To supply more CO2 to a cathode, we have been studying a gas-phase CO2 reduction reaction (GCRR) cell, in which gas-phase CO2 is supplied and consumed in a reduction reaction with a metal fiber electrode (cathode) on a proton exchange membrane.1 We have reported that the faradaic efficiency of the CO2 reduction reaction (FE CO2RR)can be improved by changing the cathode materials (Au, Cu, etc.) in the GCRR cell.2 However, we also observed that the FE CO2RR decreases over time due to carbonate accumulation on the cathode surface, which is assumed to reduce the amount of gas-phase CO₂ supplied to the cathode surface. We considered that by removing the carbonate, we could maintain the FE CO2RR. In this study, we attempted to supply water vapor to the cathode surface. In addition, we theoretically derived the humidity conditions necessary for carbonate dissolution and experimentally verified whether the water vapor supply could remove generated carbonate and maintain the FE CO2RR during 500-h continuous tests, in a world first in GCRR cell research.We hypothesized that the carbonate dissolution rate surpassing the carbonate generation rate is essential for effective carbonate dissolution. To verify this hypothesis, we prepared two humidity conditions: a high-humidity condition with a higher dissolution rate and a low-humidity condition with a higher generation rate. The high-humidity condition was achieved using the bubbling method, resulting in CO2 humidities of 0.28 g/m3 and 0.024 g/m3 under high-humidity and low-humidity conditions, respectively. The CO2 flow rate was maintained at 35 sccm. The sample was prepared by thermocompression bonding an Au fiber sheet (area: 1 x 1 cm) and a proton exchange membrane Nafion 117 (manufactured by Sigma-Aldrich) at 423 K and 3.5 MPa, then positioned between the photoanode and the cathode chamber. An AlGaN/n-GaN heterostructure grown on a sapphire substrate served as the photoanode, with AlGaN serving as the light absorption layer. A NiO thin film was fabricated on the surface of AlGaN as a co-catalyst.3 The photoanode chamber was filled with a 1.0 mol/L KOH electrolyte solution, and the photoanode was immersed in the solution. The photoanode was irradiated with light from an Xe lamp with an irradiance of 2.2 mW/cm2 (λ ≤ 365 nm). To avoid the drop in photocurrent due to deterioration of the photoanode, the photoanode was replaced after 268 h and 308 h under high-humidity and low-humidity conditions, respectively. The photocurrent between the photoanode and cathode was measured using a potentiostat with an applied potential of 0.5 V, and the amount of product gases (H2, CO, CH4, C2H4) in the cathode chamber was measured with a gas chromatograph.Figure 1 shows the amount of CO and H2 production over irradiation time alongside images of the cathode surface after 500 h, under high-humidity and low-humidity conditions. The average photocurrent density over 500 h was 0.15 mA/cm2 under both high-humidity and low-humidity conditions. Photocurrent density is defined as photocurrent normalized by the area of the photoanode. This result suggests nearly identical average resistances of the reaction cells. Observation of the cathode surface after 500 h revealed minimal carbonate formation under high-humidity conditions (Fig. 1 (a)). On the other hand, under low-humidity conditions, the cathode surface was predominantly covered with carbonate (Fig. 1 (b)). Only CO and H2 gases were detected from the cathode chamber, with no other gases detected. The cumulative CO amounts produced after 500 h per photoanode area under high-humidity and low-humidity conditions were 1330 μmol/cm2 and 970 μmol/cm2, respectively (Fig. 1). In addition, the FE CO2RR calculated from the CO amount produced was 94% and 71%, respectively, and the FE CO2RR was maintained over 500 h under high-humidity conditions compared with low-humidity conditions. These results suggest that under high-humidity conditions, where the carbonate dissolution rate exceeded the generation rate, carbonate dissolution and removal occurred as hypothesized. Additionally, we considered that the removal of the carbonate resulted in the sustained CO2 supplied to the cathode surface.[1] S. Sato et al., The 89th ECSJ Spring Meeting, 1B14 (2022) (in Japanese).[2] S. Sato et al., The 90th ECSJ Spring Meeting,1C19 (2023) (in Japanese).[3] Y. Uzumaki et al., PRiME 2020, L04-3054 (2020). Figure 1
Read full abstract7-days of FREE Audio papers, translation & more with Prime
7-days of FREE Prime access