ref (iii) either doesn't have energy consumption numbers or doesn't contain info about cell voltage for energy consumption calculations.We also updated our Table S1 and the Figure 1b based on ref (i) data.

Changes: (Line 37)
With momentum building for a formic acid economy 1,9 , multiple research efforts have focused on the optimization of catalyst selectivity [10][11][12][13][14][15][16] .However, the majority of efforts remain on smallscale H-cell or liquid flow cells operating at low current densities (<50 mA/cm 2 ).To reduce cost (Line 48)   diffusion electrode (GDE) based CO2R to formate/formic acid devices.These device configurations can be categorized into three main groups: 1. Flowing catholyte [18][19][20][21][22][23][24][25] , 2. Single membrane (either cation exchange membrane (CEM) 26 or anion exchange membrane (AEM)) 27 , and 3. Interlayer configurations [28][29][30] .Simplified cross sections for these configurations are shown in Figure 2a.For catholyte configurations, an electrolyte chamber is created between the membrane and the cathode GDE.The flowing catholyte serves to provide ionic access to the (Line 58, 59) mA/cm 2 27 .Díaz-Sainz et al 26 used a filter press setup with a single CEM membrane and can achieve 89% FE at current density of 45 mA/cm 2 ..Both approaches yielded formate as opposed (Table S1) (Figure 1b) Comment 2. This aspect is of great significance and calls for further elaboration.The author should conduct a comprehensive analysis of how their work enhances the scalability of the CO2 electroreduction process.It is crucial to provide insights into how the proposed method can be effectively scaled up to industrial levels and discuss the potential advantages or challenges associated with such scaling.
Response: Thanks for your comments.We've added more discussions about scalability of each configuration.

Changes:
(Line 121) architectures used for fuel cell and water electrolyzer stacks, enabling a more rapid transition to scale.Catholyte configuration contains a catholyte flow chamber, which can lead to pressure imbalance between the gas and liquid phase especially at larger cell configurations.For interlayer configuration with porous liquid flow layer, significant efforts are needed to optimize the porous interlayer to prevent significant pressure drop, carbon dioxide build-up within the interlayer and cell failure when scaling up to larger cells.It is also very hard to fabricate standalone thin, porous interlayer at a large scale.In contrast, the proposed new configuration is a zero-gap MEA configuration and doesn't contain any liquid flow chamber or interlayer.The uniqueness of the proposed configuration compared to other existing electrochemical cells is that it can generate formic acid instead of formate in a zero-gap configuration, and the structure is energy efficient and easy to scale-up.
Comment 3. Further elaboration on Figure 1 is warranted.The authors should place greater emphasis on the main reactions occurring in the cathodic and anodic compartments within the different configurations depicted in Figure 1.
Response: Thanks for your suggestions, we've added the corresponding reactions in the cathodic and anodic compartments to Figure 1.

Changes:
(Figure 1) Response: Thank you for your suggestions.We've added more discussions related to current state of research in this field.

Changes:
(Line 46-49) Along these lines, recent efforts have been made to develop industrially relevant gas diffusion electrode (GDE) based CO2R to formate/formic acid devices.Fernández-Caso et al. 20 made a comprehensive review summarizing all the electrochemical cell configurations for continuous reduction of CO2 to formic acid/formate.In general all existing configurations can be categorized into three main groups: 1. Flowing catholyte 19,[21][22][23][24][25][26][27] , 2. Single membrane (either Comment 5.It is crucial to provide clarity on how the authors ensure a relative humidity of 100% in the cathode gas stream.Controlling the amount of water in the CO2 input stream is one potential approach to achieve this, but further details are required.To enhance the understanding of the experimental setup and procedures, it is recommended to include a schematic diagram of the setup in the manuscript.
Response: Thanks for the suggestions.We've updated supplemental Figure S5 to include mor details of our experimental setup.Basically, the anode and cathode gas went through two separate humidifier bottles.When the cell is operated at 60C, the water temperature inside the humidifier bottle is also controlled to be 60C.The gas will bubble through the humidifier bottle and bring water vapor out to the cell.When the bottle temp is the same as the cell, the gas relative humidity is 100% RH.The RH for the electrochemical station is calibrated using a Vaisala relative humidity meter.

Changes:
(Supplemental Figure S5) Figure S1.Detailed information about all the inputs and outputs of the cell during electrochemical cell testing.The anode is supplied with a mixture of 0.8 slpm H2 at 100% RH and 10 mL/min DI water.The cathode is supplied with 2 slpm 100% CO2 gas mixed with 2 mL/min 1 molar KOH.Formic acid is collected at the anode, and the CO and H2 are collected from the cathode.
Comment 6.It is essential to discuss how the authors estimate the oxidation of formic acid, considering that it is collected in both compartments of the MEA structure with an anion exchange membrane (AEM) and a perforated cation exchange membrane (CEM).Providing insights into the methodology used to estimate the oxidation of formic acid will enhance the understanding of the experimental process and the reliability of the results.
Response: Thanks for pointing this out.We did characterize the liquid coming from both compartment, anode and cathode using an experiment set up shown in updated supplemental Figure S5.There is trace amount of formate that can be detected from the cathode dropout, about two orders of magnitude smaller compared to the anode, and accounts for < 0.5% of the total FE.This is probably because AEM ionomer and AEM membrane are used close to the cathode, making the formate ion preferably electromigrated from the cathode to the PCEM/AEM interface, instead of accumulating at the cathode catalyst layer.
Because of that, we can estimate the oxidation of the formic acid based on the mass balance.We've added discussions about the formic acid oxidation estimation.

Changes:
(Line 126-127) loss) increase (Figure S1d), illustrating the zero-sum tradeoff.The amount of formate oxidation is estimated based on total mass balance, with details in the method section.The performance of (Line 422 -432) The amount of formic acid oxidation at the anode is calculated based on the total mass balance.There are three competing reactions at the cathode, hydrogen evolution reaction, CO2 reduction to CO and CO2 redcution to formic acid.Because we have formic acid oxidation process at the andoe, the FE for formic acid can be further divided into two parts, formic acid collected and formic acid oxidized.The total mass balance can be written as below: We quantified the amount of formic acid collected from HPLC, the amount of hydrogen and CO using GC.It should be noted that the majority of the formic acid is collected from the anode, using setup depicted in supplemental figure S5.The is negligible amount of formate collected from the cathode compartment, about two orders of magnitude smaller and accounts for less than 0.5% of the total FE.
Comment 7. The authors are suggested to provide the cathode and anode potential in order to compare with others' work.
Response: Thanks for the comments and suggestions.Measuring anode and cathode potential in a zerogap membrane electrode assembly is much harder than in a liquid cell.However, we did conduct a MEA polarization curve tests with an in-house designed hydrogen reference electrode attached.
Supplemental Figure S6 shows our H2 reference electrode configuration.
(Figure S6) Figure 6a shows the potential break-down analysis using this hydrogen reference electrode.
(Figure 6a) Comment 8.The calculus of figures of merit (e.g.rate, energy efficiency) should be included in the methodology section.
Response: Thanks for pointing this out.We have the Faradaic Efficiency and Purity calculation descriptions in the method section.We added the Formate/Formic acid yield (mol/kWh) calculation to the same section, as highlighted.

Changes:
(Line 359) Faradaic Efficiency, Formate/FA Yield, and Formic Acid Purity Calculation: Formate/FA yield is the amount of FA generated per kWh of electricity consumed using certain MEA configuration, with unit mol/kWh.It is calculated based on the current density, cell voltage, and Faradaic Efficiency at certain operating conditions.

= × *
Comment 9. Can you explain the advantages of using an anion exchange membrane (at the beginning of the work) instead of a cation exchange membrane?
Response: Thanks for your comments.We tested using single CEM MEA  S1 e-g.When single CEM is used, cations like potassium, sodium or proton will migrate from the anode to the cathode through the CEM, while it is hard for formate to transport from the cathode to the anode through the CEM.Because of that, salt like potassium formate will accumulate and then get collected from the cathode.The initial performance of this system looks good, but the potassium formate accumulation at the cathode can lead to severe durability issues, like the salt will precipitate and block the cathode catalyst layer.After two hours of operation, significant amount of salt accumulation can be observed, especially at high current densities (>100 mA/cm2).Photo images of the GDE and flow field after two hours of operation at 200 mA/cm2 is also included in the Supplemental Figure S12.

Changes:
(Line 112-115) anolyte concentration is reduced to 0.1 M KOH, both cell voltage and formate oxidation (formate loss) increase (Figure S1d), illustrating the zero-sum tradeoff.The performance of using a MEA configuration and single CEM membrane is also investigated, with results shown in Figure S1 f,g.The formate FE collected from the cathode is >60% at 200 mA/cm 2 at beginning of test, but suffers from fast degradation within two hours due to cathode salt accumulation (Figure S12).Comment 10.The loading of the catalyst in both the anode and cathode is an important parameter that can significantly impact the performance of the CO2 electroreduction process.The catalyst loading refers to the amount of catalyst material deposited on the electrode surface.It plays a crucial role in determining the catalytic activity, selectivity, and overall efficiency of the electrochemical reactions.
Response: Thanks for pointing this out.We've added the loading information.

Changes:
(Line 303-305) speed of 55 mm/sec.These coated electrodes were transferred to an oven and dried at 80 Comment 14.It is recommended to provide additional details about the perforation process of the cation exchange membrane in order to enhance the readers' understanding.It would be beneficial to include illustrative figures or diagrams that depict the steps involved in the perforation process.
Response: Thanks for your suggestions.We've added a supplemental Figure S13 with details of the perforation process.We also updated the method section to include more descriptions of the perforation steps.

Changes:
(Supplemental Figure S13) In this work, the authors developed a zero-gap membrane electrode assembly (MEA) cell configuration using a perforated cation exchange membrane (PCEM) for the conversion of CO2 to formic acid.By incorporating the PCEM layer, they successfully addressed the issues associated with the forward bias operation of a bipolar membrane for CO2 reduction.The authors utilized Bi-based catalysts for CO2 reduction and Pt/C catalysts for hydrogen oxidation at the anode, resulting in a formic Faradaic efficiency of over 80% at a current density of 300 mA/cm2.Their system produces formic acid, which is collected at the anode side, and remains stable for 55 hours at a current density of 200 mA/cm2.When H2 oxidation is employed at the anode, the cell voltage decreases to below 2 V at a current density of 200-300 mA/cm2.Overall, this work presents an interesting approach to membrane modification for CO2 reduction to formic acid and potentially has a significant impact.I have some comments and questions below: 1. Details about the "pore" structure of the PCEM layer should be provided.How do the pore size and density affect the diffusion of formic products and the overall performance?
Response: Thank you for your comments and suggestions.We've added characterizations to the pore structure in the supplemental Figure S13.In general, the perforation are parallel lines on the CEM membrane, with a gap in the range of 30 µm and a spacing of 3 mm.That is about 10% of the active surface area with open perforation.We did conduct a very brief study by increasing the perforation density, and we saw worse performance with more formic acid oxidation at the anode, as shown in the illustration figure below.Our understanding is that there are three different transport pathways for formic acid that is generated at the PCEM/AEM interfaces.The first is it will transport through the perforation and gets collected at the anode exhaust, the second is that it will meet with anode catalyst and then gets oxized, the third is back diffuse to the cathode.When we increase the perforation area, there is a higher chance for formic acid to transport to the anode catalyst and gets oxidized.The perforation area or the perforation pore size should be large enough to let formic acid to come out, but not too large to cause too much anode formic acid oxidation.

Figure 1 .
Figure 1.(a) Comparison of the three most prominent device configurations for CO2R to formate/formic acid, along with the architecture proposed in this study.(b) Comparison of total current and formate/formic acid yield for catholyte configuration, interlayer configuration, single CEM configuration from literature as shown in supplemental Table S1 and our work.Hollow markers represent the production of formate salt solution, while solid markers indicate the production of formic acid.* Stands for configurations using hydrogen at the anode.(c) The structure of zero-gap MEA configuration using composite bipolar membrane with perforated cation exchange layer operating in forward-bias mode.

Figure 2 .
Figure 2. (a) Comparison of the three most prominent device configurations for CO2R to formate/formic acid, along with the architecture proposed in this study.(b) Comparison of total current and formate/formic acid yield for catholyte configuration, interlayer configuration, single CEM configuration from literature as shown in supplemental Table S1 and our work.Hollow markers represent the production of formate salt solution, while solid markers indicate the production of formic acid.* Stands for configurations using hydrogen at the anode.(c) The structure of zero-gap MEA configuration using composite bipolar membrane with perforated cation exchange layer operating in forward-bias mode.

Figure S 2
Figure S 2. (a) Schematic of the reference electrode hardware and the Nafion sensing tip.(b)Cross-sectional image of the reference electrode and the MEA.

(
Figure S3.(a) Schematic of the zero-gap MEA utilizing a single AEM membrane, with CO2R at the cathode with the Oxygen evolution reaction at the anode.(b) Cell voltage and (c) FE vs. time at 200 mA/cm 2 with 1M KOH used at the anode.(d) FE and cell voltage at different current densities when 0.1 M KOH is used at the anode.(e) Schematic of the zero-gap MEA utilizing a single CEM membrane, with CO2R at the cathode with the Oxygen evolution reaction at the anode.(f) FE and cell voltage at different current densities when 0.1 M KOH is used at the anode.(g) Cell voltage and FE vs. time at 200 mA/cm 2 with 1M KOH used at the anode.

Figure S 4 .
Figure S 4. Images of the (a) back of the cathode GDE and (b) flow field of the single CEM MEA after 2 hours of operation under 200 mA/cm 2 .Significant amount of salt accumulation can be observed.

°C.
Figure 3. (a) Cell voltage break-down using H2 reference electrode of the cell operated at 60 o C with Pt/C anode and 80 µm AEM.(b) FE and formic acid concentrations collected at 200 mA/cm 2 with different anode DI water flow rate.(c) Cell voltage at 200 mA/cm 2 when different concentrations of formic acid are collected at the anode.(d) Minimum selling price break down based on the performance at different DI water flow rate, using industrial national average electricity price of 0.068 $/kWh, and 4.5 $/kg hydrogen.(*: Assuming minimum amount of formic acid oxidation at the anode with 10M FA concentration, industrial national average electricity price of 0.068 $/kWh, and 4.5 $/kg hydrogen.**: Assuming minimum amount of formic acid oxidation at the anode with 1.3M FA concentration, projected future electricity price of 0.03 $/kWh, and 2.3 $/kg hydrogen.Dashed line represents market price for 85wt% FA)

Figure S 5
Figure S 5 (a) Details of different components for the cell assembly.From left to right are components assembled from the anode to the cathode.(b) Perforated catalyst coated cation exchange membrane preparation process.(c) Zoomed in figure of the perforation and the spacing of the parallel cutting lines.(d) 3D structure of the X-ray CT image showing the perforated cation exchange membrane with perforation gap of 32.6 µm.(e) cross sectional image of the Xray CT reconstruction showing the perforated cation exchange membrane with perforation gap of 32.6 µm, anode catalyst layer, and the anion exchange membrane.
Results are added to the Supplemental Figure