Capacitive Processes for Carbon Capture and Energy Recovery from CO2 Emissions: Shaping a New Technology Going from Water to Gas Applications

Abstract

Introduction: CO₂-CDI and CO₂-CAPMIX

This Ph.D. thesis explores the use of capacitive processes for two novel applications, i.e., CO₂ capture (CO₂-CDI) and energy recovery from CO₂ emissions (CO₂- CAPMIX). Conventionally, capacitive processes have been applied for water technologies, either to desalinate water streams (Capacitive deionization, CDI) or to harvest electrical energy from salinity gradients (capacitive mixing, CAPMIX). Both CDI and CAPMIX are electrochemical technologies, which operated based on the presence of ions in water. CDI cells adsorb ions from an aqueous solution by alternatively charging and discharging two porous electrodes (electrosorption). CAPMIX cells harvest electrical energy based on the establishment of a membrane potential due to ion electrosorption in porous electrodes.

The major goal of this Ph.D. thesis is to shape CDI and CAPMIX processes from water-based toward CO₂ gas-based technologies. Previous works lead the way toward this transition by demonstrating a proof of concept of CO₂-CAPMIX, and the possibility to form ions from CO₂ gas through reaction with deionized water. Nevertheless, the electrical power generated was strongly limited by the high internal resistance of the CO₂-CAPMIX cell, caused by the low ionic conductivity of CO₂-sparged deionized water. From that starting point, this Ph.D. thesis aims at developing CO₂-CAPMIX further and demonstrating the concept of CO₂-CDI through three different steps, i.e., (i) proof of CO₂ reactive electrosorption and direct gas feeding concepts, (ii) in-depth understanding and (iii) exploration of cell designs to reduce the internal resistance.

Proof-of-principles of CO₂ reactive electrosorption and direct gas feeding (Chapter 2 and 3)

In Chapter 2, we demonstrate for the first time that CO₂ gas can be adsorbed in a membrane capacitive deionization (MCDI) cell through CO₂ reactive electrosorption. The term “reactive” refers to the hydration reaction of CO₂ gas necessary to form bicarbonate ions, and the term “electrosorption” refers to the adsorption of HCO₃ - ions in the porous electrodes. Besides, we characterize the different mechanisms of CO₂ reactive electrosorption through a calculation framework in which we relate the charge efficiency (Λ), conventionally used in CDI, to a new figure of merit, the absorption efficiency (Λa). While the charge efficiency is the amount of ions adsorbed to the electrical charge, the absorption efficiency is the amount of CO₂ gas absorbed per electrical charge. Our results show a promising value of ΛD≈ 0.7 (corresponding to Λ≈ 0.8), which is close to values expected from conventional MCDI. Furthermore, we report energy consumption values as low as 40-50 kJ/molSTU (at 0.2-0.6 A m^-2), which is comparable to other electrochemical CO2 capture systems. Nevertheless, the CO2 absorption rate was rather low due to the low current density applied, which was limited by the internal resistance. Two limiting factors on the energy consumption were then identified, i.e., (i) the high ohmic internal resistance caused by the low ionic conductivity of the CO₂-sparged solution and (ii) the loss of absorption efficiency observed at long charging times.

In Chapter 3, we demonstrate that the CO₂ gas can be directly fed into a capacitive cell operated as CO₂-CAPMIX. We refer to this feeding mode as “direct gas feeding” as opposed to “solution feeding,” where CO₂ gas is first sparged into deionized water outside the capacitive cell. By feeding the gas directly into the cell, the reactive and electrosorption process directly co-occurs in the capacitive cell, thus avoiding the energyintensive CO₂-sparging step. The direct gas feeding mode was tested with three different cell designs, i.e., flat MCDI, wire-shaped MCDI, and a membrane electrode assembly (MEA) cell designs. Among the cell design tested, the MEA design, inspired from PEM fuel cells, shows the lowest internal resistance and the higher apparent membrane permselectivity (≈0.9), resulting in power output 100 times higher than conventional MCDI cells (0.05 against 4 mW m-2). Overall, we conclude that an MEA is an interesting alternative to MCDI cell design for capacitive cell design operated for the energy recovery from CO₂ emissions. Moreover, to further improve the power density, other strategies to decrease the internal resistance should still be investigated.

Understanding the role of membranes and electrodes during CO₂ reactive electrosorption (Chapter 4)

To understand the loss of absorption efficiency (ΛD ) with time, we investigate in Chapter 4 the individual role of the electrodes and ion exchange membranes in CO₂- (M)CDI. The role of membranes was characterized by testing different CDI configurations with and without membranes. Moreover, the role of the electrodes was theoretically investigated by adopting the amphoteric Donnan (amph-D) model for CO₂-CDI. Our result shows that the presence of membranes is essential to keep high values of ΛD , which has been found two times higher in flat CO₂-MCDI compared to flat CO₂-CDI cell. Besides, between both membranes, we show that the presence of the anion exchange membrane largely contributes to a larger extent to increase ΛD compared to the cation exchange membrane.

Concerning the behavior of the electrodes, a discrepancy between data and theoretical results was observed, suggesting that other physicochemical mechanisms not included in the amph-D could influence the CO₂ reactive electrosorption process. By including the effect of ionization of surface chemical charge, a better fit was obtained between the model and the data, thus showing the importance of the surface chemica charge on the CO₂-CDI performance. We suggest investigating other properties, such as ion-size selectivity, ion-valence selectivity, as well as the field dissociation effect. Investigating such properties is also of interest for conventional CDI in mixture solutions or acidic solutions.

Cell designs exploration to reduce the internal resistance (Chapter 3 5 and 6)

Reducing the internal resistance is a crucial challenge to tackle for developing CO₂- CDI and CO₂-CAPMIX. Therefore, Chapters 3,5 and 6 focus on exploring different cell designs to reduce the internal resistance. In this regard, the internal resistance was separated into two contributions, i.e., ohmic (time-independent) and non-ohmic (timedependent) resistances.

In Chapter 3 (CO₂-CAPMIX) and Chapter 5 (CO₂-CDI), the internal resistance of a flat MCDI and MEA cell designs are compared. Our main results show that the MEA cell design improves the internal resistance, to a certain extent. In Chapter 5, we demonstrate that the MEA cell design significantly reduces the ohmic contribution of the internal resistance (more than 100 times). Moreover, in chapters 5 and 6, we show that the internal resistance of an MEA cell further decreases (more than five times) by coating the anode and cathode with the same ionomer (more than five times). The reduced internal resistance by coating the electrodes was attributed to a decrease of non-ohmic resistance. Compared to a flat MCDI cell design, power densities 100 times higher were obtained with an MEA cell with coated electrodes, and energy consumption decreased by more than 2.5 times in CO2-CDI. Nevertheless, the energy performance was lower than expected as the reduced resistance occurs at the cost of lower ion selectivities in the MEA cell design. We recommend investigating the ion selectivity mechanisms in MEA cell design to find suitable strategies to improve the ion selectivity.

Discussion on CO₂-CDI and CO₂-CAPMIX (Chapter 7)

In Chapter 7, all the main results from previous chapters are summarized and further discussed. The discussion is mostly focused on internal resistance and ion selectivity (i.e., absorption efficiency in CO₂-CDI and membrane potential in CO₂-CAPMIX), and possible strategies to improve them. Furthermore, we share and discuss our longterm perspective for both technologies.