Recovery of carbon losses during CO2 electrolysis using solid electrolyte reactors
As renewable electricity emerges as a viable alternative to fossil fuels, the electrochemical conversion of carbon dioxide into an essential chemical feedstock has been seen by many researchers as a promising way to store and use this renewable electricity while mitigating climate change. Over the past few decades, CO2 reduction reactions (CO2RR) have made remarkable progress in terms of efficiency and selectivity. Among these efforts, catalytic material design and reactor engineering have been the two main efforts in this field. The development of CO2 electrolysis, particularly the use of gas diffusion layer (GDL) electrodes in flow cell and membrane electrode assembly (MEA) cell reactors, has played a central role in driving catalytic performance toward industrially relevant targets. The practical application of electrochemical CO2 reduction technology is greatly challenged by the fact that in conventional CO2 electrolyzers, CO2 is limited by the significant shuttle effect toward the anode side through interfacial carbonate formation and mixing with O2.
Based on this, a porous solid electrolyte reactor strategy was recently reported by Assistant Professor Hao-Tian Wang at Rice University in Houston, Texas, USA, to effectively recover carbon losses from electrochemical CO2 reduction processes. By creating a sulfonated polymer electrolyte with permeability and ionic conductivity as a buffer layer between the cathode and anode, the shuttle carbonate can combine with the protons generated from the anode to re-form CO2 gas for reuse without mixing with the anode O2. The reduction of CO2 to CO using silver nanowire catalysts resulted in the demonstration of ultra-high gas purity (>99%) shuttle CO2 recovery of up to 90% while providing over 90% CO Faraday (FE) efficiency at 200 mA cm-2 current. The researchers were able to obtain continuous CO2 conversion efficiencies in excess of 90% by recovering the recovered CO2 back into the CO2 input stream. A related paper was published as "Recovering carbonlosses in CO2 electrolysis using a solid electrolyte reactor" in Nat. Catal.
As shown in Figure 1, during CO2RR electrolysis, especially at high current densities, the cathode electrolyte interface produces large amounts of hydroxide ions (OH-), which react rapidly with the CO2 stream to form carbonate or bicarbonate ions. These carbonate ions, due to the electric field, then migrate to the cathode, anode interface (aqueous solution or anion exchange membrane) against the anode reactor and combine with the locally generated protons (hydrogen ions) for the oxygen evolution reaction (OER) to produce CO2 gas again. Unfortunately, these shuttle CO2 gas molecules cannot be used directly in CO2RR because they mix with the anode O2, resulting in significant carbon losses and reducing the overall energy efficiency of the CO2RR technology. When strong alkaline solutions are used, the carbon loss problem is even more severe due to the continuous chemical reaction between the CO2 stream and the electrolyte.
To validate and systematically quantify the CO2 shuttle problem, the researchers used a commercial Ag NW catalyst in a standard anionic MEA cell to reduce 2e- CO2 to CO (Figure 2 and Methods). Ag NW was chosen because of its commercial viability, good selectivity, and stability for the electroreduction of CO2 to CO. The results show that for a wide range of cell currents, approximately the same amount of CO2 is lost compared to the reduction of CO2 formed as CO product, indicating a significant carbon loss problem at a CO2 utilization efficiency of about 50%. On the anode side, a large amount of CO2 flow was detected in addition to the expected O2 gas produced by the OER (Fig. 2b). The CO2 flow measured from the anode side, together with the CO2-to-CO conversion rate, agrees well with the total CO2 consumption rate measured from the cathode side (Figure 2c). The agreement between these three independent measurements indicates the high accuracy of the experimental design in the carbon balance analysis. Therefore, the researchers concluded that most of the CO2 gas used in the system was crossed over to the other side and mixed with oxygen, resulting in significant energy losses and increased costs during CO2RR. CO2RR during cathodic electrolysis, where the interaction of hydroxide and CO2 at the cathodic AEM interface is dominated by carbonate rather than bicarbonate ions due to the strongly alkaline local pH. The CO2 conversion problem is exacerbated when high value-added products are targeted (Figure 2 d). Competitive by-products such as HER will further increase the CO2 shuttle rate during CO2RR electrolysis and reduce the efficiency of CO2 utilization.
Design of a PSE reactor system for shuttle CO2 recovery. As shown in Figure 3a, the solid electrolyte layer contains dense but permeable ion-conducting polymers functionalized with sulfonic acid groups, which ensure efficient proton conduction between the cathode and anode (just like the Nafion membrane) at a voltage comparable to that of the MEA device. Noting the above-mentioned carbonate shuttle phenomenon in MEA, we propose that a similar process will occur at the cathode in our solid electrolyte reactor design. The researchers hypothesized that the CO2RR on the cathode side of the solid electrolyte device would lead to the formation of hydroxide ions, which would then react with the free CO2 gas to form carbonate ions. These carbonate ions would be driven by an electric field to migrate across the AEM to the solid electrolyte layer, where they recombine with the protons generated by the anode OER to compensate for the charge. Thus, the solid electrolyte buffer layer can serve as a compound site for carbonate and protons without sacrificing the performance of CO2RR as in previous solid electrolyte reactor systems. To validate this idea and evaluate the CO2 recovery capability of the PSE reactor, it is critical to accurately measure the CO2 shuttle rate and the CO2 recovery rate. Shuttled CO2 gas will be dissolved in deionized water and then discharged as bubbles as the water begins to saturate. The researchers chose to measure the dissolved CO2 and CO2 bubbles in the DI water stream using the titration and water displacement methods, respectively (Figure 3b).
A model case study of CO2 recovery using the same Ag NW CO2-to-CO catalyst as the MEA test performed on the PSE reactor cathode (Figure 4a). The other side uses an IrO2 anode electrode to oxidize water to O2 and continuously supply protons to the solid electrolyte layer through the PEM. The scanning electron microscopy (SEM) and high resolution transmission electron microscopy (HRTEM) images of Fig. 4a show that the Ag NWs have a uniform diameter of about 70 nm. the lattice spacing and structure of the HRTEM indicate that the Ag NW surface is mainly covered by a (111) surface that is considered to be the active surface for the conversion of CO2RR to CO. The IV curve reactor of the solid electrolyte, where an additional intermediate layer is introduced, presents a similar CO2RR activity compared to that implied by the device (Figure 4b). The CO2RR selectivity of the solid electrolyte reactor is also comparable to, if not better than, the MEA, with a CO FE of about 90% at high currents up to 500 mA or 200 mA cm-2 (Fig. 4c). Our CO2 shuttle rate measured on the cathode side, which is equal to the total CO2 consumption rate (input-output) minus the CO2 conversion rate (CO in this case), is very close to the theoretical result for a wide range of cell operating currents, indicating the high accuracy of a gas analysis system specifically designed for carbon balance studies. Although the difference between these two values is small, it is interesting to note that the measured rate of CO2 shuttle is always slightly higher than the theoretical guide. This may be due to potential gas leakage from the cell assembly or pipeline connections, resulting in an underestimation of the downstream CO2 flux and thus an overestimation of the CO2 shuttle rate. The intermediate solid electrolyte layer consisted of dissolved CO2 and gas-phase CO2, and the measured CO2 recovery rate increased continuously with increasing cell current (Figure 4d). Considering that the DI water flow rate through the solid electrolyte layer is fixed at 1.1 ml min-1, the platform indicates a DI water flow rate of approximately 0.91 mlCO2/mlH2O for CO2 saturation under operating conditions, which is in good agreement with the theoretical solubility of CO2 in water. Here, all shuttled CO2 can be recovered in the gas phase only if the saturated DI water flow is continued to be circulated in practice. Over a wide range of operating currents, the cell continuously recovers total CO2 (both dissolved and bubbles) up to 90% of the amount of shuttled CO2 measured on the cathode side or approximately 100% of the theoretically calculated amount of shuttled CO2 (Figure 4e). The above assumptions can also well explain the small difference in CO2 recovery, i.e. the measured rate of CO2 shuttled may be slightly overestimated. The GC measurements (Figure 4f) confirmed that the gas purity of the recovered CO2 stream was higher than 99%, with only trace impurities, including H2 and O2, possibly from electrolysis or gas leakage.
CO2 gas recovery with different catalysts and CO2RR products. To demonstrate its general applicability, the CO2 recovery performance of the PSE design was further evaluated on different CO2RR catalysts and products. First, to investigate the possibility of using different CO2 - CO catalysts in the same recovery design, Ni-SACs were used. the TEM and SEM characterization in Figure 5a and in confirmed that the Ni single atoms are well dispersed in the carbon matrix. the CO2RR-to-CO performance of Ni-SACs has been demonstrated previously. Successful conversion in a solid electrolyte reactor provided industrially relevant currents (up to 500 mA or 200 mA cm-2) while maintaining more than 90% of CO FE (Fig. 5d,g). As shown in Figure 5j, the CO2 shuttle and recovery rates are similar compared to the Ag NW case, confirming that in the case of CO generation, the carbon loss and recovery mechanisms are independent of the type of catalyst we use. To test this hypothesis and to test the CO2 recovery capability in this scenario, the cathode catalyst was replaced with a two-dimensional Bi (2D-Bi) nanosheet catalyst (Figure 5b), which has been shown to be highly selective for formic acid in previous studies. As expected, 2D-Bi exhibited high activity and selectivity towards formic acid in the solid electrolyte cell design, with consistently high levels of FE over a wide current density range (Fig. 5e,h). Commercial CuNPs consist of nanocrystals with a diameter of 30~50 nm and uniform morphology (Figure 5c). In the PSE reactor, this CuNP catalyst exhibited significant CO2RR activity and selectivity for C2+ products, especially at higher currents, providing about 40% of the C2+ FE at 200 mA cm-2 (Fig. 5f,i). For this CuNP sample, C2H4,CO and H2 comprise most of the product, while HCOO- and CH3COO- together have only 10-14% FE, which leads to a less deviated rate of CO2 shuttling from Ag NW or Ni-SAC for Cu, as shown in Figure 5l. Encouragingly, the reactor still maintains the same high CO2 recovery performance, even in more complex cases, and the copper catalyst can yield up to seven different products, including C2+ products, which implies a wide range of future applications in the CO2RR field.
The water circulation system is stable over time. In practice, direct recovery of shuttled CO2 from the gas phase, rather than partial dissolution in water, allows direct feedback to the main CO2 stream. Continuous circulation of water in the interlayer eliminates the need for dissolved CO2 analysis, as the liquid remains saturated with CO2. After the first saturation, any CO2 gas entering the PSE layer will be extracted as pure gas. Using this setup, a practical demonstration of continuous operation on Ag NW catalyst at 250 mA or 100 mA cm-2 CO2 gas recovery was performed. As a visual demonstration, an air sphere of size (~100 mL) was filled with recovered CO2 gas from the solid electrolyte layer after 90 min of surgery, suggesting simple storage of recovered CO2 gas (Figure 6a, b). no significant changes appeared in the SEM images before and after Ag NW catalyst operation, indicating good structural stability. To evaluate the long-term performance of the gas recovery (Fig. 6c), the device was operated continuously for 750 h with no significant changes in CO selectivity (~90%) or cell voltage (~3.5 V, no iR compensation). During the stability test, the recovered CO2 remained relatively stable at 1.4 to 1.5 sccm. On average, about 80 - 90% of the shuttled CO2 was recovered from the cathode measurement side in the intermediate layer throughout the process, demonstrating the continued effective recovery potential of the device.
Recovering CO2 and improving CO2 conversion efficiency. To achieve a high continuous conversion efficiency, i.e., the proportion of supplied CO2 that is continuously converted to product (method), it is necessary to eliminate the excess CO2 gas supplied to the system. Therefore, while maintaining a 100 mA cm-2 cell current density (500 mA total current), the inlet CO2 flow to the gas recovery system is gradually reduced to monitor how it affects the CO2RR performance as well as the continuous conversion efficiency (Method). Figure 7b shows the carbon balance analysis of this system, including how much CO2 gas is converted to product and how much CO2 gas is retained at the cathode side outlet at different CO2 supply flow rates. We observe that the CO FE remains at a relatively high level (~85%) until the inlet CO2 flow rate decreases to 3.6 sccm with increasing continuous conversion efficiency (Fig. 7c). This decrease in product selectivity was mainly due to the lack of reactant CO2 and thus the rate of H2 reduction by water started to increase. By further reducing the CO2 inlet flow rate, an impressive 91% continuous conversion efficiency was obtained while maintaining over 60% CO FE. The downstream CO2 residual flow rate dropped to 0.18 sccm, while the outflows of CO and H2 were about 2.39 and 1.41 sccm, respectively, indicating a low residual CO2 concentration of 4.6%.
The main calculation and test methods of this study
- Field Test | In-situ Spherical Aberration Corrected Transmission Electron Microscope (JEOL JEM-ARM200F)
Synchrotron Radiation Find Easy Research
- Synchrotron X-ray Absorption Spectroscopy (XAFS)
Spherical Differential Electron Microscopy Find Easy Research
- Cryo-transmission electron microscopy (FEI Titan Krios)
- Double spherical aberration corrected transmission electron microscope (FEI Themis Z)
- Field Test | In-situ Spherical Aberration Corrected Transmission Electron Microscope (JEOL JEM-ARM200F)
- Field Testing | In-situ Environmental Spherical Aberration Corrected Transmission Electron Microscope (FEI Titan ETEM G2)
Doing Computing Find Easy Research
- First principles DFT calculation (VASP)
- Molecular dynamics simulations (Gromacs)
- Molecular Dynamics (LAMMPS)