Method for reversible ph-swing flow cells for co 2 capture

Abstract

A process for capture of CO2 using a reversible redox salt in an electrochemical cell is provided.

Description

Docket No.30275 / 59445 METHOD FOR REVERSIBLE PH-SWING FLOW CELLS FOR CO2CAPTURE CROSS REFERENCE TO RELATED APPLICATION

[0001] The benefit of priority to U.S. Provisional Application No.63 / 543,842 filed October 12, 2023, is hereby claimed and the disclosure is incorporated herein by reference in its entirety. FIELD

[0002] The disclosure relates to an electrochemical pH-swing process for CO2capture, and more particularly, to a process utilizing H2 and redox salt cycling to capture CO2 from a water source BACKGROUND

[0003] Carbon dioxide removal (CDR) technologies are critical to limit global warming to 2 °C. In this regard, oceans are huge sinks for carbon and have to-date captured about 40% of all anthropogenically released CO2. As the atmosphere continues to equilibrate with the oceans, a majority of all CO2 emissions will be stored in the oceans, which will further exacerbate harmful impacts of ocean acidification. Thus, making removal of CO2 emissions from ocean water as critical as its removal from air.

[0004] For both air and ocean water CDR, a wide variety of electrochemical techniques have been proposed which rely on a pH-swing to thermodynamically shift the speciation of CO2. As a polyprotic acid, CO2can buffer a changing pH by bonding with or releasing protons. At the typical pH of ocean water of 8.1, bicarbonate (HCO−3) is the main constituent. However, at a lower pH, dissolved CO2becomes increasingly more dominant. This CO2can be degassed from the liquid and captured for further use. Subsequently, the pH is increased (basified) to neutralize the ocean water while negating the compositional changes incurred during acidification.

[0005] Prior work on ocean water CDR shift pH with many strategies including: (a) electrolysis, (b) bipolar membranes, and more recently with (c) hydrogen looping and (d) proton / chloride mediated electrodes. Specifically, electrolysis provides a pathway to produce renewable H2and O2via water splitting in addition to pH shifting. In electrolysis cells, pH gradients are generated from O2and H2evolution reactions at electrode surfaces, which is used to collect aqueous CO2in either its gaseous phase or as solid carbonates. However, electrolysis of water to produce O2and H2involves a large thermodynamic penalty of 1.23 V. This is in addition to the voltage requirement to maintain a pH difference between the anode and theDocket No.30275 / 59445 cathode. To avoid the requirement of water splitting, bipolar membranes (BPMs), which are an assembly of an anion and a cation exchange membrane, have been proposed. Water dissociates in these membranes to produce OH−and H+, and to create the necessary pH- gradient for CO2capture from ocean water. This approach yields a lower thermodynamic penalty of 0.83 V for a maximum pH shift of 14. A BPM-based device was demonstrated to be capable of capturing 71% of dissolved CO2 from simulated ocean water streams with an improved energy intensity of 3.5 GJ / tCO2 (0.98 kWh / kg CO2) compared to electrolysis systems.

[0006] H2 looping is another technique that can be used to achieve a pH gradient. With H2 looping, water is reduced to form H2 at the cathode, and the produced H2 is oxidized to form H+at the anode. This results in a thermodynamic penalty similar to the BPM system, consisting of only the pH shift. However, compared to the BPM system, H2 looping has been able to achieve even lower energy intensity of 2.38 GJ / tCO2 (660 kWh tCO2−1) due to better reaction kinetics.

[0007] Electrified / electrochemical CDR technologies present attractive prospects to integrate CO2 removal with renewable electricity. However, when operated at the mega-ton scale, which is a common goal for CDR technologies, it can pose formidable challenges to the grid. Demand- side flexibility, which could be easier to achieve with newly electrified loads, not only enhances operational efficiency, but also can decrease production and levelized costs.24This concept of integrating demand-side flexibility to lower energy and possibly capital costs has been considered for electrolytic H2 production systems. This aspect has not been examined as extensively in state-of-the-art CDR technologies for ocean water. Frey et al. demonstrate the value of pairing electrochemical water splitting for H2 and O2 production with a ceria redox shuttle. This pairing not only spatially decouples where H2 and O2 are produced, but provides flexibility and energy cost savings by temporally alternating between electricity charging and discharging states. Ruggles et al. showcase the benefits of operating electrolyzers as flexible loads to not only make use of low-cost electricity, but to also enable high capacity-factors for electrolyzer operation. Additionally, this study makes the case for the availability of significant amounts of free or low-cost electricity from wind and solar generation; this can be harnessed to power nascent technologies, provided they have the feature of load or demand-side flexibility. Even dynamically altering the current density to change power-consumption for a polymer- exchange membrane electrolyzer is shown to reduce the levelized cost for H2production by over half. In the same vein, powering electrified / electrochemical DAC processes by excess electricity is projected to power 20-140 Mt of CO2from 2030-2050.Docket No.30275 / 59445 SUMMARY

[0008] The process of the disclosure provides reversible, pH-shifting electrochemical process that cycles hydrogen (H2) and a redox salt to capture CO2 from a water source, such as ocean water. The process of the disclosure provides significant advantage in its dual capability to capture CO2 while offering demand-side flexibility to enable deeper penetration of renewables in the grid. This flexibility stems from having a reversible, two-step process, where the acidification of the water source to capture CO2 is power producing, while the basification step to neutralize the water is power consuming.

[0009] Processes of the disclosure provide a CDR process that implements pH-shifting with H2 and redox salt cycling with distinct capabilities to reversibly operate the same process in power consumption (electrolyzer) and production (fuel cell) modes. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1: Schematic of the proposed, reversible, oceanic CO2removal (CDR) process with H2 and redox salt (Na4Fe(CN)6 / Na3Fe(CN)6) cycling between two steps – (top) acidification, where pH-neutral ocean water is acidified by H2oxidation coupled with the reduction of Na3Fe(CN)6, and (bottom) basification, where the reverse reactions occur with Na4Fe(CN)6oxidation together with H2evolution at the cathode.

[0011] Figures 2A to 2D: Graphs showing predicted performance from 0-D equivalent circuit modeling for (a,c) acidification and (b,d) basification to determine (a,b) polarization behavior and (c,d) Faradaic efficiencies; including (c) power density produced during acidification.

[0012] Figure 3: (a-c) Graphs showing predictions for electrochemical energy intensity normalized by the amount of CO2 captured (Eq. (12)) in (GJ / tCO2) as a function of the current density when jbase = jacid (a) without and (b) with competing reactions modeled. Dashed lines on the plot indicates a baseline cell resistance of 2.18 Ω cm2and the shaded areas are indicative of this ohmic resistance in the range of 0 - 5 Ω cm2; the shades of blue are indicative of the liquid-side boundary layer thicknesses, δl, modeled. Best-case scenarios with the lowest electrochemical energy intensities from (a) and (b) are shown in (c), and compared with state-of-the-art ocean water CDR process intensities reported by Digdaya et al., Kim et al., and Yan et al., (d) Overall energy intensity (GJ / tCO2) including electrochemical and parasitic loads (for the proposed process, calculated at an industrially- relevant current density of 100 mA cm-2, the baseline cell resistance, and pH shift of 6 toDocket No.30275 / 59445 9.6, compared with electrochemical and thermally-driven direct air capture (DAC) processes with data obtained from Singh et al.; error bars for the thermal DAC processes are from Viebahn et al.

[0013] Figure 4 are graphs showing (a) integration of solar and wind electricity production in California, CAISO 2022-2023, with predicted energy demand for the proposed oceanic CO2 removal process as a function of varying extents of yearly CO2 captured, 0 – 16 MtCO2,year. (b) Hourly power demand to capture 1 tCO2 per day. The top gray curve indicates the process shown in (a), while the bottom curve shows a constant power demand process occurring at an equivalent energy intensity, i.e., the area underneath both curves are the same. (c) Hourly cost to capture 1 tCO2 per day. The gray curve indicates the hourly cost associated with variable energy process shown in (b) if all produced electricity is sold at market value, while the red curve compares it to the hourly cost associated with constant power demand. Operating conditions of the proposed, variable process shown in (a-c) include jacid = 75 mA cm−2and jbase = 150 mA cm−2, while still constraining the same volume of ocean water being treated over a 24-hour day in the individual steps (i.e., tbase = 0.5tacid). The process switches between acidification (early mornings and evenings / nights) and basification (mid-day) based on the supply and the cost of electricity purchased from CAISO. (d) Electricity cost savings for the proposed process operation compared to a process with constant energy usage at different ratios of tbase to tacid, assuming a time- weighted average operating current density of 100 mA cm−2. Different line styles indicate different proportions of market value at which produced electricity can be sold (0%, 50%, and 100%). For all cases: competing reactions are not considered, a thin boundary layer is assumed (δl = 10 µm), and the pH shift is from 6 to 9.6. Parasitic energy costs are included for CO2 phase separation, but not for pumping as we assume co-location with a desalination plant. Parasitics are assumed to only operate during the basification process, at high enough power to remove all CO2treated over the day.

[0014] Figure 5: Equivalent circuit model that is used to model the performance of the acidification step (pink) and the basification step (orange). For any one species, Nernstian required potentials are indicated by U, mass transfer overpotentials are indicated by ηMT,and kinetic overpotentials are indicated by ηK. ηKof competing reactions (shown by light gray lines) is neglected for a conservative estimation. All competing reactions are modeled as mass transport limited once they are thermodynamically favorable. The reactions are abbreviated as follows: Hydrogen Oxidation Reaction (HOR), Oxygen Reduction ReactionDocket No.30275 / 59445 (ORR), Redox Salt Reduction Reaction (SRR), Redox Salt Oxidation Reaction (SOR), Sulfate Reduction Reaction (SO42-).

[0015] Figures 6A to 6C are schematic illustrations of an electrochemical flow cell for performing a process in accordance with the disclosure.

[0016] Figure 7 is a graph showing LSV scans for the acidification experiments at multiple flow rates. Additional model curves have been added to the plot to compare the model projected performance against experimental results. The best fit model is the scenario with an 18 Ωcm2 areal resistance, a 100μm or 200μm liquid boundary layer, and assuming a pH shift from 4 to 10.7.

[0017] Figure 8 is a graph showing LSV scans for a process in accordance with the disclosure at constant current.

[0018] Figure 9 is a schematic illustrating the precipitation cleaning benefit of a process in accordance with the disclosure, with (a) showing Pt-C GDE before the experiment, (b) the overall electrode and a microscope image after the Pt-C GDE is used during a constant current basification experiment. Significant precipitation was observed on the electrode surface. Finally (c) the same electrode is shown after use in an acidification process, showing that the electrode is cleaned during the process and that the precipitation filling the cracks along the GDE surface was cleaned off the surface.

[0019] Figures 10A-10C are graph showing electrochemical energy intensity of a process in accordance with the disclosure at constant current and extracted energy intensity for the best fit model from the LSV experiment, compared to prior known processes disclosed in the literature.

[0020] Figure 11 is a graph showing the fraction of inorganic carbon present as CO2, HCO3- and CO32-in solution as a function of pH.

[0021] Figure 12 is a graph showing different pH shifts for use in the process of the disclosure.

[0022] Figure 13 is a graph showing the reversible desired reactions and undesired mass transfer limited reactions modeled via the O-D equivalent circuit model.

[0023] Figures 14A to 14D are graphs showing predicted performance without competing reactions from 0-D equivalent circuit modeling for (a,c) acidification and (b,d) basification to determined (a,b) polarization behavior and (c,d) Faradaic efficiencies including (c) power density produced during acidification.Docket No.30275 / 59445

[0024] Figure 15A and 15B are graphs showing predictions for electrochemical energy intensity normalized by the amount of CO2captured as a function of current density when jop,base= jop,acid (a) without and (b) with competing reactions modeled. Dashed lines on the plot indicates a baseline cell resistance of 2.18 Ωcm2and the shaded areas are indicative of this ohmic resistance in the range of 0 – 5 Ωcm2.

[0025] Figure 16 is a graph showing total energy intensity of different device layouts at industrially relevant current densities (jbase = jacid = 100 mA cm-2). All layouts are for no competing reactions with an ohmic resistances of 2.8 Ωcm2. Four different cases of water pumping energy intensities are shown to demonstrate the impact of water pumping. No water pumping (no WP) represents the process being co-located with desalination. The energy intensity water pumping are considered for both pH shifts.

[0026] Figure 17 is a graphs showing that there is more CO2 stored in ocean water than air at steady-state.

[0027] Figure 18 is a summary of Butler-Volmer mass transport limited kinetics.

[0028] Figure 19 is a schematic of an experimental set-up used in Example 5.

[0029] Figure 20 are graphs showing current density for different reactor designs (thick plate, rigded plate, thin plate) and flow rates.

[0030] Figure 21 is a graph comparing performance of a system in accordance with the disclosure to literature results

[0031] Figure 22 is a graph and associated imaging showing fouling after repeated cycles. DETAILED DESCRIPTION

[0032] Processes for capturing CO2from a water source in accordance with the disclosure can be performed in an electrochemical flow cell that includes a flow channel defined between an anode and an oppositely disposed cathode. A membrane is arranged in the flow channel to define a first side of the flow channel between the anode and the membrane and a second side of the flow channel between the cathode and the membrane. The process can include performing an acidification step followed by performing a basification step. The acidification step and basification step can be performed in a cycle, alternatingly. Alternatively, the acidification step can be performed repeatedly for a period of time or continuously for a period of time followed by the basification step. The acidification and basification steps can be performed in the same or in separate electrochemical flow cells. Processes of the disclosure contemplateDocket No.30275 / 59445 operation of multiple electrochemical flow cells for performing the acidification and / or basification steps. Acidification and basification steps can be performed repeatedly, with removal of the phase-separated CO2 being performed between each change from acidification to basification. The acidification and basification can be performed alternatingly or can be performed for durations of time before switching between process steps.

[0033] For a process in which acidification and basification are performed in a single flow cell, the process can include performing the acidification step by flowing the water source and H2 gas through the first side of the flow channel in a first direction. The process further includes flowing an electrolyte comprising a reversible redox salt through the second side of the flow channel also in the first direction. Upon application of a first potential to the electrochemical cell, an H2 oxidation reaction occurs at the anode with the water source being an anolyte for the H2 oxidation reaction and a reduction of the reversible redox salt occurs at the cathode, thereby acidifying the water source and causing a shift in speciation within the water source from HCO3- to dissolved CO2 thereby allowing the dissolved CO2 to phase separate from the water source, The phase-separate CO2 can then be removed from the electrochemical cell.

[0034] The process then further includes performing the basification step. In embodiments in which acidification and basification are performed in the same flow cell, the basification step includes reversing the polarization of the anode and cathode thereby providing a reversed electrochemical cell in which the anode becomes the cathode of the reversed cell and the cathode becomes the anode of the reversed cell. In this reversed cell, the first side is defined between the reversed cell cathode and the membrane and the second side is defined as being between the revised cell anode and the membrane. The basification step then includes flowing the acidified water having CO2 removed therefrom through the first side of the flow channel in a second direction, opposite to the first direction, and flowing the reduced redox salt through the second side of the flow cell also in the second direction. Upon application of a second potential to the electrochemical cell, an H2evolution reaction occurs at the cathode of the reverse cell and the reduced redox salt is oxidized at the anode of the reverse cell, thereby basifying the acidified water and generating H2, wherein the oxidized redox salt and generated H2can be collected for use in a subsequent acidification step. The acidification and basification steps can be repeatedly performed for any number of times or cycles, either continuously or in a batch- style process.

[0035] In embodiments in which the acidification and basification processes are performed in separate electrochemical cells, each cell can include a flow channel defined between an anodeDocket No.30275 / 59445 and an oppositely disposed cathode. In the first electrochemical cell, a membrane is arranged in the flow channel to define a first side of the flow channel between the anode and the membrane and a second side of the flow channel between the cathode and the membrane. In the second electrochemical cell, a membrane is arranged in the flow channel to define a first side of the flow channel between the cathode and the membrane and a second side of the flow channel between the anode and the membrane.

[0036] The process includes flowing the water source and H2 through the first side of the flow channel of the first electrochemical cell and flowing the electrolyte comprising the reversible redox salt through the second side of the flow channel of the first electrochemical cell. Upon application of a potential to the electrochemical cell, an H2 oxidation reaction occurs at the anode with the water source being an anolyte for the H2 oxidation reaction and a reduction of the reversible redox salt occurs at the cathode, thereby acidifying the water source and causing a shift in speciation within the water source from HCO3- to dissolved CO2 thereby allowing the dissolved CO2 to phase separate from the water source. The phase separated CO2 can then be removed from the first electrochemical cell. The acidified water and the electrolyte comprising the now reduced redox salt are the remaining outputs of the first electrochemical cell and serve as inputs into the second electrochemical cell.

[0037] The first and second electrochemical cells can be in fluid communication with one another such that the outputs of the first electrochemical cell are provided as the inputs of the second electrochemical cell. The fluid communication can be direct, such that the outputs of the first electrochemical cell are fed directly into the second electrochemical cell as inputs. The fluid communication can be indirect, such that the outputs of the first electrochemical cell are transferred to one or more storage vessels. The second electrochemical cell can then be fluidly couple to the one or more storage vessels for use as the inputs into the second electrochemical cell.

[0038] The second electrochemical cell performs the basification reaction using the outputs from the first electrochemical cell, with the CO2-depleted acidified water from the first electrochemical cell being flowed through the first side of the second electrochemical cell and the electrolyte comprising the reduced redox salt being flowed through the second side of the flow channel of the second electrochemical cell. Upon application of a potential to the second electrochemical cell, an H2evolution reaction occurs at the cathode and the reduced redox salt is oxidized at the anode, thereby basifying the acidified water and generating H2. The oxidized redox salt and generated H2are the outputs of the second electrochemical cell, which can beDocket No.30275 / 59445 fluidly coupled to the inlet of the first electrochemical cell and serve as inputs to the first electrochemical cell.

[0039] The first and second electrochemical cells can be in fluid communication with one another such that the outputs of the second electrochemical cell are provided as the inputs of the first electrochemical cell. The fluid communication can be direct, such that the outputs of the second electrochemical cell are fed directly into the first electrochemical cell as inputs. The fluid communication can be indirect, such that the outputs of the second electrochemical cell are transferred to one or more storage vessels. The first electrochemical cell can then be fluidly coupled to the one or more storage vessels for use as the inputs into the first electrochemical cell. The first and second electrochemical cells can be fluidly coupled in to provide a closed loop circulation of the H2 gas and electrolyte during the process.

[0040] Once the process has been initiated, the first and second electrochemical cells can be operated substantially simultaneously to provide a substantially continuous process for the capture of CO2 from the water source. Alternatively, the first and second electrochemical cells can be run at separate times, for example, alternatingly or intermittently with one another in a batch-style process.

[0041] The acidification step represents a power producing step in which power is generated by the process, while the basification step represents a power consuming step in which applied power is required to perform the process step. Power generated during the acidification step can be optionally stored and used for at least partial powering of the basification step. Power generated during the acidification step can be otherwise stored and removed from the system for alternative use. The processes of the disclosure advantageously allow for control over the timing for performing the power consuming verses power producing steps of the process, thereby providing demand flexibility. This can allow for integration of the process of the disclosure into a variety of power supply sources, such as renewable sources where the amount of energy available may vary throughout the day.

[0042] Figure 1 illustrates a schematic of a process in accordance with the disclosure, shown with reference to ocean water as the water source and Na3Fe(CN)6 / Na4Fe(CN)6as the redox salt by way of example only. During the process, the water pH decreases from 8.1 during acidification to produce a shift in speciation from HCO3- to dissolved CO2, which is phase- separated to capture pure CO2. Basification negates these changes through the addition of the same concentration of OH ions as H+ ions added to ocean water during acidification, resulting in pH > 8.1 at the end of basification. CO2-starved ocean water will re-equilibrateDocket No.30275 / 59445 with the atmosphere and get buffered back to a neutral-pH to re-enter acidification and the cycle continues. The net thermodynamic potential and the electrochemical energy intensity of this process is dictated by the extent of pH-swing between acidification and basification. Cell potentials shown imply a pH-shift of 14. Both step can occur in the same device while being temporally decoupled in a semi-batch process or the steps can be separately performed on different devices.

[0043] The cell operates between two steps (1) acidification to decrease the pH of the water source and (2) basification to negate changes incurred during acidification by increasing pH. To produce the pH-swing, H2 and a reversible redox salt are cycled in the flow cell. The reaction potentials are shown in Equations (1a) to (2d) below for an example system, assuming different pH for acidification and basification. Reaction potentials are expressed in reference to the standard hydrogen electrode (SHE). Acidification (Power Production) Cathode: [Fe(CN)6]3+ e ^ [Fe(CN)6]4, E°= 0.358 V vs. SHE (1a) Anode: H2^ 2e + 2H+, E°= −0.059 pHlowV vs. SHE (1b) Net: 2NaCl + H2 + 2Na3[Fe(CN)6] ^ 2Na4[Fe(CN)6] + 2HCl (1c) Acidification cell potential: E°cell = 0.358 + 0.059 pHlow V (1d) Basification (Power Consumption) Cathode: 2H2O + 2e ^ H2 + 2OH , E°= −0.059 pHhigh V vs. SHE (2a) Anode: [Fe(CN)6]4^ e + [Fe(CN)6]3-, E°= 0.358 V vs. SHE (2b) Net: 2H2O + 2Na4[Fe(CN)6] ^ 2Na3[Fe(CN)6] + 2NaOH + H2 (2c) Basification cell potential: E°cell = −0.059 pHhigh − 0.358 V (2d)

[0044] In the example shown, during the acidification step, H2 oxidation reaction (HOR) occurs at the anode with CO2-rich water as the anolyte, and ferricyanide ([Fe(CN)6)]3) reduction occurs at the cathode, as described in the reactions above. To maintain charge neutrality, an ion exchange membrane, such as a cation exchange membrane (CEM) can be used to facilitate migration of primarily Na+ions due to its relatively larger concentration compared to H+in ocean water. The low pH environment created at the anode shifts the species equilibrium towards dissolved CO2, which can be phase-separated from theDocket No.30275 / 59445 solution. The dissolved CO2can be separated, for example, using various methods, such as vacuuming using a membrane contactor. Because the net reaction is thermodynamically downhill (E°cell > 0 V), the acidification step is power producing.

[0045] During basification H2evolution (HER) occurs at the cathode and ferrocyanide ([Fe(CN)6)]4)) oxidation takes place at the cathode, which instead consumes power (E° cell < 0 V28). To negate the compositional changes incurred during acidification, the concentration of OH- ions added during basification is equal to that of the H+ions added. This balance prevents permanent changes to the water composition and reduces unanticipated consequences of CO2 capture and water treatment. However, this results in pH > 8.1 at the end of basification due the loss of buffering capacity from CO2, which was removed following acidification. The CO2-starved water can be buffered back to a neutral pH before discharge to the ocean or re-entering the acidification step, by equilibrating with the atmosphere and absorbing CO2. In a semi-batch process, the two steps of acidification and basification can be temporally decoupled, and the same rector / device can be used for both steps by reversing the flow of reactants and the polarization on the electrodes. The reversal provided in the methods of the disclosure can be useful in alleviated local salt precipitation at the electrode-electrolyte and membrane – electrolyte interfaces. Salt precipitation can be detrimental to the operation of the flow cell. The self-cleaning feature of the processes of the disclosure are advantageous in preventing precipitating build-up, thereby mitigating salt-precipitation problems from which other known processes currently suffer.

[0046] The redox salt can include can be a ferro / ferri-cyanide, iodide, bromide, and / or chloride salt with any one or more cations selected form sodium, lithium, and potassium. For example, the redox salt can be Na4Fe(CN)6 / Na3Fe(CN)6, K4Fe(CN)6 / K3Fe(CN)6

[0047] The water source can be, for example, oceans, lakes, rivers, and other surface waters. The water source can further include alkaline solutions resulting from Direct Air Capture processes or other bicarbonate containing aqueous sources, such as electrolytes.

[0048] The water source can have a pH of about 7.5 to about 8.5. The acidified water source after the acidification step can have a pH of about 4 to about 6. The basified water after performing basification step can have a pH of about 9 to about 11.

[0049] In any of the embodiments of the process of the disclosure, the process can include storing the outputs of any of the acidification or basification steps. For example, the process can include storing the CO2depleted acidified water and electrolyte containing the reducedDocket No.30275 / 59445 redox salt. The process can include repeatedly performing or performing in a continuous manner, the acidification step and storing the outputs thereof. The process can then utilize the stored CO2 depleted acidified water and electrolyte containing the reduced redox salt to perform the basification process either in the same electrochemical cell or in a separate electrochemical cell. This can provide flexibility in the timing in which the acidification and basification steps are performed, thereby providing flexibility in the energy needs by allowing for switching and control over when the power producing acidification process is performed vs when the power consuming basification process is performed.

[0050] For example, the process can include performing repeated cycles, each cycle comprising performing the acidification step, removing the phase-separated CO2 and performing the basification step. Alternatively, an acidification cycle and CO2 removal can be performed followed by a basification cycle, whereby in each cycle acidification or basification is performed on two or more volumes of inputs introduced into the electrochemical cell or for a defined period of continuous flow of the inputs into the electrochemical cell.

[0051] The processes of the disclosure can be performed with operating current densities from 0 to about 500 mA cm-2. This can result in a generated potential for acidification of about 0 to 1 V and an applied potential for basification of about 1 to 2.5 V between the anode and cathode.

[0052] Removal of the phase-separated CO2 can be performed using any known methods. For example, the phase-separated CO2 can be removed using air contactors or with vacuum separations.

[0053] The electrochemical cell includes the anode, the cathode, and a membrane disposed in the flow channel. The electrochemical cell can be free of a bipolar membrane. The membrane can be for example an ion exchange membrane. The cathode and anode can be any suitable materials. For examples, either or both of the anode of the acidification process and cathode of the basification process can be a porous electrode layer, made of stable material such as carbon or titanium, with a catalyst for suitable hydrogen reactions, which are broadly based on noble and transition metals (e.g., Pt, Ir, Ru, Rh, Au, Ag, Ni, Cu, Ti, Mn, Fe, Cr) that may be alloyed and / or engineered as composites with oxides and other organic hosts (e.g., nano-engineered Pt-C).

[0054] The process of the disclosure can be performed at an operating temperature of about 5 °C to 80 °C.Docket No.30275 / 59445

[0055] Flow rate of the water source, electrolyte having the reversible redox salt, and H2gas (inputs) into the electrochemical cell for acidification can be selected to be equivalent to the product of the Faradaic efficiency of the acidification process, the current density of the acidification process divided by the concentration of added protons and Faraday’s constant. The flow rate of the acidified water and the electrolyte with the reduced redox salt (inputs) into the electrochemical cell for basification can be selected to be equivalent to the product of the Faradaic efficiency of the basification process, the current density of the basification process divided by the concentration of added hydroxides and Faraday’s constant.

[0056] It has been advantageously observed that the process of the disclosure can remove precipitates that can form on the electrodes during the basification process. Precipitates formed during basification were observed to be removed from the electrode surface during the acidification step when performed in the same electrochemical cell. In embodiments of the method of the disclosure utilizing separate electrochemical cells for acidification and basification, it may be advantageous to reverse the polarization of the electrodes of the first and second electrochemical cells and switch performance of the acidification to the second electrochemical cell to thereby allow for removal of precipitates that may have built-up on the electrodes during the basification process. The beneficial removal of precipitates from the electrode surface provided by the reversible nature of the processes of the disclosure can greatly enhance the life cycle of the components of the electrochemical cell and avoid or at least reduce deterioration in performance resulting from precipitates. Electrochemical Performance Modeling

[0057] Fig.2 depicts predicted performance for different conditions (Table 1) of: (a) mass- transfer, which is dictated by the diffusion boundary layer thicknesses modeled for the gas and the liquid species, and (b) the pH-shift, with different low and high pH values at the end of acidification and basification respectively- (pHlow, pHhigh): (4, 10.7) and (6, 9.6). A smaller pH- shift was analyzed to quantify the tradeoffs in the extent of CO2captured for an operating scenario that is less prone to salt precipitation during basification as compared to the more typically experimented / modeled pH shift from 4 to 10.7. For all mass-transfer cases modeled, the different pairs of pH-shift conditions resulted in a constant offset in potential. This Nernstian shift is proportional to the pH change (∆V = 0.059∆pH) in the respective steps 0.12 V for acidification ending at pH 6 compared to pH 4, and 0.06 V when basification ends at pH 9.6 compared to pH 10.7 (Fig.2(a) and (c)). As a consequence, the smaller pH-shift condition had smaller reaction onset potentials. Amongst the various mass-transport cases modeled (TableDocket No.30275 / 59445 1), overlapping polarization curves were obtained when the liquid-side boundary layer is 10 µm thick, irrespective of whether the gas-side boundary layer was 10 or 50 nm thick. Therefore, the liquid-side boundary layer thickness limits performance. Hence, all results are only shown as a function of the liquid-side boundary layer thickness.

[0058] In the acidification step (Fig.2(a)), the voltage produced initially dropped rapidly with an increasing current density, because of kinetic limitations. The subsequent linear region in the polarization plot was due to significant ohmic losses. This ohmic region extends from 0-0.5 V for cases with the thin liquid-side boundary layer, δl = 10 µm, as the ohmic potential losses became increasingly significant at high current densities. When the boundary layer became even thicker (δl = 50 µm), the mass-transfer losses became dominant, which produced a steep voltage drop-off while the current asymptotes to near- steady values of ∼75 mA cm−2. Therefore, to access industrially relevant current densities that are at least 100 mA cm−2, excellent liquid-side mass transport must be in effect, translating to a boundary layer thickness near 10 µm. The Nernstian shift in the onset potentials leads to an almost 50% increase in the peak power produced during acidification at pH 6 compared to at pH 4. High Faradaic efficiencies (>91%) are predicted for the acidification step in Fig.2 (b), as the competing oxygen reduction is severely mass transport limited (0.12 mA cm−2) due to the low solubility of O2 in water (Table 2) Similar trends were generally observed in the basification step (Fig.2(c)) — higher operating currents densities are achieved at any potential with a decrease in the liquid-side boundary layer thickness. Distinct from the acidification step, two onsets were observed in the basification step (2(c)), where the early onset is due to competing reactions. The competing sulfate and oxygen reduction reactions were conservatively modeled to always be mass- transport limited. As a consequence, at small operating potentials / currents, these reactions were more dominant than the desired hydrogen evolution. This outcome was also reflected in the Faradaic efficiency in Fig.2(d), where a maximum Faradaic efficiency of only 84% is achieved for the basification cell.

[0059] To compute electrochemical energy intensity of the overall process, operating potentials from the polarization plots in Fig.2 were extracted while constraining operating currents to be equal during acidification and basification, i.e., jacid = jbase. While this was done for both pH-shifts modeled (Fig.15), Fig.3(a) and (b) presents results only for the smaller pH-shift of 6-9.6, which provides dual advantages of reduced energy intensities compared to the larger pH-shift, and a pathway to alleviate precipitation challenges duringDocket No.30275 / 59445 basification. Without competing reactions, energy intensities at non-zero current densities start near 1 GJ / tCO2and grow with increasing current density (Fig.3(a)). For the thinner boundary layer, the slope was nearly constant at higher current densities indicative of the ohmic resistance being the dominant contributor. Electrochemical energy intensities became increasingly dependent on the modeled ohmic drop at larger current densities; at 93 mA cm−2, the electrochemical energy intensity more than doubles from 2.3 GJ / tCO2 without an ohmic drop to 5.3 GJ / tCO2 at 5 Ω cm2. Therefore, ensuring small ohmic loses for Na+transport was observed to be important for lowering electrochemical energy requirements, which could be accomplished using thin or highly conductive membranes. In comparison, the thicker boundary layer case is mass-transport limited, and resulted in a dramatic electrochemical energy intensity rise at ∼70 mA cm−2.

[0060] While the changes in the energy intensities as a function of the boundary layer thickness are subdued in Fig.3(a), these changes get amplified when competing reactions are included in Fig.3(b). Generally, the inclusion of competing reactions increases the energy intensity, by up to 2.5 times, compared to fully selective reactions for the respective mass-transfer conditions. However, the relative extent of increase was larger for the thinner boundary layer, as it can access larger current densities for both desired and undesired reactions. Undesired reactions change the shape of the curves in Fig.3(b). Electrochemical energy intensities initially decreased with increasing current density because the Faradaic efficiency increases in Fig.3(b). For the thicker boundary layer, energy intensity subsequently increases with increasing ohmic resistance and mass-transport limitations, however for the thinner boundary layer case, this inflexion point is not observed for current densities smaller than 150 mA cm−2. Overall, the thicker boundary layer case achieved lower electrochemical energy intensities when competing reactions are modeled.

[0061] Fig.3(c) demonstrates the proposed process’s competitiveness on the basis of electrochemical energy intensities with state-of-the-art, pH-swing CDR processes for ocean water. Comparisons were made against mass-transfer conditions that led to the lowest energy intensity scenarios without (Fig.3(a)) and with (Fig.3(b)) competing reactions. Compared to the H2 looping process demonstrated by Yan et al., a modest reduction in en- ergy intensity by 35% is predicted for the process of the disclosure without competing reactions. This was a fair comparison, as this study used ideal saline solutions with only bicarbonate species added, which precludes the possibility of sulfate reduction (Fig.13). At even lower current densities (< 4 mA cm−2), greater reductions by 66% and 68% wereDocket No.30275 / 59445 projected compared to recent lab-scale demonstrations with bipolar membranes (Digdaya et al.) and chloride-mediated electrode materials, respectively (Kim et al.) Both studies report non-negligible competing reactions, including precipitation of Mg(OH)2, which is expected to be sensitive to the pH-swing and operating mass-transfer conditions in the experimental cells. Even so, the experimentally reported energy intensities are significantly smaller than theoretical projections with competing reactions — this reinforces the extremely conservative nature of our model assumptions for competing reactions and the low likelihood of competing sulfate reduction reaction to be mass-transfer limited.

[0062] Overall Energy intensity was evaluated as shown in the Examples to draw comparisons against non-electrochemical, DAC technologies, Fig.3(d) shows these comparisons by considering: (a) operation at industrially relevant currents (Jacid = Jbase =100 mA cm−2), (b) co-location with desalination, which eliminates a separate energy input for ocean water pumping, and (c) energy inputs for CO2 phase-separation from aqueous to gaseous state; and (d) gas-phase compression to 1 atm. Without ocean water pumping, the process of the disclosure was predicted to be is least on par with, if not more competitive than, state-of-the-art DAC technologies. However, when pumping is factored in, it becomes the dominant contributor and at minimum constitutes 69% of the overall energy intensity. Consequently, the overall process energy intensities of the process of the disclosure can exceed state-of-the-art DAC projections. While all the results presented in Fig.3 consider a pH-shift of 69.6, when energy costs for water pumping are included, a pH-shift of 410.7 becomes more competitive as it captures more CO2 on a volumetric basis.

[0063] Processes in accordance with the disclosure can achieve substantial reductions in the electrochemical energy intensities, up to 66% to 35% compared to experimental data reported for electrochemical pH-swing processes with bipolar membranes involving redox salts and H2 looping, respectively. Processes of the disclosure can be coupled with a renewable power supply, for example, from solar and wind, to power the CO2removal phase. Demand-flexibility can allow for improved integration with renewable energy uses, allowing the process to be flexible in the timing of the power consumption phase, for example, during high sun or high wind periods of the day.

[0064] Fig.4 illustrates how the process of the disclosure can be integrated with variable solar and wind electricity produced in California, USA, and how demand-flexibility translates to cost-savings with variable electricity pricing. The process of the disclosure can follow variable renewable electricity from solar and wind through its: (1) ability to alter the state ofDocket No.30275 / 59445 operation between power production (acidification) and consumption (basification), and (2) control of the duration, and equivalently the operating current densities, for the two steps. Fig.4 (a) plots the time-profile of the process operation where basification current densities are twice as large as acidification, but basification occurs for half the time duration as the acidification step (jbase= 2jacidand tbase= 0.5tacid). The energy-intense basification can occur during time periods with low electricity prices, while power-producing acidification can occur when the prices are high. Additionally, parasitic energy requirements assuming co-location with desalination are included, but are modeled to entirely take place during times of basification. In this way, extraction of the CO2 occurs as a semi-batch process on a daily basis, once the water source has been acidified. This operational scenario was applied for increasing amounts of CO2 captured and fitted to the CAISO solar and wind electricity production curve, which projects renewables in California can power an impressive 16 Mt of CO2 capture per year. However, co-location even with the largest desalination plant will constrain the maximum CO2 captured to be several orders-of-magnitude less, at 100 ktCO2. This large gap again underscores the necessity for creative solutions for ocean water pumping to enable larger-scale CDR.

[0065] Beyond integration with renewable electricity, Figs.4(b), (c), and (d) together reveal the cost-savings that this process can be accessed by flexibly switching between the different states of operation compared to a steady energy usage. For a ton of CO2 captured per day, Fig. 4(b) depicts the hourly power demand for the proposed process, with variable / dual-state operation between power production (acidification) and consumption (basification), and an equivalent constant power process with the same net energy consumption as the variable process. Fig.4(c) maps these power demand profiles to predict corresponding cost profiles based on representative time-of-day electricity costs for the CAISO grid (see Sec.8.5), where the differences in the shaded areas of the cost curves translate to operational (electricity) cost savings. The variable process with demand flexibility is able to arbitrage and ultimately benefit from fluctuating electricity prices compared to a steady power consumption CDR process.

[0066] Fig.4(d) predicts cost-savings as a function of the relative time spent in basification versus acidification. The time-averaged current density is held constant at an industrial viable, 100 mA cm−2. Therefore, as the time ratio increases towards 1, the current density of acidification increases, while for basification it decreases, until both are equal to 100 mA cm−2at a time ratio of 1. The minimum duration of basification time is 3.4 hours, amounting to a time ratio of 0.17, which is dictated by operating at the limiting currentDocket No.30275 / 59445 density for basification. Three different selling price scenarios are considered for the electricity produced during acidification, including no selling, selling at 50%, and selling at 100 % of the market price at any given time. For all pricing scenarios, electricity cost- savings decrease when increasing the relative duration of the basification step, as the power consumption process increasingly occurs during times with high electricity prices. At the minimum time ratio, substantial cost-savings up to $34 per tCO2 are projected even without any re-sale of electricity. This reveals the value of load flexibility enabled by the two operational states of this process. Compared to modeling no sale of electricity, cost-savings increase to the range of $19.3 - 65.2 per tCO2 when selling at 100% of the market price. As renewable energy continues to expand, the time-of-day fluctuations in the electricity prices will become more drastic. Thus, these monetary savings are only expected to grow in the future. While a comprehensive techno-economic analysis is outside the scope of this study, predicted cost-savings are contextualized with reported data for CO2 production costs from oceanic and atmospheric CDR processes. Operational cost-savings driven by demand- flexibility are substantial — 9–17% and 19–52% of CO2 production costs for oceanic pH- shifting co-located with desalination and DAC technologies respectively. EXAMPLES Example 1: Modeling of Current-Voltage Behavior

[0067] A zero-dimensional (0-D) equivalent circuit model was developed to predict current- voltage behavior for the proposed H2and salt cycling ocean water CDR process (Fig.1). Kinetic and mass-transfer overpotentials are modeled and additionally ohmic losses were included in these evaluations. Two scenarios were modeled by considering, (1) ideal selectivity, i.e., no competing reactions, and (2) parallel reaction pathways for desired and competing reactions. In the acidification step, the desired reactions are the H2oxidation reaction (HOR, Eq. (1b)) and ferricyanide salt reduction reaction (SRR, Eq. (1a)) at the anode and cathode respectively. Competing oxygen reduction reaction (ORR, O2+ 2e−^ 2H2O) was considered at both electrodes to account for the presence of dissolved O2 in aqueous solutions. For basification, the desired reactions are ferrocyanide salt oxidation reaction (SOR, Eq. (2b)) and hydrogen evolution reaction (HER, Eq. (2a)). Competing reactions included oxygen reduction reaction (ORR) at both the anode and the cathode, and additionally sulfate reduction (SO42-+ 2e- ^ SO32-) at the cathode (5). Sulfate reduction was included as a competing reaction during basification, as it is present at a substantial concentration (28 mM) in ocean water present in the catholyte. However, this is not relevant forDocket No.30275 / 59445 acidification, as ocean water is present as the anolyte. Oxygen evolution and chlorine evolution were not considered as competing reactions, as the modeled range of voltages is not oxidative at the anode to result in either of these reactions. Hydrogen oxidation (HOR) was not included as a competing pathway at the anode as it is not naturally dissolved in ocean water. Crossover of all species other than Na+is assumed to be negligible across the cation exchange membrane (Fig.1).

[0068] To determine current density, jop as a function of potential, Vop, for both operational states of acidification and basification, a set of equations were solved simultaneously. These equations enforce current equality and summation in series and parallel components respectively. Current equality was enforced at the anode and the cathode by Eq. (3) for acidification (acid) or basification (base) operational state. In Eq. (4), the net current density at the anode (ja) and the cathode (jc) is the sum of current densities for the desired and the undesired reactions, as these are parallel pathways. Because the desired and competing reactions are parallel pathways, these potentials are equal (Eq. (5)). Eq. (6) shows voltage summation across series components. Overall, the operating potential was evaluated in Eq. (6) as the difference in the anodic (Va) and the cathodic (Vc) potential, based on the sign- convention assumed, and accounts for the ohmic potential drop in the membrane (ηohm = jopROhm). The anodic / cathodic potential (Va / Vc) was modeled as the sum of the standard redox potential (U), which were adjusted for the bulk species concentrations by the Nernst equation (Eq. (1d, 2d); Table 2), kinetic (ηK), and mass-transfer (ηMT) overpotentials (Eq. (7)). jop = ja = |jc|, op = acid / base (3) ji = ji,desired + Σji,competing, i = a, c (4) Vi, desired = Vi, competing i = a, c (5) Vop = Vc − Va − jopROhm (6) Vi= Ui+ ηK,i+ ηMT,i,i = a, c (7)

[0069] Ohmic overpotential, ηohm,is linearly proportional to operating current density and the ohmic resistance of the membrane (Eq.6). Baseline membrane resistance was modeled as 2.18 Ωcm2, based on the correlation determined by Zhu et al. assuming a NaCl concentration of 0.5 M to represent ocean water. While other literature reports higher specific resistances closer to 5 Ωcm2for CEMs in electrodialysis systems, lower values of 1.05Ω cm2for 0.55 NaCl solutions have been obtained through modeling.Docket No.30275 / 59445

[0070] The kinetic overpotential (ηK) determines the driving force of electron-transfer reactions, and was evaluated based on concentration-dependent, reversible Butler-Volmer kinetics (Eq. (8)). The rate expression in Eq. (8) is dictated by the reaction-specific exchange current density, (jo), charge-transfer coefficients(αa / c),and additionally includes dependency on the surface concentrations of the reactant / product species. Surface species concentrations were obtained as a function of the desired reaction current density and the limiting current densities as in the term (1 −^^^) and the bulk concentrations related to a reference concentration of the respective species. Exchange current density, reference concentrations and charge-transfer coefficients were determined from reported data for platinum-loaded GDEs and platinum electrodes (Table 2). Eq. (8) is only relevant for the desired reactions during acidification / basification. For the competing reactions, the kinetic overpotential is assumed to be 0 (see Fig.5) as these reactions are always assumed to occur at mass-transfer limited currents, and therefore, ji,competing= ji,l(Eq.10).reduced / oxidized species respectively, jl,c / a is the limiting current density of the cathodic / anodic reaction, νc / a is the stoichiometric coefficient for the reduced / oxidized species, n is the number of electrons transferred in the redox reaction.

[0072] The mass-transfer overpotential (Eq. (9)) arises due to species concentration differences between the electrode surface and the bulk solution, and is in turn dependent on the operating current and the limiting current densities of the respective oxidized (o) and the reduced (r) species for any reversible redox reaction. The maximum rate with which reactants diffuse from the bulk electrolyte to the electrode surface, and equivalently for products to diffuse away from the electrode surface is dictated by the steady-state diffusive fluxes in the concentration boundary layer. Therefore, anodic / cathodic limiting currents (jl,a / c) were quantified using Eq. (10), where Dr / o is the reduced or oxidized species diffusivity, and δa / c is the anodic / cathodic diffusion layer thickness. Competing reactions (O2 and SO42-reduction reactions) were conservatively modeled at the limiting current in Eq. (10); therefore, ηK does not need to be calculated, as the entire overpotential is ηMT.Therefore, Eqs. (8)-(9) were not solved for competing reactions. Species concentrations and other parameters required to compute the respective limiting currents are listed in Table 2. For theDocket No.30275 / 59445 redox salt, the concentrations were assumed to be corresponding to their solubility limits at room temperature — for the Na3Fe(CN)6and Na4Fe(CN)6species it is 0.5 M and 0.56 M respectively. Dissolved gas species concentrations for H2 and O2 were obtained from their solubility and based on assuming equilibration with 1 atm of the respective gases above the solution. For H2evolution that requires water as a reactant (when the product is OH−), the mass-transfer limited current is a much larger value and on the order of (106A cm−2).

[0073] Using convergeddetermined as the minimum between the current efficiencies (ji,desired / jop) at the anode and cathode (Eq.11). Additionally, because the acidification produces power, its area-specific power density can be computed as the product of the operating voltage and the current density during acidification, Pacid= jop,acidVop,acid.

[0074] Different modelingbetween acidification and basification, (2) ohmic, (3) mass-transfer, and (4) reaction selectivity. Variations in the pH-shifts influence the Nernstian potential, U, for the acidification / basification steps, (Eqs. (1d), (2d), Table 2) and therefore the minimum required potential to drive the net reaction. Three different ohmic resistances were considered to capture variations in the reported resistances for CEMs, including the ideal case of no ohmic drop across the species. For all reactions except hydrogen oxidation, which involves gaseous H2as the reacting species, limiting currents were obtained based on the liquid boundary layer thickness as the reactants are in aqueous form. Baseline case models δg= 50 nm and δl= 50 µm for gas and liquid species respectively. A best-case scenario with much thinner boundary layers of δg= 10 nm and δl= 10 µm was also modeled. Finally, reaction selectivity is either assumed to be ideal with a 100% FE at all electrodes, or by also accounting for competing reactions.Docket No.30275 / 59445 Table 1: Thermodynamic, ohmic, and mass transport properties and competing reaction extends implemented in the equivalent circuit model Parameter Values UnitsExample 2: Modeling Energy Intensity

[0075] The net electrochemical energy intensity (EEI) normalized by mass of CO2 removed was computed in Eq.12, and was primarily dictated by the operating current densities (jacid and jbase), operating potential (Vbase and Vacid), and the time duration (tacid and tbase) during acidification and basification respectively. Current densities and the time duration were related to each other with constraints of: (a) both steps treating the same total volume of ocean water to ensure that there was no acidified water accumulation on a daily basis, Qacidtacid = Qbasetbase, and (b) the process cycling between the two states over a day, i.e., tacid+tbase = 24 hrs. Qacid and Qbase are the electrode area specific volumetric flow rates (Ls-1m-2) of the ocean water treated in the acidification and basification steps, respectively. The flow rate Qacid, is related to the rate of desired reaction for acidification (i.e., FEacidjacid) and the corresponding concentration of protons added (CH+acid) by Eq.13.Docket No.30275 / 59445

[0076] In Eq. (12),and calculated in Eq. (14), where CCO2,aqis the concentration of dissolved CO2in ocean water that is dictated by the pH shift, ηcollis the percentage of dissolved CO2collected by vacuum pumping (90%) and MCO2is the molar mass of CO2.

[0077] Inincludes ocean water pumping energy intensity (PEI) and vacuum pumping for CO2phase separation from its aqueous form (VPEI), as described in Eq. (15). Recent reports from the desalination community point to energy intensities for ocean water intake varying widely with plant loca- tion inland, its altitude, type of intake, and motor and pump efficiencies. Additionally, it has been proposed in the literature to co-locate an ocean CDR plant with desalination, as the ocean intake and outfall already exists, which results in a water pumping energy intensity value of 0. Therefore, a wide range of ocean pumping energy intensities were analyzed from 0 – 1.358 kWh m−3. VPEI accounts for both pre-treatment involving degassing of O2 and N2 prior to acidification, to ensure a purified CO2 stream, and the phase-separation / degassing of CO2 post acidification. The energy required for CO2 phase-separation was estimated from the adiabatic compression work needed to compress from the CO2 equilibrium pressure of acidified ocean water and water vapor, 0.1059 atm at 25°C, to atmospheric pressure, accounting for a vacuum pump efficiency of 34% and only 90% removal of dissolved CO2 at 0.77 GJ per tCO2. Vacuum degassing of O2 and N2 was similarly estimated at 0.41 GJ per tCO2, yielding a total VPEI = 1.18 GJ per tCO2.Docket No.30275 / 59445

[0078] The net process energy intensity, EI, for ocean water CO2capture was computed as the sum of the electrochemical and the parasitics, EI = EEI + Parasitics Example 3: Process of the Disclosure in a GDE Flow Cell

[0079] The redox salt electrolyte used in the example included 0.5 M Fe(CN)63- / 4-ions. All salt masses were measured using an analytical balance with ±0.1 mg accuracy. For the acidification process, 0.5 M K3Fe(CN)6 (Sigma-Aldrich ACS Reagent ≥99.0% purity) solution was prepared from DI water with a resistivity of 18.2 MΩ.cm. For basification, 0.5 M Na4Fe(CN)6 (Sigma- Aldrich ≥98.0% purity) was prepared from 18.2 MΩ.cm DI water. Artificial oceanwater was prepared from Instant Ocean aquarium salt. To simulate ocean water, 35.95 g / L of the Instant Oceanwater Aquarium sea salt is added to DI water with a conductivity of 18.2 MΩ.cm.

[0080] Two electrodes and one membrane were required for the assembly of a single flow chamber for this reaction. A single piece of Nation NR-211 membrane was used, which had been soaked in 0.5 M NaCl for at least 24 hrs. The membrane was inspected for visual damage or leakage prior to an experiment. Each LSV experiment includes 5-6 flow rate scans (0, 10, 20, 30, 40, 50 ml / min for basification or 0, 20, 40, 60, 80 ml / min for acidification). A wider range of flow rates was chosen for the acidification experiments, due to the longer voltage scan range that was required with the same finite solution volume from the syringe pump. The syringe pump was a Harvard Apparatus PHD Ultra model fitted with 2100 ml syringes. For each LSV experiment, either new or newly cleaned electrodes were used. For acidification experiments, a carbon paper electrode (Sigracet 22 BB) was used for the redox salt, while a gas diffusion electrode deposited with Pt is (0.5 mg / cm2Pt-C 60% coated on Sigracet 22 BB carbon paper from Fuel Cell Store). New pieces of Carbon paper or GDE were cut for each experiment. For basification, similarly a new GDE was cut for each experiment for the electrode touching the ocean water; however, the electrode for the redox salt was in this case a platinum foil. Prior to each experiment, the platinum foil was electrochemically cleaned using the following procedure. The electrode was soaked in 0.1M HCl and was cycled from -0.7V to 0.3V vs Ag / AgCl electrode for 20 cycles at a scan rate of 100 mV / s. At the endpoints, the potential was held for 1 second. The produced curve shape was checked for the following features to ensure the Pt surface did not contain impurities: H2evolution peak beginning around -0.25V vs. Ag / AgCl and H2underpotential deposition signal for adsorption / desorption between -0.25 and 0.4 V vs. Ag / AgCl. Over the course of one experiment, multiple LSV scans were performed, one at each probed flow rate. The trial order of the flow rates was randomized, so that for each experiment the order of the tested flow rates was for basification, 20, 30, 0, 50, 10, 40 ml / min, and for acidification, itDocket No.30275 / 59445 was 40, 60, 0, 20, 80 ml / min. The randomization of the flow rates with the trial number allowed for deconvolution of changes due to trial changes with changes caused by the differing mass transport condition.

[0081] To assemble the test cell prior to its operation, the multiple layers of the cell were clamped together. Referring to Figure 6, first, an insulating plate with a gas port provided the base. Next a stainless-steel current collector was placed, which enables the connections to the potentiostat. On top of the current collector was a stainless-steel mesh, a stainless-steel foam, which was topped by the 0.5 mg / cm2Pt-C Gas Diffusion Electrode. Next, a flow channel gasket was placed, followed by the ocean water flow channel, followed by another flow channel gasket. A Nafion NR-211 membrane sat between the two flow plates and between two flow gaskets. On top of the redox flow channel, another flow gasket was used to seal the flow channel to the platinum foil. Finally, the second non-conducting plate was used to stabilize the opposite side of the cell during the clamping. When conducting an LSV experiment, gas must be flowing through one gas port on the Pt-C GDE side. The other port was left open to vent the gas without building up pressure inside the gas compartment. For acidification tests, the supplied gas was Hydrogen and for the basification tests an inert gas, without significant levels of Oxygen was used. Nitrogen or Argon may be used. Once the cell was assembled, both electrolyte and gas flow were connected. Electrolyte flow was connected through the Luer lock fitting. Oceanwater was flowed through the flow plate touching the Pt-C GDE and the redox salt was flowed through the flow channel touching the Pt foil. Inlets of the electrolyte flow were connected at the bottom of the flow channel and the outlet tubing was connected at the top of the flow channel. With electrolyte tubing connected, electrolyte was input to the flow chamber. Approximately 5 ml was needed of both electrolytes to fill the flow chambers. Once the flow chambers were full, the gas supply was connected to the outside gas port. To check that gas was flowing, the outlet gas port was covered. If gas bubbles emerged from the outlet of the electrolyte, the gas was determined to be successfully flowing. Additional electrolyte was then pushed through the flow chamber to ensure it was filled again.

[0082] Prior to running each experiment on the potentiostat, the Open Circuit Voltage (OCV) and cell resistance was recorded. Typical OCV values for acidification were >0.8 V. For acidification, the OCV was important to maintain at least this high. If the OCV was low, there was likely not enough H2supply to the Pt-C GDE. In this example, it was necessary to adjust the gas flow until a better OCV was obtained if it became too low. For basification, the OCV reading was less critical. It generally started near 0.6 V, but could continuously fall over multiple trials.Docket No.30275 / 59445 For either experiment the cell resistance reading was generally near 3 Ohm. If it was significantly larger, the electrode-potentiostat connections were adjusted, the screws which clamp the cell together were tightened, additional electrolyte was pushed through the cell, and / or the gas pressure flowing into the cell was reduced.

[0083] The results for the acidification LSV experiment following the above procedure are shown in Fig.7. From the experimental data, it was observed that there was an initial onset, followed by a largely straight line, indicating the cell was ohmically limited. Different effects were observed from the different flow rates though, indicating that the different flow rates do have different mass transport limitations. As the flow rate increased, the max current density reached also increased, which verified the cell was also mass transport limited. These solid-colored experimental curves were also compared to dashed model curves. The curve indicated by “Model Results” shows the closest match to the experimental data at the higher flow rates, while the thicker boundary layer modeled in the curve labeled “δl = 200 μm” better matched with the lower flow rates. This curve shows a largely straight trend which matches the data for a majority of the current densities. However, at the very low current densities the experimental data had a significant higher voltage than what the model projects. Since the model simulates the reaction occurring at only 1 pH (marked at pHlow), it cannot account for the initial changing pH that was observed with the solution input at ~pH 8.1. The pH starts the scan at low current densities with a higher pH, which quickly dropped with the increasing rate of reaction. The explanation was corroborated by the model results from the pHlow = 6, pHhigh = 9.6 scenario. This result onsets at a higher pH more similar to the onset of the experimental results. Therefore, this initial pH of the solution from the modeled pH explains the deviation between the model and experimental results at low current densities.

[0084] The basification LSV experimental and model results are shown in Fig.8. Again, similar trends were observed in terms of the ohmic and mass transport conditions of the experimental data. The model results at a cell resistance of 18 Ωcm2, a δl= 100 and 200 μm, and pHlow= 4, pHhigh= 10.7 scenario match most closely with the data at 30 and 0 ml / min, respectively. However, there was again significant deviation between the model and the onset of the reaction. This deviation was even more extreme than in the acidification scenario, as the reaction onsets almost 1 V earlier than the model onset. Without intending to be bound by theory, it is believed that this could come from possible competing reactions such as ORR and sulfate reduction, which onset at lower applied potential. This explanation matched the model as the competing reactions were expected to be significant in the basification scenario. The initialDocket No.30275 / 59445 linear increase shown in the basification model data was from the mass transport limited competing reactions of sulfate reduction. Since the kinetic onset of these reactions was not modeled, there was a step change as opposed to gradual increase. Once the competing reactions were accounted for in the model, the model did match the LSV data well. Specifically, the 30 ml / min and 0 ml / min results are well mirrored for the δl= 100 and 200 μm, respectively.

[0085] To reinforce the validity of extrapolating the scan and model data to the cyclic process operation, the stability of the cell components over multiple cycles is necessary. Since this is a pH-shifting process with ocean water, during the pH increasing step, basification, it is likely to have fouling on the electrode surface. Fouling during basification has been well reported in the literature and can increase the cell resistance, decrease the transport to the reacting surface, and prevent further reaction from taking place. However, the reversible cell design of the combined H2 and redox salt cycling process of the disclosure was observed to allow for the polarity of the electrode to switch intermittently, thereby helping to clean the electrode surface. A constant current experiment was performed to validate this finding. The constant current experiment was initially performed with a new Pt-C GDE, as specified above. Further details of this experiment are shown in Fig.9. First, basification was performed at a constant current density of 14.6 mA / cm2for 4 mins to treat 60 ml of ocean water. The pH increased from 7.19 to 8.99. Significant precipitation was observed on the electrode surface and microscope images show the precipitation fills the cracks that cover the GDE surface. The same electrode was used for the acidification experiment. Following the acidification at the same current density for the same amount of time, the electrode looked visibly cleaner with only small regions of precipitate remaining. Microscope images on these remaining regions revealed that even though some precipitate may still remain on the surface the cracks were no longer filled with precipitation, but instead appeared clean. This demonstrated that the process in accordance with the disclosure is capable of intermittently cleaning itself to maintain optimal operation performance.

[0086] From the above cycling experiment and from the best fitting model data from the LSV experiments, the electrochemical energy intensity was calculated, shown in Fig 10. While CO2was not explicitly measured during the experiment, it can be calculated by assuming a 2.2 mM DIC concentration of the 60 ml of oceanwater treated and assuming 90% of the present DIC can be captured in the form of CO2(comparable to the model assumptions). Plotting the electrochemical energy intensity compared to notable literature works showed that the electrochemical energy intensity of the process in accordance with the disclosure was higherDocket No.30275 / 59445 than those reported in the literature. However, the experimental obtained electrochemical energy intensity was lower than the modeled results. Example 4: Additional Modeling information Dissolved Inorganic Carbon as a function of pH

[0087] Dissolved Inorganic Carbon (DIC) species buffer changing pH in ocean water and exist in different states depending on the pH. Using equilibrium equations (Eqs. S1 - S2) and equilibrium constants (K1= 10-5.86and K2= 100-8.92),the transitions between different states of the molecule can be described as a function of the pH. The proportion of CO2, HCO3- and CO32-at a constant value of DIC (2.2 mM assuming natural oceanwater) can then be calculated, at any given pH as shown in Figure 11. CO2 species are dominant at pH lower than 6, making up nearly all of the DIC at pH lower than 4. Bicarbonate, HCO3- , species are dominant at near neutral pH and carbonate, CO32-, species are present as nearly 100% of the DIC above pH 11.Calculation of pH shift range

[0088] The total pH shift is calculated to determine prospective thermodynamic cases for the devices operation (both low to the acidification pH, and high to the basification pH). The total pH shift is calculated in three steps: (1) the addition of protons through acidification, (2) the reduction of the DIC through the CO2extraction, and (3) the addition of hydroxide in the basification step. First, for the acidification step, the amount of added protons is calculated assuming a final desired pH (CH+). The final concentrations of CO2, HCO3−, and CO32−are found using the equilibrium equations (Eqs. (S1) - (S2)) and Eq. (S3), given the DIC remains constant.Docket No.30275 / 59445

[0089] Following the calculation of the DIC contents, the COH−can be calculated from the ionization constant for water (pKw), using Eq. (S4). At a reference temperature of 25 °C and a salinity of 35, the pKw is shown to be 13.22.

[0090] Finally, the required concentration of added H+can be calculated based on the change in the DIC species and pH, from Eq. (S5).

[0091] by removing the fraction of DIC that is present as CO2 at the new acidified pH, as in Eq. (S6).

[0092] species.

[0093] Additionally, from summing the species in the carbonate alkalinity, it can be determined the extent to which removing the CO2cause the other species concentrations to shift their equilibrium, without any net addition of CH+or COH−, from Eq. (S7).

[0094] concentrations of all species following the removal of CO2:

[0095] neutralize the previously added CH+. For this system, equivalent basification to the extent of acidification is considered, described in Eq. (S8).

[0097] This constraint is necessary to ensure that overall the ocean water will not be either net acidified or basified by the CO2 removal process, instead only the concentrations of DICDocket No.30275 / 59445 species will change. This added COH−will contribute to changing the carbonate alkalinity as shown in Eq. (S9).

[0098] be obtained following the basification step. Using this methodology, the theoretical solution pH at all points in this reaction scheme can be tracked and are shown in Fig.12. On the x-axis, the pH of the solution exiting from the electrochemical process is shown, where acidification is shown in blue and basification in red. Following acidification, the y-axis describes the equilibrium pH after the CO2 species have been removed. Following the green line from the blue to the red curve traces the expected pH shift from equivalent basification, shown on the x-axis again. In Fig.12, the low pH achieved upon exiting the acidifying step and removing CO2 are shown by the line labeled (1) on the x and y axes respectively. Note that pH increases simply by removing CO2. Corresponding arrow heads on the curves indicate the input (on (1) curve, y axis) and output (on (2) curve, x axis) pH that would be achieved through the basification process. The dashed line labeled (3) at pH 8 indicates the initial starting pH of ocean water. The dashed line labeled (4) at pH 10 indicates the suggested maximum pH before significant precipitation as described by Sharifian et al.

[0099] To maximize the amount of CO2 extracted during acidification, a pH shift from 8.1, which is the natural pH of ocean water, to 4 is all that is necessary. However, when basification follows to the same extent as acidification, the pH of the decarbonized ocean water does not return to 8.1. As shown in Fig.12, when the pH is acidified to a value of 4, the maximum pH achieved at the end of basification is 10.7, because of the loss of the buffering CO2species. This could be problematic as the tendency of hydroxide precipitation increases dramatically above pH 10. This increases component fouling challenges and has been widely observed. The process of the disclosure can provide unique benefits for precipitate management. Minor precipitation caused by local concentration gradients can be cleaned off by reversing the flow and polarization between the two steps, similar to reversal techniques used in electrodialysis. To minimize bulk precipitation, the extent of acidification can be varied. For instance, shifting to a pH of 6 during acidification will result in a pH of 9.6 at the end of basification. However, operating the device in this method would reduce the total amount of CO2extracted per unit volume of treated ocean water. Whereas 99% of inorganic carbon is present as CO2at a pH 4, this is more than halved to 43% at a pH of 6Docket No.30275 / 59445 (Fig.11). Therefore, two cases of pH shifts have been considered: (1) the previously explored pH 4 - 10.7 and (2) a less extractive, but possibly more practical pH 6 - 9.6. Model formulation and parameters

[0100] Kinetic overpotentials are evaluated based on concentration-dependent, reversible Butler-Volmer kinetics (Eq. (8)); Eq. (8) is only relevant for the desired reactions during acidification / basification. Redox salt ([Fe(CN)6]3−) / ([Fe(CN)6]4−) and H2 species (H+ / H2)) were modeled to be reversible with symmetric charge-transfer coefficients (αc = αa = 0.5) for the cathodic / anodic reactions in Eq. (8). jo for ([Fe(CN)6]3−) / ([Fe(CN)6]4−) was taken from Daum et al. for a platinum electrode. Hydrogen oxidation and reduction were modeled on a platinum electrode with pH-specific jo values taken from Zheng et al. and are listed in Table 1. The temperature for all reactions is assumed to be 300 K. However, due to the pH shifting nature of the device, both H2 oxidation and evolution will not take place at the same pH. In the acidification step, H2 oxidation was modeled at low pH (4 or 6), whereas H2 evolution during basification occurs at the high pH (10.7 or 9.6). Therefore, while these reactions were modeled to be reversible (Eq. (8)), H2 oxidation is only calculated at more positive potentials than its onset at low pH. Equivalently, H2 evolution occurs only at potentials more negative than its onset at high pH. Additionally, because pH-specific jo values are modeled, the ratio of bulk and reference concentrations is assumed to be 1 in Eq. (8).Docket No.30275 / 59445 Table 2: Equivalent Circuit Model ParametersIndividual Reactions Modeling Results

[0101] Fig.3 shows the predicted current-potential curves for the various reactions (desired and undesired) considered by the best fluid flow conditions: δg = 10 nm and δl = 10 µm. Neg-Docket No.30275 / 59445 ative and positive currents signify cathodic and anodic reactions, respectively. Amongst the desired reactions during acidification, the [Fe(CN)6]3−reduction is more mass-transfer limited than the HOR (jl,c = −432 mA cm−2), indicating that [Fe(CN)6]3−reduction will limit the composite cell currents. Despite the low solubility of H2in water, HOR is able to achieve currents significantly larger than 1000 mA cm−2(jl,a=∼ 7.5A cm−2), due to the 3 orders-of-magnitude smaller boundary layer thickness modeled for the gas diffusion layer. ORR, included as a competing reaction at both the anode and the cathode, is severely mass-transfer limited at a current density of -0.12 mA cm−2. Therefore, even at large enough operating potentials it does not significantly compete with the desired reactions. For the basification step, the [Fe(CN)6]4−oxidation is more mass-transfer limited than the HER, dictated by the smaller concentration of [Fe(CN)6]4−species. HER is essentially only kinetically limited due to the large concentration of water; the lower limiting current of the [Fe(CN)6]4−) species (jl,a= 354 mA cm−2) will limit the composite cell current. Since the SO42-reduction competes significantly with the HER at a current density of -58 mA cm−2, the maximum Faradaic efficiency for the basification device will be limited. The difference in pH of the ocean water between the acidification and basification steps of produces the shift between the onsets of the HOR and the HER curves in Fig.3 indicated by the double arrow. In contrast, there is no offset in the [Fe(CN)6]3− / [Fe(CN)6]4−polarization curves as this is not a proton-coupled redox reaction. For both the acidification and basification steps, the [Fe(CN)6]3− / [Fe(CN)6]4−redox reactions will limit the maximum current densities. Therefore, the liquid boundary layer thickness, δl, is a crucial performance determining parameter, whereas the sensitivity to the gas boundary layer thickness, δg, is not significant for thickness <100 nm. Competing reactions are modeled as only mass transport limited and therefore, have a step-wise onset at the thermodynamic potential. Sulfate reduction onset immediately increases to its limiting current density at the voltage of -0.64 V vs SHE, shifted from the standard redox potential of -0.93 V vs SHE assuming a pH of 8.1 for ocean water and based on the relative concentrations of sulfate / sulfite in ocean water (Table 2). ORR is sustained at its limiting current density of only -0.12 mA cm−2across the entire shown voltage range.

[0102] Possible competing reactions of oxygen evolution (O2 + 2H2O + 4e−^ 4OH−, E° = 1.229 - 0.059 · pH) and chlorine evolution (Cl2+ 2e−^ 2Cl−, E° = 1.358)as they occur outside of the considered voltage range (V vs. SHE) listed in Fig.13. In Fig.13, it can be seen that ORR is mass transport limited over the entire voltage range shown. CurrentDocket No.30275 / 59445 density values shown in Fig.13 are representative of the best fluid flow conditions evaluated (δl= 10 µm and (δg= 10 nm) Electrochemical performance results without competing reactions

[0103] Fig.14 shows the predicted performance for the acidification and basification steps for the different (1) mass transport conditions and (2) pH shift conditions explored (Table 1) without considering competing reactions. Similar onsets and curve trends are observed for the polarization data in Fig.14 (a) and (c), compared to the electrochemical performance with competing reactions (Fig.2). Major differences include the elimination of an initial competing reactions onset in Fig.14(c) and the plotted Faradaic efficiency at a constant 100% following the reaction onset (d). Faradaic efficiencies are initially 0%, as this is prior to the reaction onset, but become 100% once the desired reaction onsets

[0104] Different mass-transfer conditions and therefore limiting currents were modeled by varying the liquid-side boundary layer thicknesses, (δl). Two different pH-shift conditions were modeled for the low and high pH values at the end of acidification and basification, that were respectively constrained by the concentration of H+and OH−additions to be equal during both steps – (pHlow, pHhigh): (4,10.7) and (6,9). Baseline parameter values listed in Table 2 was used in the equivalent circuit models with consideration of only desired reactions. There were differences in the current and potential ranges plotted between Fig. 14a acidification and 14b basification steps. Electrochemical energy intensity at pH 4-10.7

[0105] Similar to the pH shift of 6 - 9.6 (Fig.3), the electrochemical energy intensity (EEI) for the larger pH shift of 4 to 10.7 was also plotted as a function of current density. Fig. 15(a) shows the EEI without any competing reactions at both liquid boundary layers considered. For this scenario, both boundary layers nearly overlap until about 50 mA cm−2. After this they begin to deviate, where the thicker boundary layer has a dramatic upswing in the EEI, as it becomes mass transfer limited. The thinner boundary layer increases linearly though, indicating the impact of the ohmic resistance on the EEI. Comparatively, Fig.15(b) illustrates the EEI when mass transfer limited competing reactions are included. In this case, the EEI is initially decreasing as the Faradaic efficiency of the reaction is increasing. For the thicker boundary layer, a minimum is reached near 4.5 GJ / tCO2, before later increasing due to the increasing ohmic drop and the mass transport limitation. The thinner boundary layer does not reach a minimum within the current density range shown in Fig.5.Docket No.30275 / 59445 Energy intensity of considered parasitics

[0106] Vacuum Degassing of CO2: To find the vacuum required, the reduction in CO2partial pressure was calculated using Henry’s Law. Henry’s constant of CO2 in seawater at 25 °C is 2.947E-2M / L atm. At acidic pH (∼4) approximately all of the dissolved inorganic carbon is present as CO2, meaning the concentration of dissolved CO2is about 2.2 mM. The equilibrium partial pressure for this concentration is 0.0746 atm, based on Henry’s Law (Eq. (S10)).

[0107] whereconstant for CO2in ocean water, and PCO2is the partial pressure of CO2.

[0108] The vapor pressure at 25°C is 0.0313 atm. The sum of the vapor pressure and PCO2 yields the total vacuum pressure required of 0.1059 atm for equilibrium removal of CO2 in a wet stream, as shown in Eq. (S11).

[0109] to a per mass of CO2 basis, the molar mass of CO2 (44 g / mol) and the mole fraction of CO2, x, is needed and the later can be deduced from the ideal gas law and the PCO2 as in Eq. (S12). To account for imperfect separation with only 90% removal of the available CO2, the partial pressure of CO2 in Eq. (S12) will be 90% of the ideal partial pressure of CO2.

[0110] where nCO2is the number of moles of CO2 removed and ηtotis the total number of moles.

[0111] By knowing the necessary equilibrium pressure for degassing, the vacuum energy to achieve this pressure can be estimated by finding the adiabatic work required of the com- pression process, supposing an efficiency of a vacuum pump. The adiabatic work of expansion can be written as Eq. (S13)Docket No.30275 / 59445

[0112] ^ is the adiabaticconstant of 1.3 for CO2, P1 is the pressure at state 1 (low, inlet pressure), and P2 is the pressure at state 2 (high, outlet pressure). ηvp is the estimated vacuum pump efficiency, which is needed to estimate the actual work from the adiabatic work. Budama et al. have compiled vacuum pump efficiencies based on the desired low pressure, with applications to thermochemical energy systems. Based on the low pressure that is required by this system, 0.1059 atm, the estimated vacuum pump efficiency, ηvp, is 34%. The overall work to vacuum degas the CO2 from the ocean water is then estimated at 0.77 GJ / tCO2 (VPEICO= 0.7777 GJ / tCO2).

[0113] Vacuum pre-degassing of O2and N2: To ensure that the CO2wet stream is of high purity and is not mixed with significant concentrations of O2and N2, vacuum degassing of the ocean water as pre-treatment prior to acidification is possible to removed dissolved O2and N2. At a salinity of 35 and at 298 K, the N2and O2solubility in ocean water is projected at 10.1 mg / L and 6.4 mg / L, respectively. These mass concentrations give molar concentrations of 0.36 mM and 0.20 mM for N2and O2respectively. Considering fully acidified ocean water, the dissolved CO2content is projected to be near 2.2 mM, which would yield about 79.7% pure CO2without any pre-degassing of O2 and N2. For partially acidified ocean water at pH 6, the dissolved CO2content is 0.95 mM, which would yield 62.9% pure CO2. At either pH shift, the content is below the proposed purity of 95.5% of CO2 for pipeline transportation as proposed by de Visser et al. Therefore, pre-degassing O2and N2from the ocean water will be necessary.

[0114] Assuming at least a 95.5% pure stream is desirable, the amount of dissolved O2and N2 must be reduced by about 12.5 times. The necessary reduction in dissolved gases leads to equilibrium partial pressures of O2 and N2 at 0.0168 and 0.0632 atm, respectively. The sum of these partial pressures and the saturation vapor pressure at 298 K yield a vacuum pressure of 0.1113 atm required to degas air from ocean water to achieve 95.5% pure CO2, shown in Eq. (S14).

[0115] The adiabatic compression work can again be used to estimate the degassing work. However, this gives the work in units of J per total moles of the air and water vapor mixture, which undergo the compression process. To reduce the dissolved gas amount byDocket No.30275 / 59445 the goal of 12.5 times, from 1 liter of ocean water, 0.331 and 0.184 mmol of N2and O2will be removed, respectively. Dividing the total amount of dry air removed per liter of ocean water by the mole fraction (Eq. (S15)), yields the total amount of moles compressed to pre-degas 1 liter of ocean water.

[0116] Withtotal amount of N2, O2, and water vapor compressed to pre-treat 1 liter of ocean water is 0.716 mmol. By determining the amount of mixture to be vacuumed degassed per liter of water, the energy intensity of the degassing process per unit mass CO2can be calculated assuming the concentration of CO2in the degassed water will be 0.96 mM (conservatively, assuming acidification to only pH 6) with a collection efficiency of 90%. This calculation yields 18800 moles which are vacuum compressed per tCO2removed from ocean water.

[0117] The work of the pre-treatment degassing process can be calculated using the adiabatic work of expansion in Eq. (S16)., adiabatic constant of 1.4 for air, P1 is the absolute low pressure at state 1 (0.1113 atm), and P2 is the absolute high pressure at state 2 (1 atm). With the similar vacuum pressure, the same vacuum pump efficiency of 34% can be estimated. The adiabatic compression work for pre-degassing can be calculated using Eq. (S13). This yields the work for degassing the stream, VPEIAir = 0.41 GJ / tCO2. The total vacuum pumping work (VPEI) is then equal to the sum of these parts, shown in Eq. (S17), yielding 1.18 GJ / tCO2.

[0119] Ocean water pumping: Ocean water pumping for this process is an additional energy intense step that must be considered when estimating the overall energy intensity of ocean water CDR processes. Extensive research on the energy intensity of ocean water pre-treatment and pumping has been completed by the desalination community. According to Rao et al., aDocket No.30275 / 59445 wide range of expected energy intensity values are possible for both pre-treatment and water pumping, separately. Estimates for water intake range between 0.05 - 0.58 kWh / m3of water, depending on the type and design of intake. Pre-treatment estimates range between 0.223 - 0.778 kWh / m3, when multiple kinds of pre-treatment are summed including chlorination, flocculation, sedimentation, dissolved air flotation, membrane and media filtration. In total, this yields a wide possibility for total energy intensities, from 0.275 - 1.358 kWh / m3. Since there is a wide range of predicted energy intensity values, several different cases for water pumping are considered: 0.275, 0.75, and 1.358 kWh / m3. Additionally, there is the possibility of co-locating an ocean water CDR plant with desalination, so that the CDR plant would not have to treat ocean water pumping as an additional parasitic, as the pumping would already be completed for desalination. In this case, the energy intensity for the CDR process being implemented at an already functioning desalination cite would be 0 kWh / m3. Therefore, four total ocean water pumping cases are considered for both pH shifts: PEI = 0, 0.275, 0.75, and 1.358 kWh / m3.

[0120] Fig.16 shows the electrochemical energy intensity and the considered parasitics are compared across the four different pumping cases and two different pH shifts. The electro- chemical energy intensity is based on an industrially relevant current density of 100 mA cm−2for both the acidification and basification steps, for the scenario without competing reactions. Competing reactions are not considered here as it has been determined that competing reactions need to be limited in order to make this process viable. Water pumping amounts per mass of CO2 collected are determined by the amount of dissolved CO2, determined by the desired pH shift, as described by Eq. (14). Dividing the PEI by mCO2 yields the water pumping contribution to the parasitic energy usage, as described by Eq. (15). While the pH shift of 6-9.6 has reduced energy intensity when co-located with desalination (No WP), this trend shifts if any water pumping is necessary. In such cases, the pH shift of 4 - 10.7 is less energy intense, as less water needs to be pumped to remove the same amount of CO2. Therefore, when water pumping is considered (i.e., a system not co-located with desalination) water pumping becomes the driving factor behind the overall energy intensity. If a stand-alone ocean water CDR plant is considered, the energy intensity of water pumping would need to be minimized as much as possible. Example 5

[0121] The redox salt solution used in the constant current testing of a system in accordance with the disclosure was a mixture of 0.25M potassium ferricyanide and 0.25M sodiumDocket No.30275 / 59445 ferrocyanide in deionized water. The ocean water was a solution of 35.95 grams of Instant Ocean aquarium salt mixture per 1 liter of deionized water.

[0122] Constant current tests were performed on the two-electrode cell with a Biologic potentiostat while connecting the reference electrode probe directly to the counter electrode probe. The length of the tests and the constant current value the tests were performed at were dependent on the flow rate of the ocean water and combined ferri / ferrocyanide solution. For acidification, the reacted redox salt solution was collected and set aside to be reused for basification, while the reacted ocean water flowed into a vacuum chamber without being exposed to air in order to prevent CO2 loss.

[0123] The flow cell consisted of two plastic outer plates, two current collectors, two 1” squares of titanium mesh, one of which was platinized, a 1” square gas diffusion electrode of platinized carbon paper, a 1” square counter electrode of platinum foil, two flow plates, one of which was ridged, a 1.2” square of Nafion NR-211, and four gaskets, arranged as seen in Fig 19. For acidification, the platinized carbon paper was used as the working electrode, while the platinum foil was used as the counter electrode, and vice versa for basification. The platinum foil electrode was electropolished in 0.1M hydrochloric acid prior to use. For acidification, hydrogen was supplied to the cell throughout the test at a low flow rate. For basification, nitrogen gas was used to flood the cell to prevent secondary reactions with atmospheric oxygen.

[0124] For acidification, prior to the constant current test, the flow cell underwent an open- circuit voltage (OCV) test in order to calibrate the flow of hydrogen to the cell and prevent hydrogen mass transport limitation during the constant current test. The hydrogen flow was approved once the OCV showed a value of < -0.7V. Additionally, the impedance across the flow cell was measured by ZIR prior to the constant current test in order to prevent large hydrogen bubbles forming within the cell., with an acceptable impedance being < 4.5Ω.

[0125] The effect of reactor design on performance was evaluated. Without intending to be bound by theory, it was believed that a thinner flow plate should linearly reduce the ohmic resistance, and a thin or finned flow plate or higher flow rates would decrease the liquid boundary layer and therefore reduce the mass transport limitation, resulting in higher performance.

[0126] Referring to Figure 20, it was observed that for acidification, the mass transport was less important, but reduced ohmic resistance of thin plate significantly helped performance. For basification, there was an improvement from a thin designed due to reduced ohmic resistance,Docket No.30275 / 59445 However mass transport is more important here. The ridged plate was observed to increase the limiting current density by about 50% compared to the standard plate and the thin plate were somewhere between the two due to the larger ohmic resistance numbers at the same flow rate. There were no significant competing reactions until high voltages.

[0127] Figure 21 illustrates a comparison of operating current density for both acidification and basification, vs electrochemical energy intensity. The points from the first acidification- basification cycle that were taken are shown in the stars, where the error bars come from the gas mass flow and concentration measurement error at a 95% confidence interval. Literature comparisons are shown in the other shaped markers. The electrochemical equivalent circuit model at comparable resistance is shown at the dotted line, and the thermodynamic minimum based on the Nernstian voltage pH shift is shown in the dashed line.

[0128] At lower currents, the system in accordance with the disclosure was observed to be comparable with the literature. At increasing current density there was a linear increase in the electrochemical energy intensity, observed, due to the ohmic resistance. This followed the trend from the modeling with data being shifted up. This shift could be attributable to a number of factors, but the overall correlation to the trend of the model was observed.

[0129] Referring to Figure 22, the change in fouling over 4 total cycles is shown.1 total cycle was an acidification followed by a basification. Therefore, at the stars at half cycles there was a clear electrode and at the full cycles significant precipitation across the electrode was observed. The precipitation especially sticks where the fluid advection was reduced in the edges of the flow channel and the turns of the serpentine path. The fouling coverage increased from cycles 1 to 3, but the clean coverage remained about the constant.

[0130] Electrochemical energy intensity data is also shown in Figure 22. These data were obtained at a fixed current density of 17 mA / cm2and the energy was on the higher range near 6-7 GJ / tCO2 as a result. The energy intensity variations with cycling was within measurement error.

[0131] The foregoing description is given for clearness of understanding only, and no unnecessary limitations should be understood therefrom, as modifications within the scope of the disclosure may be apparent to those having ordinary skill in the art.

[0132] All patents, patent applications, government publications, government regulations, and literature references cited in this specification are hereby incorporated herein by reference in their entirety. In the case of conflict, the present description, including definitions, will control.Docket No.30275 / 59445

[0133] Throughout the specification, where the compounds, compositions, methods, and / or processes are described as including components, steps, or materials, it is contemplated that the compounds, compositions, methods, and / or processes can also comprise, consist essentially of, or consist of any combination of the recited components or materials, unless described otherwise. Component concentrations can be expressed in terms of weight concentrations, unless specifically indicated otherwise. Combinations of components are contemplated to include homogeneous and / or heterogeneous mixtures, as would be understood by a person of ordinary skill in the art in view of the foregoing disclosure. References (1) Smith, P. et al. Biophysical and economic limits to negative CO2 emissions. Nature Climate Change 2016, 6, 42–50. (2) Fuss, S.; Jones, C. D.; Kraxner, F.; Peters, G. P.; Smith, P.; Tavoni, M.; van Vu- uren, D. P.; Canadell, J. G.; Jackson, R. B.; Milne, J.; Moreira, J. R.; Nakicenovic, N.; Sharifi, A.; Yamagata, Y. Research priorities for negative emissions. Environmental Research Letters 2016, 11, 115007. (3) Keith, D. W. Why Capture CO2 from the Atmosphere? Science 2009, 325, 1654– 1655. (4) Davila, X.; Gebbie, G.; Brakstad, A.; Lauvset, S. K.; McDonagh, E. L.; Schwinger, J.; Olsen, A. How Is the Ocean Anthropogenic Carbon Reservoir Filled? Global Biogeo-chemical Cycles 2022, 36. (5) Caldeira, K.; Akai, M. In IPCC Special Report on Carbon dioxide Capture and Storage; Metz, B., Davidson, O., de Coninck, H., Loos, M., Meyer, L., Eds.; Cambridge University Press, 2005; Chapter 6 Ocean Storage. (6) Archer, D.; Kheshgi, H.; Maier-Reimer, E. Multiple timescales for neutralization of fossil fuel CO 2. Geophysical Research Letters 1997, 24, 405–408. (7) Kheshgi, H. S.; Smith, S. J.; Edmonds, J. A. Emissions and Atmospheric CO2 Stabilization: Long-Term Limits and Paths. Mitigation and Adaptation Strategies for Global Change 2005, 10, 213–220. (8) Sharifian, R.; Wagterveld, R. M.; Digdaya, I. A.; Xiang, C.; Vermaas, D. A. Electro- chemical carbon dioxide capture to close the carbon cycle. Energy Environmental Science 2021, 14, 781–814. (9) Renfrew, S. E.; Starr, D. E.; Strasser, P. Electrochemical Approaches toward CO2 Capture and Concentration. ACS Catalysis 2020, 10, 13058–13074. (10) Eisaman, M. D.; Parajuly, K.; Tuganov, A.; Eldershaw, C.; Chang, N.; Littau, K. A. CO 2 extraction from seawater using bipolar membrane electrodialysis. Energy En- vironmental Science 2012, 5, 7346–7352. (11) Willauer, H. D.; DiMascio, F.; Hardy, D. R.; Lewis, M. K.; Williams, F. W. Develop- ment of an Electrochemical Acidification Cell for the Recovery of CO2 and H2 from Seawater. Industrial Engineering Chemistry Research 2011, 50, 9876–9882. (12) Willauer, H. D.; DiMascio, F.; Hardy, D. R.; Lewis, M. K.; Williams, F. W. Develop- ment of an Electrochemical Acidification Cell for the Recovery of CO2 and H2 from Seawater II. Evaluation of the Cell by Natural Seawater. Industrial Engineering Chemistry Research 2012, 51, 11254–11260.Docket No.30275 / 59445 (13) de Lannoy, C.-F.; Eisaman, M. D.; Jose, A.; Karnitz, S. D.; DeVaul, R. W.; Hannun, K.; Rivest, J. L. Indirect ocean capture of atmospheric CO2: Part I. Prototype of a negative emissions technology. International Journal of Greenhouse Gas Control 2018, 70, 243– 253. (14) Digdaya, I. A.; Sullivan, I.; Lin, M.; Han, L.; Cheng, W.-H.; Atwater, H. A.; Xiang, C. A direct coupled electrochemical system for capture and conversion of CO2 from ocean-water. Nature Communications 2020, 11, 4412. (15) Willauer, H. D.; DiMascio, F.; Hardy, D. R.; Williams, F. W. Development of an Electrolytic Cation Exchange Module for the Simultaneous Extraction of Carbon Dioxide and Hydrogen Gas from Natural Seawater. Energy & Fuels 2017, 31, 1723–1730. (16) Willauer, H. D.; DiMascio, F.; Hardy, D. R.; Williams, F. W. Feasibility of CO2 Extraction from Seawater and Simultaneous Hydrogen Gas Generation Using a Novel and Robust Electrolytic Cation Exchange Module Based on Continuous Electrodeionization Technology. Industrial & Engineering Chemistry Research 2014, 53, 12192–12200. (17) Eisaman, M. D.; Alvarado, L.; Larner, D.; Wang, P.; Garg, B.; Littau, K. A. CO2 separation using bipolar membrane electrodialysis. Energy Environ. Sci.2011, 4, 1319– 1328. (18) Park, H.; Lee, J.; Han, J.; Park, S.; Park, J.; Min, B. CO2 Fixation by Membrane Separated NaCl Electrolysis. Energies 2015, 8, 8704–8715. (19) Prajapati, A.; Singh, M. R. Preventing Over-Electrodialysis for Efficient CO2 Capture from Seawater. ACS Sustainable Chemistry & Engineering 2022, 10, 12466–12474. (20) Sharifian, R.; Boer, L.; Wagterveld, R.; Vermaas, D. Oceanic carbon capture through electrochemically induced in situ carbonate mineralization using bipolar membrane. Chemical Engineering Journal 2022, 438, 135326. (21) Yan, L.; Bao, J.; Shao, Y.; Wang, W. An Electrochemical Hydrogen-Looping System for Low- Cost CO2 Capture from Seawater. ACS Energy Letters 2022, 7, 1947–1952. (22) Kim, S.; Nitzsche, M. P.; Rufer, S. B.; Lake, J. R.; Varanasi, K. K.; Hatton, T. A. Asymmetric chloride-mediated electrochemical process for CO2 removal from oceanwa-ter. Energy & Environmental Science 2023, (23) Singh, U.; Colosi, L. M. Capture or curtail: The potential and performance of direct air capture powered through excess renewable electricity. Energy Conversion and Management: X 2022, 15. (24) Zhou, E.; Mai, T. Electrification Futures Study: Operational Analysis of U.S. Power Systems with Increased Electrification and Demand-Side Flexibility. 2021; https: / / www.nrel.gov / docs / fy21osti / 79094.pdf. (25) Frey, D.; Kim, J.; Dvorkin, Y.; Modestino, M. A. Spatiotemporal Decoupling of Water Electrolysis for Dual-Use Grid Energy Storage and Hydrogen Generation. Cell Reports Physical Science 2020, 1, 100226. (26) Ruggles, T. H.; Dowling, J. A.; Lewis, N. S.; Caldeira, K. Opportunities for flexible electricity loads such as hydrogen production from curtailed generation. Advances in Applied Energy 2021, 3, 100051. (27) Ginsberg, M. J.; Venkatraman, M.; Esposito, D. V.; Fthenakis, V. M. Minimizing the cost of hydrogen production through dynamic polymer electrolyte membrane elec-trolyzer operation. Cell Reports Physical Science 2022, 3, 100935. (28) Van´ysek, P. In Electrochemical Series, 103rd ed.; Rumble, J., Ed.; CRC Press / Taylor & Francis, 2022. (29) Dermentzis, K. Continuous electrodeionization through electrostatic shielding. Elec- trochimica Acta 2008, 53, 2953–2962. (30) Viebahn, P.; Scholz, A.; Zelt, O. The Potential Role of Direct Air Capture in the German Energy Research Program—Results of a Multi-Dimensional Analysis. Energies 2019, 12, 3443.Docket No.30275 / 59445 (31) Eisaman, M. D.; Rivest, J. L.; Karnitz, S. D.; de Lannoy, C.-F.; Jose, A.; DeVaul, R. W.; Hannun, K. Indirect ocean capture of atmospheric CO2: Part II. Understanding the cost of negative emissions. International Journal of Greenhouse Gas Control 2018, 70, 254– 261. (32) Plante, E. C. L.; Simonetti, D. A.; Wang, J.; Al-Turki, A.; Chen, X.; Jassby, D.; Sant, G. N. Saline Water-Based Mineralization Pathway for Gigatonne-Scale CO2 Management. ACS Sustainable Chemistry & Engineering 2021, 9, 1073–1089. (33) California ISO Oasis Wind and Solar Forecast. 2023; http: / / oasis.caiso.com / mrioasis / logon.do. (34) IEA, Direct Air Capture 2022. 2022; https: / / www.iea.org / reports / direct-air-capture-2022. (35) Beck, V. A.; Ivanovskaya, A. N.; Chandrasekaran, S.; Forien, J.- E. B.; Worsley, M. A. Inertially enhanced mass transport using 3D-electrodes with periodic lattice structures. Proceedings of the National Academy of Sciences 2021, 118. (36) Vedharathinam, V.; Qi, Z.; Horwood, C.; Bourcier, B.; Stadermann, M.; Biener, J.; Biener, M. Using a 3D Porous Flow-Through Electrode Geometry for High-Rate Electrochemical Reduction of CO ¡sub,2¡ / sub, to CO in Ionic Liquid. ACS Catalysis 2019, 9, 10605– 10611. (37) Barrera, L.; Chandran, R. B. Harnessing Photoelectrochemistry for Wastewater Nitrate Treatment Coupled with Resource Recovery. ACS Sustainable Chemistry Engineering 2021, 9, 3688–3701. (38) Zhu, S.; Kingsbury, R. S.; Call, D. F.; Coronell, O. Impact of solution composition on the resistance of ion exchange membranes. Journal of Membrane Science 2018, 554, 39– 47. (39) Geise, G. M.; Curtis, A. J.; Hatzell, M. C.; Hickner, M. A.; Logan, B. E. Salt Concentration Differences Alter Membrane Resistance in Reverse Electrodialysis Stacks. Environmental Science Technology Letters 2014, 1, 36–39. (40) Gómez-Coma, L.; Ortiz-Martı´nez, V. M.; Carmona, J.; Palacio, L.; Prádanos, P.; Fal-lanza, M.; Ortiz, A.; Ibañez, R.; Ortiz, I. Modeling the influence of divalent ions on membrane resistance and electric power in reverse electrodialysis. Journal of Membrane Science 2019, 592, 117385. (41) Eliaz, N.; Gileadi, E. Physical Electrochemistry, 2nd ed.; Wiley-VCH, 2019; pp 31–33. (42) Wang, X.; Gautam, R. K.; Jiang, J. Strategies for Improving Solubility of Redox-Active Organic Species in Aqueous Redox Flow Batteries: A Review. Batteries Supercaps 2022, 5. (43) Weng, L.-C.; Bell, A. T.; Weber, A. Z. Modeling gas-diffusion electrodes for CO2 reduction. Physical Chemistry Chemical Physics 2018, 20, 16973–16984. (44) Burdyny, T.; Smith, W. A. CO2 reduction on gas-diffusion electrodes and why catalytic performance must be assessed at commercially-relevant conditions. Energy Environ- mental Science 2019, 12, 1442–1453. (45) Clark, E. L.; Resasco, J.; Landers, A.; Lin, J.; Chung, L.-T.; Walton, A.; Hahn, C.; Jaramillo, T. F.; Bell, A. T. Standards and Protocols for Data Acquisition and Reporting for Studies of the Electrochemical Reduction of Carbon Dioxide. ACS Catalysis 2018, 8, 6560–6570. (46) Gupta, N.; Gattrell, M.; MacDougall, B. Calculation for the cathode surface concentrations in the electrochemical reduction of CO2 in KHCO3 solutions. Journal of Applied Electrochemistry 2006, 36, 161–172. (47) Bandwidth Study on Energy Use and Potential Energy Savings Opportunities in U.S. Seawater Desalination Systems.2017,Docket No.30275 / 59445 (48) Rao, P.; Aghajanzadeh, A.; Sheaffer, P.; Morrow, W. R.; Brueske, S.; Dollinger, C.; Price, K.; Sarker, P.; Ward, N.; Cresko, J. Volume 1: Survey of Available Information in Support of the Energy-Water Bandwidth Study of Desalination Systems.2016. (49) CAISO (California ISO) Day-Ahead Price. http: / / www.energyonline.com / Data / GenericData.aspx?DataId=22&CAISO___Day-Ahead_Price. (50) Dickson, A. G.; Goyet, C. Handbook of methods for the analysis of the various parameters of the carbon dioxide system in sea water. Version 2.1994, (51) Sharifian, R.; Boer, L.; Wagterveld, R.; Vermaas, D. Oceanic carbon capture through electrochemically induced in situ carbonate mineralization using bipolar membrane. Chemical Engineering Journal 2022, 438, 135326. (52) Digdaya, I. A.; Sullivan, I.; Lin, M.; Han, L.; Cheng, W.-H.; Atwater, H. A.; Xiang, C. A direct coupled electrochemical system for capture and conversion of CO2 from ocean-water. Nature Communications 2020, 11, 4412. (53) Willauer, H. D.; DiMascio, F.; Hardy, D. R.; Lewis, M. K.; Williams, F. W. Development of an Electrochemical Acidification Cell for the Recovery of CO2 and H2 from Seawater II. Evaluation of the Cell by Natural Seawater. Industrial & Engineering Chemistry Research 2012, 51, 11254–11260. (54) Xu, Y.; Edwards, J. P.; Liu, S.; Miao, R. K.; Huang, J. E.; Gabardo, C. M.; O’Brien, C. P.; Li, J.; Sargent, E. H.; Sinton, D. Self-Cleaning CO2Reduction Systems: Unsteady Electrochemical Forcing Enables Stability. ACS Energy Letters 2021, 6, 809–815. (55) Plante, E. C. L.; Simonetti, D. A.; Wang, J.; Al-Turki, A.; Chen, X.; Jassby, D.; Sant, G. N. Saline Water-Based Mineralization Pathway for Gigatonne-Scale CO2 Management. ACS Sustainable Chemistry & Engineering 2021, 9, 1073–1089. (56) Park, H.; Lee, J.; Han, J.; Park, S.; Park, J.; Min, B. CO2 Fixation by Membrane Separated NaCl Electrolysis. Energies 2015, 8, 8704–8715. (57) Rau, G. H. Electrochemical CO2 capture and storage with hydrogen generation. Energy Procedia 2009, 1, 823–828. (58) Sano, Y.; Hao, Y.; Kuwahara, F. Development of an electrolysis based system to con- tinuously recover magnesium from seawater. Heliyon 2018, 4, e00923. (59) Dermentzis, K. Continuous electrodeionization through electrostatic shielding. Elec- trochimica Acta 2008, 53, 2953–2962. (60) Daum, P. H.; Enke, C. G. Electrochemical kinetics of the ferri-ferrocyanide couple on platinum. Analytical Chemistry 1969, 41, 653–656. (61) Zheng, J.; Sheng, W.; Zhuang, Z.; Xu, B.; Yan, Y. Universal dependence of hydrogen oxidation and evolution reaction activity of platinum-group metals on pH and hydrogen binding energy. Science Advances 2016, 2. (62) Van´ysek, P. In Electrochemical Series, 103rd ed.; Rumble, J., Ed.; CRC Press / Taylor & Francis, 2022. (63) Rumble, J., Ed. CRC Handbook of Chemistry and Physics, 103rd ed.; CRC Press / Taylor & Francis, 2022. (64) Wang, X.; Gautam, R. K.; Jiang, J. Strategies for Improving Solubility of Redox-Active Organic Species in Aqueous Redox Flow Batteries: A Review. Batteries & Supercaps 2022, 5. (65) Sheng, W.; Zhuang, Z.; Gao, M.; Zheng, J.; Chen, J. G.; Yan, Y. Correlating hydrogen oxidation and evolution activity on platinum at different pH with measured hydrogen binding energy. Nature Communications 2015, 6, 5848.Docket No.30275 / 59445 (66) Barrera, L.; Chandran, R. B. Harnessing Photoelectrochemistry for Wastewater Nitrate Treatment Coupled with Resource Recovery. ACS Sustainable Chemistry & Engineering 2021, 9, 3688–3701. (67) Canfield, D. E.; Farquhar, J. Animal evolution, bioturbation, and the sulfate concentration of the oceans. Proceedings of the National Academy of Sciences 2009, 106, 8123–8127. (68) Zhang, J.-Z.; Millero, F. J. The rate of sulfite oxidation in seawater. Geochimica et Cosmochimica Acta 1991, 55, 677–685. (69) IUPAC-NIST Solubility Database. NIST Standard Reference Database 106 (70) Cengel, Y.; Boles, M. Thermodynamics: An Engineering Approach, 9th ed.; McGraw-Hill US Higher Ed USE, 2018. (71) Matsumiya, N.; Teramoto, M.; Kitada, S.; Matsuyama, H. Evaluation of energy con- sumption for separation of CO2 in flue gas by hollow fiber facilitated transport membrane module with permeation of amine solution. Separation and Purification Technology 2005, 46, 26–32. (72) Crone, B. C.; Garland, J. L.; Sorial, G. A.; Vane, L. M. Significance of dissolved methane in effluents of anaerobically treated low strength wastewater and potential for recovery as an energy product: A review. Water Research 2016, 104, 520–531. (73) Budama, V. K.; Brendelberger, S.; Roeb, M.; Sattler, C. Performance Analysis and Optimization of Solar Thermochemical Water-Splitting Cycle with Single and Multiple Receivers. Energy Technology 2022, 10, 2100220. (74) Sverdrup, H. U.; Johnson, M. W.; Fleming, R. H. The Oceans: Their Physics, Chemistry, and General Biology. Journal of Chemical Education 1943, 20, 517. (75) de Visser, E.; Hendriks, C.; Barrio, M.; Mølnvik, M. J.; de Koeijer, G.; Liljemark, S.; Gallo, Y. L. Dynamis CO2 quality recommendations. International Journal of Greenhouse Gas Control 2008, 2, 478–484. (76) Rao, P.; Aghajanzadeh, A.; Sheaffer, P.; Morrow, W. R.; Brueske, S.; Dollinger, C.; Price, K.; Sarker, P.; Ward, N.; Cresko, J. Volume 1: Survey of Available Information in Support of the Energy-Water Bandwidth Study of Desalination Systems.2016. (77) Eisaman, M. D.; Rivest, J. L.; Karnitz, S. D.; de Lannoy, C.-F.; Jose, A.; DeVaul, R. W.; Hannun, K. Indirect ocean capture of atmospheric CO2: Part II. Understanding the cost of negative emissions. International Journal of Greenhouse Gas Control 2018, 70, 254–261. (78) California ISO Oasis Wind and Solar Forecast. 2023; http: / / oasis.caiso.com / mrioasis / logon.do.

Claims

Docket No.30275 / 59445 What is claimed is:

1. A process of capturing CO2 from a CO2-containing water source in an electrochemical flow cell comprising a flow channel defined between an anode and an oppositely disposed cathode, and a membrane arranged in the flow channel to define a first side of the flow channel between the anode and the membrane and a second side of the flow channel between the cathode and the membrane, the process comprising: performing an acidification step comprising: flowing, in a first direction, the water source and H2 gas between through the first side of the flow channel, and flowing, in the first direction, an electrolyte comprising a reversible redox salt through the second side of the flow channel, wherein upon application of a first potential to the electrochemical cell, a H2 oxidation reaction occurs at the anode with the water source being an anolyte for the H2 oxidation reaction and a reduction of the reversible redox salt occurs at the cathode, thereby acidifying the water source and causing a shift in speciation within the water source from HCO3- to dissolved CO2thereby allowing the dissolved CO2to phase separate from the water source; removing the phase separated CO2from the electrochemical cell; performing a basification step comprising: reversing a polarization of the anode and the cathode thereby providing a reversed electrochemical cell in which the anode becomes a cathode of the reversed cell and the cathode becomes an anode of the reversed cell, the first side being defined between the cathode of the reversed cell and the membrane and the second side being defined between the anode of the reversed cell and the membrane, flowing, in a second direction, the acidified water having CO2removed therefrom through the first side, and flowing, in the second direction, the electrolyte comprising the reduced redox salt through the second side, wherein upon application of a second potential to the electrochemical cell, an H2evolution reaction occurs at the cathode of the reverse cell and the reduced redox salt isDocket No.30275 / 59445 oxidized at the anode of the reverse cell, thereby basifying the acidified water and generating H2, wherein the oxidized redox salt and generated H2are used in a subsequent acidification step, wherein the first direction is opposite the second direction.

2. A process of capturing CO2from a CO2-containing water source, comprising: performing an acidification step in a first electrochemical flow cell comprising a flow channel defined between an anode and an oppositely disposed cathode, and a membrane arranged in the flow channel to define a first side of the flow channel between the anode and the membrane and a second side of the flow channel between the cathode and the membrane, the acidification step comprising flowing the water source and H2 gas between through the first side of the flow channel of the first electrochemical cell, and flowing an electrolyte comprising a reversible redox salt through the second side of the flow channel of the first electrochemical cell, wherein upon application of a first potential to the first electrochemical cell, a H2 oxidation reaction occurs at the anode with the water source being an anolyte for the H2 oxidation reaction and a reduction of the reversible redox salt occurs at the cathode, thereby acidifying the water source and causing a shift in speciation within the water source from HCO3- to dissolved CO2 thereby allowing the dissolved CO2 to phase separate from the water source; removing the phase separated CO2 from the first electrochemical cell; performing a basification step in a second electrochemical cell comprising a flow channel defined between an anode and an oppositely disposed cathode, and a membrane arranged in the flow channel to define a first side of the flow channel between the cathode and the membrane and a second of the flow channel between the anode and the membrane, the basification step comprising: flowing the acidified water having CO2removed therefrom through the first side of the flow channel of the second electrochemical cell, and flowing the electrolyte comprising the reduced redox salt through the second side of the flow channel of the second electrochemical cell,Docket No.30275 / 59445 wherein upon application of a second potential to the second electrochemical cell, an H2evolution reaction occurs at the cathode of the second electrochemical cell and the reduced redox salt is oxidized at the anode of the second electrochemical cell, thereby basifying the acidified water and generating H2, wherein the oxidized redox salt and generated H2used in a subsequent acidification step, wherein the first electrochemical cell and the second electrochemical cell are fluidly coupled such that the acidified water having CO2 removed and the electrolyte comprising the reduced redox salt from the first electrochemical cell are flowed into the second electrochemical cell and the H2 and electrolyte comprising the oxidized redox salt generated in the second electrochemical cell are flowed into the first electrochemical cell.

3. The process of claim 1 or 2, wherein the reversable redox salt comprises anions from salts selected from ferro / ferri-cyanide, iodide, bromide, or chloride salts and cations selected from sodium, lithium, or potassium.

4. The process of claim 3, wherein the reversible redox salt is Na4Fe(CN)6 / Na3Fe(CN)6 or K4Fe(CN)6 / K3Fe(CN)6.

5. The process of any one of the preceding claims, wherein the water source has a pH of about 7.5 to 8.5, the acidified water source has a pH of about 4 to about 6, and the basified water source has a pH of about 9 to about 11.

6. The process of any one of the preceding claims, further comprising storing the CO2 depleted acidified water and the electrolyte comprising the reduced redox salt and repeating the acidification step prior to performing the basification step, wherein the basification step is performed using the stored CO2 depleted acidified water and the electrolyte comprising the reduced redox salt.

7. The process of any one of the preceding claims, comprising performing repeated cycles, each cycle comprising performing the acidification step, removing the phase-separated CO2, and performing the basification step.

8. The process of any one of claims 1 to 7, comprising performing a first cycle comprising performing the acidification step and removing the phase-separated CO2for a first time and then performing a second cycle comprising performing the basification step for a second time.

9. The process of claim 8, wherein the first time is different from the second time.Docket No.30275 / 59445 10. The process of claim 8 or 9, wherein the first cycle is performed during a period of high power usage on a power source to which the electrochemical cell is connected and the second cycle is performed during a period of low power usage on the power source.

11. The process of any one of the preceding claims, wherein the membrane is an ion exchange membrane.

12. The process of any one of the preceding claims, wherein the anode comprises a base material selected from carbon or titanium and a catalyst for suitable hydrogen reaction selected from Pt, Ir, Ru, Rh, Au, Ag, Ni, Cu, Ti, Mn, Fe, and Cr.

13. The process of any one of the preceding claims, wherein the cathode comprises a base material selected from carbon or titanium and a catalyst for suitable hydrogen reaction selected from Pt, Ir, Ru, Rh, Au, Ag, Ni, Cu, Ti, Mn, Fe, and Cr.

14. The process of claim 2, comprising performing the acidification step and the basification step substantially simultaneously.

15. The process of claim 1, comprising performing the acidification and basification steps in a plurality of electrochemical cells, each electrochemical cell performing both acidification and basification steps.

16. The process of claim 2, comprising performing the acidification and basification steps in a plurality of electrochemical cells, each electrochemical cell performing only one of the acidification step or the basification step.

17. The process of claim 1, comprising performing a further acidification step after the basification step, wherein the H2 generated during the basification step is flowed into the first side of the flow channel with a water source and the electrolyte comprising the oxidized redox salt generated during the basification step is flowed into the second side of the flow channel, wherein precipitates accumulated on the anode and / or cathode during the basification step are removed during the acidification step performed after the basification step.

Images