Efficient CO2 Electroreduction over Silver Hollow Fiber Electrode

Electrocatalytic reduction of CO2 to fuels and chemicals is one of the most attractive routes for CO2 utilization. However, low efficiency and poor stability restrict the practical application of most conventional electrocatalysts. Here, a silver hollow fiber electrode is presented as a novel self-supported gas diffusion electrode for efficient and stable CO2 electroreduction to CO. A CO faradaic efficiency of over 92% at current densities of above 150 mA·cm−2 is achieved in 0.5 M KHCO3 for over 100 h, which is comparable to the most outstanding Ag-based electrocatalysts. The electrochemical results suggest the excellent electrocatalytic performance of silver hollow fiber electrode is attributed to the unique pore structures providing abundant active sites and favorable mass transport, which not only suppresses the competitive hydrogen evolution reaction (HER) but also facilitates the CO2 reduction kinetics.


Introduction
The electroreduction of carbon dioxide (CO 2 ) to useful chemicals leads to a promising pathway for both CO 2 utilization and the storage of renewable electricity, which is of great significance for achieving carbon neutrality [1,2]. Diverse valuable compounds including syngas, formate [3], methane [4], ethylene [5] and ethanol [6] with a collective market size of over 500 megatonnes per year can be obtained from electrocatalytic CO 2 reduction reaction (eCO 2 RR) [7]. Among them, carbon monoxide (CO) as an important component of the commodity syngas (a mixture of CO and H 2 ) is widely used in the various chemical engineering processes such as Fischer-Tropsch synthesis and methanol synthesis [8,9]. Producing CO on a large scale via CO 2 electrochemical reduction driven by renewable electricity shows an application potential. With respect to the other products such as formic acid (or formate), alcohols and hydrocarbons from CO 2 electroreduction, CO is not only more easy to separate from the aqueous electrolyte solution, but also efficiently generated with fewer electrons (two-electron transfer) at slight negative potentials [10]. Therefore, there have been many efforts focusing on electrocatalyst development to achieve highly efficient CO 2 conversion to CO [7,[10][11][12][13][14][15].
Among various materials that have been studied as electrocatalysts, silver (Ag) is a promising material that possesses the efficient capability to electroreduce CO 2 into CO and also costs much less than other precious metal catalysts [16,17]. Its all-inorganic nature would be more stable than homogeneous catalysts [18,19]. Although faradaic efficiency (FE) of up to 90% for eCO 2 RR towards CO has been achieved over various Ag-based catalysts, such as OD-Ag [20], BD-Ag [21], POD-Ag [22], P-Ag [23] and AE-Ag [24], in an aqueous H-cell type electrolyzer, the current densities of these catalysts are generally All mentioned electrochemical experiments with the Ag HF electrode as both the working electrode and gas diffuser were conducted in CO2-saturated KHCO3 solution ( Figure S7, see Section 4 for details). In this work, a flow rate of 30 mL•min −1 of CO2 was chosen to ensure that the reaction is not limited by a lack of CO2 supply and to compare our performance with other studies [34,36]. As shown in Figure 2A, only CO and H2 were detected over the Ag HF electrode with a total faradaic efficiency (FE) of ~100% in the potential range of −0.7 to −1.4 V. Note that the generation of H2 was well suppressed within the whole potential range (−0.7 to −1.4 V), consistently resulting in CO FEs greater than 80%. The exclusive formation of CO for eCO2RR led to a CO FE greater than 90%, especially at −0.9 to −1.3 V. In addition, the corresponding current densities (j) of eCO2RR products showed the same trend ( Figure 2B). Both total current densities and CO partial current densities (jCO) increased rapidly at more negative potentials, while the H2 partial current densities (jH2) increased slowly, giving a ~141 mA•cm −2 jCO with a CO FE as high as 92.7% at −1.2 V. The corresponding CO formation rate, cathodic energy consumption, outlet CO concentration and CO energy efficiency of Ag HF were 2636.8 µmol•h −1 •cm −2 , 91.5 mW•cm −2 , 1.79% and 51.1%, respectively (Table S1), comparable to those of other prominent electrocatalysts [31,35,38]. The CO FEs of Ag foil were lower than 60% in all potentials, giving the largest CO FE of ~56% at −0.9 V with a lower 5 mA•cm −2 jCO ( Figure  S8). In contrast, the jCO of Ag HF reached as high as 194 mA•cm −2 at −1.4 V, which is about 42 times that for Ag foil (4.6 mA•cm −2 ), evidencing the striking promotion in the intrinsic activity of Ag.
The long-term performance of electrocatalysts is of great importance for their practical applications [26]. Although many studies [20][21][22][23][24][25] have reported that Ag-based electrocatalysts possess the capability to selectively electroreduce CO2 into CO, most of their current densities for long-term tests are restricted below 100 mA•cm −2 ( Figure 2C and  Figure S6) for Ag HF showed Ag 3d 5/2 and Ag 3d 3/2 peaks at binding energies of 368.2 and 374.2 eV, respectively, indicating the metallic Ag 0 characteristics of the sample surface, which is identical to commercial Ag powder and Ag foil. These results verified that both the bulk and surface compositions of Ag HF are completely identical to those of Ag powder and Ag foil as metallic Ag 0 .
All mentioned electrochemical experiments with the Ag HF electrode as both the working electrode and gas diffuser were conducted in CO 2 -saturated KHCO 3 solution ( Figure S7, see Section 4 for details). In this work, a flow rate of 30 mL·min −1 of CO 2 was chosen to ensure that the reaction is not limited by a lack of CO 2 supply and to compare our performance with other studies [34,36]. As shown in Figure 2A, only CO and H 2 were detected over the Ag HF electrode with a total faradaic efficiency (FE) of~100% in the potential range of −0.7 to −1.4 V. Note that the generation of H 2 was well suppressed within the whole potential range (−0.7 to −1.4 V), consistently resulting in CO FEs greater than 80%. The exclusive formation of CO for eCO 2 RR led to a CO FE greater than 90%, especially at −0.9 to −1.3 V. In addition, the corresponding current densities (j) of eCO 2 RR products showed the same trend ( Figure 2B). Both total current densities and CO partial current densities (j CO ) increased rapidly at more negative potentials, while the H 2 partial current densities (j H2 ) increased slowly, giving a~141 mA·cm −2 j CO with a CO FE as high as 92.7% at −1.2 V. The corresponding CO formation rate, cathodic energy consumption, outlet CO concentration and CO energy efficiency of Ag HF were 2636.8 µmol·h −1 ·cm −2 , 91.5 mW·cm −2 , 1.79% and 51.1%, respectively (Table S1), comparable to those of other prominent electrocatalysts [31,35,38]. The CO FEs of Ag foil were lower than 60% in all potentials, giving the largest CO FE of~56% at −0.9 V with a lower 5 mA·cm −2 j CO Catalysts 2022, 12, 453 4 of 11 ( Figure S8). In contrast, the j CO of Ag HF reached as high as 194 mA·cm −2 at −1.4 V, which is about 42 times that for Ag foil (4.6 mA·cm −2 ), evidencing the striking promotion in the intrinsic activity of Ag.  Table S2).
The long-term performance of electrocatalysts is of great importance for their practical applications [26]. Although many studies [20][21][22][23][24][25] have reported that Ag-based electrocatalysts possess the capability to selectively electroreduce CO 2 into CO, most of their current densities for long-term tests are restricted below 100 mA·cm −2 ( Figure 2C and Table S2). On the other hand, although relatively higher current densities have been achieved through GDEs in highly alkaline electrolytes, the use of basic media poses significant stability challenges [27,32]. For example, the carbon-based GDE was found to degrade over 2 h in a basic electrolyte during eCO 2 RR [38]. In sharp contrast, the durability of a Ag HF electrode was evaluated in a continuous CO 2 electrolysis test operated at −1.2 V with 0.5 M KHCO 3 . As shown in Figure 2C, the CO FE remained between 92% and 93% with a fluctuating total current density of 150-160 mA·cm −2 . No sign of decline was observed during the 100 h test. Such excellent long-term performance is comparable to that of excellent Ag-based GDEs in a flow cell system and much higher than that of outstanding Ag-based electrocatalysts in an aqueous H-cell electrolyzer ( Figure 2D and Table S2) [11,[19][20][21][22][23][24][25]32,[39][40][41][42][43][44][45][46][47]. The postreaction XRD and XPS ( Figure S9) revealed the stable compositions of Ag HF after electrolysis, which were responsible for the steady CO 2 electroreduction performance, providing great prospects for scalable eCO 2 RR applications.
The electrochemically active surface areas (ECSAs) of Ag HF and Ag foil were examined by measuring their double-layer capacitance (C dl ) values from their cyclic voltammetry curves ( Figures 3A and S10). The resulting ECSAs were 6.0 and 1.9 mF·cm −2 for Ag HF and Ag foil, respectively. Note that the ECSA value for Ag HF is 3.2 times that for Ag foil, far away from the disparity in j CO (38 times) (Figures 2B and S11) at the same potential (−1.2 V). Then, we normalized the j CO and j H2 by their ECSAs ( Figure S12). The results showed that the j CO of Ag HF was always much higher than that of Ag foil at the same potentials regardless of whether it was normalized or not (Figures 2B, S11A and S12A). In addition, even though the j H2 plot showed that the Ag HF had a similar j H2 to the Ag foil ( Figure S11B), the normalized j H2 of the Ag HF was much lower than that of the Ag foil at the same potentials ( Figure S12B). These results implied that the high value of ECSA may only play a partial role in the efficient formation of CO over the Ag HF electrode [35]. The intrinsic activity of the Ag HF may largely be promoted by its unique structure, facilitating the eCO 2 RR while the HER was suppressed. Then, a Tafel analysis was further performed to gain insight into the underlying kinetic mechanism for the eCO 2 RR over the Ag HF ( Figure 3C). The Tafel slopes of both Ag foil (162 mV·dec −1 ) and Ag HF (112 mV·dec −1 ) were close to 118 mV·dec −1 , which is the commonly suggested Tafel slope when the rate-determining step for eCO 2 RR is the initial electron transfer to CO 2 to form a surface-adsorbed *COOintermediate [12,48] (step 1 of Figure S13). Note that the Tafel slope of Ag HF was much lower than that of Ag foil, implying a faster initial electron transfer to a CO 2 molecule for CO 2 activation [49], which may improve the intrinsic catalytic activity of Ag HF towards CO formation. Such a faster electron transfer was also verified by the lower interfacial charge transfer resistance (R ct ) of Ag HF (0.8 Ω·cm 2 ) compared with that of Ag foil (2.3 Ω·cm 2 ), as shown in Figure 3D. The improved charge transfer indicates that faster electrochemical reduction occurs on the Ag HF [37], and this is in accordance with the eCO 2 RR performance results shown in Figure 3B. Consequently, these results suggested that the improved initial one-electron transfer enhanced the intrinsic CO 2 reduction activity over the Ag HF, resulting in such high activity for the electrocatalytic reduction of CO 2 to CO.

Discussion
Besides the improved initial one-electron transfer and the higher-ECSA-enhanced intrinsic CO 2 reduction activity of Ag HF, the high activity and selectivity for the electrocatalytic reduction of CO 2 to CO might also be associated with the favorable mass transfer and abundant three-phase reaction interfaces of the Ag HF due to its compulsory gas flow-through configuration [34]. The CO 2 molecules are forced to penetrate through the porous wall of the Ag HF electrode, resulting in compulsive interaction of CO 2 with the reaction active sites and further effective activation, which synergistically facilitates CO formation [34,35]. On the other hand, the hollow fiber configuration of the Ag HF electrode might also be beneficial for the removal of CO from the electrode surface, induced by the very high local concentration of CO 2 near the electrode surface, which might boost the CO production rate over the Ag HF [34].
Compared with the conventional GDEs with multiple components, the Ag HF exhibits the following striking merits: (1) It is formed from a single component. High-purity commercial silver powder was subjected to the combined phase inversion/sintering process to obtain the Ag HF without any additive binder. (2) It has a tough and integral substrate. Due to the sintering step, the Ag particles in Ag HF are fused, indicating the formation of a tough and integral substrate, which not only exhibits structural stability but also facilitates electron transport. (3) It has tunable pore structures. The pore structures of Ag HF can be further improved via the optimized preparation procedures to promote CO 2 supply and dispersion, resulting in enhanced CO 2 electroreduction performance.
As a matter of fact, although the Ag HF we reported here could efficiently and stably electroreduce CO 2 to CO by virtue of abundant active sites and favorable mass transport due to its unique pore structures, the ≤300 mA·cm −2 current densities are still limitations to affording an economically viable CO 2 electrochemical conversion [26,30]. Future studies could be aimed at the optimization of hollow fiber configurations, such as tuning the surface morphology and size of Ag nanoparticles to obtain more active sites through synthesizing nanostructured [50] or oxide-derived Ag catalysts [20] to give full play to the improved favorable mass transport of the Ag HF for more efficient CO 2 electroreduction for scalable applications.

Catalyst Preparation
The Ag HF was fabricated by a combined phase-inversion/sintering process ( Figure S1). Briefly, commercially available polyetherimide (PEI, 24 g) was added to N-methyl-2pyrrolidone (NMP, 96 g), followed by ultrasonic treatment for 1 h to obtain a homogeneous and transparent solution. Then, Ag powder (80 g) was added to the above solution. The as-obtained mixture was further treated by ball-milling (300 rpm) for 24 h to form a uniform slurry. After cooling to room temperature, the slurry was vacuumized (1 mbar) for 5 h to remove bubbles to obtain a casting solution. Next, the casting solution was extruded through a spinning machine and shaped in a water bath via the phase-inversion process. After spinning, the as-formed tubes were kept in a water bath for 24 h to eliminate the solvent completely, followed by stretching and drying for 48 h to obtain a green body. The green body was cut into appropriate lengths and then calcinated in an air flow (100 mL·min −1 ) at 600 • C (heating rate: 1 • C·min −1 ) for 6 h to remove PEI. After being naturally cooled to room temperature, the calcined green body was then reduced in a 5% H 2 (argon balance) flow (100 mL·min −1 ) at 300 • C (heating rate: 1 • C·min −1 ) for 3 h to obtain Ag HF.
The Ag HF with an exposed length of 4 cm was stuck into a copper tube using conductive silver adhesive for electrical contact, while the end of the Ag HF tube as well as the joint between the Ag HF and copper tube were sealed and covered with gas-tight and nonconductive epoxy. After drying at room temperature for 24 h, a Ag HF electrode was obtained with an exposed geometric area of 0.5 cm 2 (S = πDL = 3.14 × 400 × 10 −4 × 4 = 0.5 cm 2 ) and a silver loading of 14 ± 1 mg·cm −2 .
A piece of Ag foil was ultrasonically cleaned in acetone and ethanol, and after drying in air, the side and back of the Ag foil were sealed with epoxy to obtain a Ag foil electrode with an exposure geometric area of 1 cm × 0.5 cm and a silver loading of 1000 ± 50 mg·cm −2 .

Catalyst Characterization
The cross-section and surface morphologies of Ag HF were observed via scanning electron microscopy (SEM) using a Supra 55 microscope with an accelerating voltage of 5.0 kV. Transmission electron microscopy (TEM) investigations were conducted with a JEM-ARM300F microscope operated at 300 kV. X-ray diffraction (XRD) measurements were performed on a Rigaku Ultima 4 X-ray diffractometer using a Cu Kα radiation source (λ = 1.54056 Å) at 40 kV and 40 mA. X-ray photoelectron spectroscopy (XPS) tests were conducted using a Quantum 2000 Scanning ESCA Microprobe instrument with a monochromatic Al Kα source (1486.6 eV). The binding energies in all XPS spectra were calibrated according to the C 1s peak (284.8 eV).

Electrochemical Measurements
Electrochemical characterization was performed on a Biologic VMP3 potentiostat (Bio-Logic Inc., Seyssinet-Pariset, France) in a two-compartment electrolysis cell with a three-electrode configuration at room temperature. The Ag HF electrode was used as the working electrode, with a KCl-saturated Ag/AgCl reference electrode in the cathodic compartment and a platinum mesh counter electrode in the anodic compartment ( Figure S7). The electrochemically active surface area (ECSA) of the electrode was evaluated by the double-layer capacitance (C dl ). The C dl was determined by performing cyclic voltammetry (CV) in the potential range of 0.4 to 0.5 V (vs. RHE) at different scan rates in CO 2 -saturated 0.5 M KHCO 3 . The electrochemical impedance spectroscopy (EIS) measurements were performed in CO 2 -saturated 0.5 M KHCO 3 at −1.2 V (vs. RHE), and the frequency limits were typically set in the range of 0.1 Hz to 100 kHz with a voltage amplitude of 50 mV. Prior to the experiments, the electrolysis cell was vacuumized and then purged with high-purity CO 2 (99.999%, Shanghai Pujiang Special Gas Corp., Shanghai, China) for 30 min, after which CO 2 was continuously delivered into the cathodic compartment at a constant rate of 30 mL·min −1 . All the applied potentials were recorded against the KCl-saturated Ag/AgCl reference electrode and then converted to those versus the reversible hydrogen electrode (RHE) with iR corrections by the following equation: where E (vs. Ag/AgCl) is the applied potential, pH is the pondus hydrogenii value of the electrolyte solutions (~7.2, CO 2 -saturated 0.5 M KHCO 3 ), i is the current density at each applied potential, and R s is the solution resistance obtained by EIS measurements (~5.7 Ω·cm 2 ). All applied potentials in the main text and Supplementary Materials refer to the RHE unless otherwise stated.
During the stability test of Ag HF CO 2 electroreduction, the potential was fixed at −1.2 V (vs. RHE), the electrolyte was CO 2 -saturated 0.5 M KHCO 3 and the CO 2 flow rate was kept at 30 mL·min −1 . The catholyte and anolyte were cycled at a flow rate of 10 mL·min −1 by using two identical peristaltic pumps (Jihpump BT-50EA 153YX), accompanied by the supplement of ultrapure water to maintain a constant concentration of 0.5 M KHCO 3 . In addition, the postreaction catholyte and anolyte were subjected to inductively coupled plasma element measurements, and no dissolved silver or platinum ions were found.

Product Quantifications
Gas-phase products from the cathodic compartment were directly vented into a gas chromatograph (GC-2014, Shimadzu Co., Ltd., Kyoto, Japan) equipped with a Shincarbon ST80/100 column and a Porapak-Q80/100 column with a flame ionization detector (FID) and a thermal conductivity detector (TCD) during the electroreduction tests and analyzed online. The FID detector was used for CO quantification (as well as CH 4 , C 2 H 4 and C 2 H 6 ), while TCD was used for H 2 quantification. All faradaic efficiencies reported were based on at least three different GC runs. High-purity argon (99.999%) was used as the GC carrier gas. In all the CO 2 electrolysis tests, only H 2 and CO were the gas-phase products, and their faradaic efficiencies were calculated as follows: where C product is the concentration of the gas-phase products (ppm), ν CO2 is the flow rate of CO 2 (30 mL·min −1 ), t is the reaction time, α is the number of transferred electrons for producing CO or H 2 , F is the Faraday constant, V m is the gas mole volume and Q is the total quantity of the electric charge. The CO formation rate was calculated using the following equation: where S is the geometric area of the electrode (cm 2 ). The cathodic energy consumption was calculated as follows: where E is the applied potential vs. RHE after iR compensation. By assuming that the overpotential of oxygen evolution reaction on the anode side is zero, the cathodic energy efficiency for CO was calculated as follows [38]: where E CO is −0.11 V (vs. RHE); 1.23 V is the thermodynamic potential for water oxidation on the anode side. Possible liquid-phase products from the cathodic compartment after CO 2 electrolysis for 1 h were analyzed using another off-line GC-2014 (Shimadzu Co., Ltd., Kyoto, Japan) equipped with a headspace injector and an OVI-G43 capillary column (Supelco ® , Sigma-Aldrich Inc., St. Louis, MO, USA). No liquid-phase products were detected by the off-line GC. The postreaction catholyte solution was also analyzed by a 600 MHz NMR spectrometer (Bruker, Karlsruhe, Germany) for possible liquid-phase products, especially formate and acetate. After an hour of electrolysis, an aliquot of catholyte solution (0.5 mL) was mixed with 0.1 mL DSS (6 mM) as internal standard and 0.1 mL D 2 O. No liquid-phase product was detected by NMR.

Conclusions
In this work, we report a three-dimensional silver hollow fiber electrode used as both a working electrode and a gas diffuser for highly efficient and stable electroreduction of CO 2 to CO. A CO faradaic efficiency of over 92% at current densities of above 150 mA·cm −2 with a 100 h sustained performance was achieved in 0.5 M KHCO 3 , which is comparable to the most outstanding Ag-based electrocatalysts. The experimental results suggested that the excellent electrocatalytic performance of the electrode is attributed to the unique pore structures, providing abundant active sites in addition to favorable mass transport, which not only suppressed the competitive HER but also facilitated the CO 2 reduction. In addition, the Ag HF may become an ideal industrial electrode due to its tough framework and mature preparation process, showing great potential for scalable applications.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/catal12050453/s1, Figure S1. Schematic illustration showing the general procedures for the fabrication of the Ag HF. Figure S2: SEM image of commercially available silver powder; Figure S3: SEM images of the (A,B) outer and (C,D) inner surfaces of the Ag HF. Figure S4: XRD patterns of Ag foil, Ag powder and Ag HF. Figure S5: TEM images of (A) Ag foil and (B) Ag HF. Figure S6: XPS spectra of Ag foil, Ag powder and Ag HF. Figure S7: Schematic illustration of porous silver hollow fiber for efficient CO production via eCO 2 RR. Figure S8: (A) CO and H 2 faradaic efficiencies and (B) current densities of eCO 2 RR over Ag foil in the potential range from −0.7 to −1.4 V. Figure S9: (A) XRD patterns and (B) XPS spectra of Ag foil and Ag HF before and after eCO 2 RR. Figure S10: Cyclic voltammetry curves of (A) Ag foil and (B) Ag HF in 0.5 M KHCO 3 . (C) Plot of ∆j (the difference of cathodic and anodic current densities, j c − j a ) against the scan rates from cyclic voltammetry curves. The plots in Figure S10C are the same as those in Figure 3A in the main text. Figure S11: (A) CO and (B) H 2 partial current densities over Ag foil and Ag HF in the potential range of −0.7 to −1.4 V. Figure S12: (A) ECSA-normalized CO and (B) H 2 partial current densities over Ag foil and Ag HF in the potential range of −0.7 to −1.4 V. Figure S13: Proposed reaction steps for the electroreduction of CO 2 to CO on silver catalysts. Table S1: Detailed eCO 2 RR performances of the Ag HF electrode. Table  S2: Electrocatalytic performances for CO 2 to CO over typical recently reported Ag-based catalysts.