Nanoscale Management of CO Transport in CO2 Electroreduction: Boosting Faradaic Efficiency to Multicarbon Products via Nanostructured Tandem Electrocatalysts

Tandem catalysis presents a promising strategy to improve the selectivity toward multicarbon products in the electrocatalytic carbon dioxide reduction reaction (CO2RR). For CO2RR, CO is a critical intermediate for producing multicarbon products. However, the management of CO localization and CO diffusion remains underexplored despite its critical role. Herein, a 3D tandem catalyst electrode with silver nanoparticles (Ag NPs) is designed to generate CO as an intermediate product within a copper (Cu) nanoneedle array. Via this nanostructured design, CO2 forms C2+ products with a high Faradaic efficiency (FEC2+) of 64% in an H‐cell and 70% in a flow cell with a current density of 350 mA cm−2. These figures‐of‐merit are currently among the top literature reports. More importantly, in situ Raman spectroscopy and finite‐element method calculations are employed to elucidate the origins of enhanced selectivity. These approaches reveal the crucial role of prolonging the CO diffusion path length for improving CO utilization during CO2 conversion with tandem catalyst systems. The favorable CO2RR FEC2+ in two distinct environments (H‐cell and flow cell) further corroborates that this effect is not limited to a particular reactor environment. Overall, this study provides new insights for designing tandem catalysts for improved CO2RR selectivity to C2+ products.

and initial 2-electron reduction of CO 2 . Therefore, one approach to promote the formation of C 2+ products is to create nanoconfined spaces in Cu-based materials. The nanoconfined spaces can enhance the concentration and interaction between *CO by trapping more CO adsorbates. [24,25] For instance, Zeng et al. boosted C 2+ selectivity on Cu by introducing Cu hollow multishell structures to increase the coverage of localized CO. [26] Another representative approach is to increase CO generation rates to elevate the generation of C 2+ products. [27,28] Among various strategies, combining CO-generating metal atoms (e.g., silver [Ag], [29] gold, [30] and nickel [31] ) with Cu-based materials to form tandem catalysts is an effective way to increase the local concentration of CO.
In the tandem catalytic system, CO 2 is first reduced to CO by a CO-forming metal with high selectivity. The CO then spills over to nearby Cu sites, where CO is further reduced to C 2+ products through a C-C coupling and hydrogenation process. [32] However, if the CO generation rate exceeds the C-C coupling rate, the CO utilization rate plateaus, resulting in low C 2+ product selectivity. [33,34] For this reason, although several tandem catalysts have been reported to exhibit better C 2+ product selectivity compared to pure Cu catalysts, their CO FE remains high. [35][36][37] Prolonged CO residence time can increase the possibility of CO resorption and further reduction, which can reduce CO FE and increase C 2+ FE. Therefore, the residence time of CO at Cu sites in Cu-based tandem catalysts is an important yet often neglected performance indicator. [38] Proper distribution and localization of CO-forming catalysts in the tandem system should enable the electrode to form C 2+ products with high activity and selectivity. Inspired by these findings, for a tandem CO 2 RR catalyst, we believe that we can promote C 2+ products with high activity and selectivity by tuning the amount of CO-forming catalyst and the nanoscale morphology of the C 2+ selective catalyst to modulate CO diffusion.
Thus, in this work, we design and fabricate a tandem catalyst structure consisting of Ag nanoparticles (Ag NPs) deposited on the bottom of a Cu nanoneedle array for selective CO 2 reduction to C 2+ products. The dense Cu needle subjects the outflow of CO from the Ag NPs to high diffusion resistance. In addition, the CO undergoes a long diffusion path length, which makes it easier for local C 1 species to dimerize to C 2+ products. Additionally, the spatial management of the CO intermediates along the Cu needle length is achieved by controlling the mass ratio between the Cu needle and Ag NPs. In particular, when the mass ratio of the Cu needles to Ag NPs is 1:2, the Cu needle-Ag electrode can effectively achieve a superior FE of about 64% for C 2+ products at the potential of −1.00 V versus RHE in an H-cell and 70% at 350 mA cm −2 in a flow cell. In situ Raman spectroscopy combined with finite-element method calculations revealed that, in the Cu-Ag tandem catalyst electrode, adjusting the length of the Cu nanoneedles (C 2+ selective part) can maximize the utilization of CO spilled from the Ag catalyst, which thus leads to the enhancement of the selectivity of C 2+ products. Compared to other literature reports, this strategy achieves high C 2+ FE and current density within both H-cell and flow-cell environments, demonstrating the robustness of this approach to controlling CO diffusion even within different reactor configurations. Overall, this work constructs an effective tandem electrode structure for CO 2 RR with high C 2+ selectivity, and this work demonstrates the importance of adjusting the spacing and length of the C 2+ selective component in the tandem catalyst for better CO utilization.

Catalyst Synthesis and Characterization
The Cu(OH) 2 nanoneedle substrate was prepared as the precursor to the C 2+ selective component for the tandem CO 2 RR catalyst. To this end, a Cu-coated carbon paper was first prepared using a previously reported electroless Cu plating method. [42,43] Briefly, a superhydrophobic carbon paper was used ( Figure S1a, Supporting Information), and scanning electron microscopy (SEM) confirms that the Cu plating layer is uniformly and well-attached to the carbon fibers ( Figure S1b, Supporting Information). Only one side of the superhydrophobic carbon paper is coated with Cu. The other side maintains the carbon paper's original superhydrophobic structure (as shown in Figure S1c,d and Figure S2, Supporting Information). The maintenance of the hydrophobic structure on the back of the carbon paper allows the gas to pass from the back to the surface of the catalyst; this feature enables the samples to be applied directly into different CO 2 reduction reactors. Powder X-ray diffraction (XRD) of the carbon paper shows only the peak of carbon at 26.4° (Figure S1e, Supporting Information). After Cu plating, peaks at 43.4°, 50.4°, and 73.9° could be assigned to Cu (111), Cu (200), and Cu (220), respectively. Cu(OH) 2 needles were prepared on the Cu-coated carbon paper by an anodization process ( Figure S1f, Supporting Information). The morphology and density of Cu(OH) 2 needles can be controlled by controlling the reaction time. The density of the Cu(OH) 2 needles increased as the reaction time increased from 5 to 10 min. However, as the reaction time increases to 15 min, Cu(OH) 2 dehydrates through a breaking of the interplanar hydrogen bonds, and CuO flakes begin to form. As reaction time further increases, the needle-like structure eventually disappears ( Figure S3, Supporting Information). [44] In order to obtain a high density of Cu(OH) 2 needles, the reaction time was set at 10 min in this work. Transmission electron microscopy (TEM) imaging reveals that the average diameter of the typical Cu(OH) 2 needle is 160 nm ( Figure S1g, Supporting Information). As displayed in Figure S1h,i (Supporting Information), the corresponding energy-dispersive X-ray spectroscopy (EDS) elemental mapping exhibits elemental Cu and O in Cu(OH) 2 . Additionally, the XRD pattern further proves that Cu(OH) 2 needles were successfully grown on the Cu-coated carbon paper ( Figure S1j, Supporting Information).
The fabrication of the tandem electrode (Cu needle-Ag) with needle structure is schematically presented in Figure 1a. Ag NPs solution was dispersed onto the prepared Cu(OH) 2 needles and then dried to deposit the Ag NPs at the bottom of the Cu(OH) 2 needle array ( Figures S4 and S5, Supporting Information). The mass loading of Ag NPs can be tuned by controlling the dosage of the Ag NP solution. XRD patterns of the Cu(OH) 2 needle-Ag verify that it was composed of Cu(OH) 2 and Ag ( Figure S6, Supporting Information).  Figure 1e. Before the CO 2 RR test, Cu needles-Ag were obtained via electrochemical reduction of Cu(OH) 2 needle-Ag at −0.9 V versus RHE for least 5 min ( Figure S7, Supporting Information). Figure S8a,b (Supporting Information) shows SEM and TEM images of prepared Ag NPs by redox reaction (see Supporting Information for details). The average diameter of the Ag NPs is 52.1 ± 0.6 nm ( Figure S7c, Supporting Information). The lattice spacing of 0.24 nm corresponds to the (111) planes of the Ag ( Figure S7d,e, Supporting Information). Moreover, the XRD pattern also displays typical diffraction peaks of Ag ( Figure S7f, Supporting Information). The catalytic performance of the as-synthesized Ag NPs was also tested in an H-cell. As shown in Figure S9 (Supporting Information), a high CO FE is observed across all tested potentials. In particular, it reaches an average peak FE of 90.4% for CO at −0.9 V versus RHE, indicating its excellent catalytic selectivity toward CO product in CO 2 RR.

Electrochemical CO 2 Reduction Performance
The CO 2 reduction performances of the Cu needle and Cu needle-Ag electrodes in 0.1 m KHCO 3 electrolyte were first evaluated in a H-type cell. Linear sweep voltammetry (LSV) in N 2 -and CO 2 -saturated electrolytes showed that the Cu needle-Ag catalyst affords a more positive onset potential and larger current densities than Cu needle catalyst, indicating the enhanced activity with the cooperation of Cu needle and Ag NPs ( Figure S10, Supporting Information). Additionally, double-layer capacitance measurements of the Cu needle and Cu needle -Ag in 0.1 m KHCO 3 were used to determine the electrochemical active surface areas (ECSA) of the electrodes ( Figure S11a,b,g, Supporting Information). The Cu needle-Ag shows a larger ECSA value than the Cu needles, indicating that it has more electrochemically active sites for CO 2 RR (Figure S11e, Supporting Information). Furthermore, Brunauer-Emmett-Teller (BET) measurements were used to determine the specific surface areas of Cu needle and Cu needle-Ag. N 2 adsorption-desorption isotherms and corresponding results indicate that the specific surface areas of Cu needle and Cu needle-Ag are very close (10.97 and 13.36 m 2 g −1 , respectively) ( Figure S11c,d,f, Supporting Information). Therefore, the specific surface area is not the main reason for the difference in their catalytic performance. Figure 2a compares the current densities exhibited by the two samples during CO 2 RR under different potential. The total current densities (j total , dotted line) of Cu needle -Ag remains higher than that of Cu needle under different potentials. The partial current densities of C 2 H 4 (j C2H4 , solid line) far exceed that of the Cu needle at −0.9 V and −1.0 V, while it is almost the same as that of the Cu needle at −0.8 and −1.1 V. In addition, the total current density and C 2 H 4 partial current density of Cu needle and Cu needle-Ag in H-cell normalized by ECSA also show similar results to the current density calculated from the geometric area ( Figure S12, Supporting Information).
The FE of each sample was assessed. For the Cu needle catalyst alone, the maximum C 2 H 4 FE is 35.5% (Figure 2b; Figure S13a competitive reactions (such as the formation of H 2 and formate) on Cu needle-Ag are suppressed. These results imply that Cu needle-Ag can inhibit the production of H 2 and improve the selectivity to C 2 H 4 . Figure 2d depicts the total FE of C 2+ products on Cu needle and Cu needle-Ag at different potential. The Cu needle-Ag demonstrates a significantly improved selectivity for C 2+ products compared with the Cu needle at the full range of tested potentials. Remarkably, Cu needle-Ag displays FE of 64% for C 2+ products at −1.0 V, which is much higher than that of Cu needle (40.7%). The stability of Cu needle and Cu needle-Ag catalyst during electrocatalysis was further examined at −1.0 V for 10 h. As shown in Figure 2e, the change in FE of C 2 H 4 is <4% throughout the test.
The morphology change of Cu needle-Ag during electrocatalysis was also characterized by SEM. During the electrocatalytic process, the Cu needles in the Cu needle-Ag electrode gradually bend slightly, and the surface becomes rough, but the Cu still maintains a needle-like structure ( Figure S14, Supporting Information). EDS elemental mapping of Cu needle-Ag after catalysis reveals that the structure and the position of Ag NPs in Cu needle-Ag remain unchanged after electrocatalytic CO 2 reduction (Figure S15a-d, Supporting Information). The XRD pattern of Cu needle-Ag after catalysis further proves that the composition of the electrode also remains unchanged ( Figure S15e, Supporting Information). X-ray photoelectron spectroscopy (XPS) of the Cu needle-Ag also shows that there is oxide on the surface of Cu nanoneedles before the reaction ( Figure S15f,g, Supporting Information). After CO 2 RR, the presence of Cu 2+ may be attributed to the oxidation of Cu during sample transfer for XPS analysis. As displayed in Figure S14h (Supporting Information), the atomic percentage of Ag increases slightly relative to Cu after the reaction. This observation could be caused by the slight bending of the Cu needle after reaction. When preparing Cu needle-Ag electrode, Ag NPs can be deposited on the bottom of Cu needles along with the dispersing liquid. During the XPS testing, part of the Ag signal from Ag NPs will be blocked by the Cu needles above. After CO 2 RR, the Cu needles bend to a certain extent, allowing more Ag from the Ag NPs to be detected by the detector. Overall, the above results confirm that the composition of Cu needle-Ag catalyst is robust during the electrocatalytic process.
The electrochemical CO 2 RR performance was compared for tandem Cu needle-Ag electrodes with increasing mass loadings of Ag NPs ( Figure S16, Supporting Information). As shown in Figure 2f, the FE of C 2 H 4 increases from 33.5% on Cu:Ag = 1:1 to over 48% on Cu:Ag = 1:2 and maintains a comparable value when further increasing the Ag NPs loading to Cu:Ag = 1:3. However, when Ag loading is further increased to Cu:Ag = 1:4, the FE of C 2 H 4 decreases to 29.1%. Overall, Cu needle-Ag electrode exhibited the highest FE of C 2+ products when Cu:Ag = 1:2. The results suggest that the optimal mass ratio of Cu and Ag allows a good match of CO production and further CO conversion process, thereby helping to improve C 2+ conversion. To elucidate the effect of Cu needle density, the low-density Cu needle-Ag electrode was prepared using a low-density Cu(OH) 2 needle sample ( Figure S3a, Supporting Information) and tested under the same conditions. Compared with the high-density Cu needle electrode, the FE of CO of the low-density Cu needle electrode was significantly higher and the C 2+ product FE was lower ( Figure S17, Supporting Information). This is because the CO generated by the Ag NPs at the bottom of the low-density Cu needle is not sufficiently further reduced due to the CO diffusion process. The results suggest that a high density of Cu needle contributes to the enhanced selectivity of C 2+ products.
Cu needle-Ag electrode exhibits good C 2+ products selectivity in H-cell. However, the limitation of mass transfer in H-cells makes it difficult to obtain current densities approaching industrially relevant values (>200 mA cm −2 ). Therefore, Cu needle-Ag catalysts were further tested in a flow cell to improve the mass transfer of CO 2 for increased current density. (Figure S18, Supporting Information). A potential range from −0.8 to −1.1 V versus RHE was used in the H-cell testing since this is the typical potential range used for Cu-based catalyst testing. However, in the flow cell testing, we used a potential range from −0.6 to −1.1 V versus RHE in order to obtain performance of electrode at current densities below 200 mA cm −2 . As exhibited in Figure 2g, the Cu needle-Ag has high selectivity for C 2+ products at applied current densities between 200 and 350 mA cm −2 . Figure 2h shows the evolution of C 2+ products, and the FE of C 2+ products increase from 58.2 to 70% for current densities of 100 mA cm −2 and 350 mA cm −2 , respectively. C 2 H 4 and ethanol are the major C 2+ products, while small amounts of propanol and negligible acetic acid can be detected (Figures S19 and S20, Supporting Information). In addition, the FE of C 1 products drops from 25.9 to 10.9% when the current density increases from 100 to 350 mA cm −2 . However, at 400 mA cm −2 , Cu needle-Ag gives a lower FE of C 2+ and a higher FE of H 2 . Compared with the H-cell, Cu needle-Ag obtained better performance and higher current density in the flow cell. Here, the alkaline electrolyte used in the flow cell was beneficial to suppress hydrogen evolution reaction (HER). In addition, the shorter electrode distance and faster mass transfer rate of flow cell were beneficial to enhance performance and increase the current density. Encouragingly, among the reported state-of-theart Cu-based catalysts, the catalytic performance of our catalysts to C 2+ products is comparable or superior in both H-cell and flow cell systems (Figure 2i, Table S1, Supporting Information).

Investigation of Structural Effects
To elucidate the effect of electrode structure on the performance of the tandem catalyst, we further tested and compared nanoneedles to other morphologies of Cu-Ag tandem electrodes with varying structures. These comparisons revealed that the Cu needle-Ag offered the most favorable FE to C 2+ products. These comparable samples included Cu foil-Ag compared with Cu foil and Cu-Ag particles compared with Cu particles (Figure 3a). Figure S21a,b (Supporting Information) show an SEM image and typical XRD pattern of Cu foil. Cu foil-Ag electrode was prepared via a galvanic replacement reaction. SEM image and EDS element line scan mapping demonstrate that Ag particles were uniformly deposited on the surface of Cu foil ( Figure S21c,d, Supporting Information). In order to optimize the performance of the Cu foil-Ag electrode, the Cu foil was immersed in the AgNO 3 solution for different reaction times to obtain different amounts of Ag on the surface of the Cu foil. The CO 2 RR performance of Cu foil and the as-prepared Cu foil-Ag  Figure S22 (Supporting Information) shows the CO 2 RR performance of the Cu foil-Ag electrode obtained at different reaction times. All electrodes obtained the best catalytic performance at −1.0 V. However, when compared with Cu foil, Cu foil-Ag exhibits a lower FE of H 2 and a higher C 2+ FE. In addition, Cu foil-Ag obtains the highest C 2+ products FE when the galvanic replacement reaction is carried out for 30 s. Therefore, Cu foil-Ag with 30 s reaction time will be used for further discussion. Figure S23a,b (Supporting Information) demonstrates the SEM and TEM image of as-prepared pure Cu particles. The XRD pattern indicates no impurities in the as-prepared Cu particles ( Figure S23c, Supporting Information). The Cu-Ag particles electrode was prepared by mixing Cu NPs and Ag NPs and depositing them on hydrophobic carbon paper. SEM imaging and elemental mapping confirmed that Cu NPs and Ag NPs were fully mixed ( Figure S23d,e, Supporting Information). The XRD pattern confirmed that the asprepared electrode was composed of Cu and Ag ( Figure S18f, Supporting Information). Figure S24 (Supporting Information) shows the FE of products obtained on Cu particles and Cu-Ag particles with different ratio of Cu particles and Ag particles. Cu-Ag particles obtains the highest C 2+ products FE when the mass ratio of Cu particles and Ag particles is 1:2, which was consistent with other reported results. [53] Therefore, Cu-Ag particles electrode with mass ratio of Cu and Ag is 1:2 will be used for further discussion. Note that the performance of Cu particles and Cu-Ag particles is similar at low potential. At more reducing potentials of −1.0 and −1.1 V, Cu-Ag particles exhibit lower FE of H 2 , lower CO, and higher FE of CH 4 than Cu particles. Remarkably, the FE of C 2+ products obtained on Cu particles and Cu-Ag particles still remains similar, which means that mixing Ag particles with Cu particles promotes the formation of hydrocarbons, yet the promotion of C-C coupling is minimal. Note that different from Cu needle-Ag electrode, Cu-Ag particles and Cu foil-Ag electrodes had only trace detectable propanol, which was considered to be zero. Figure 3b,c shows the comparisons of the total FE for C 2+ products of Cu foil with Cu foil-Ag and Cu particles with Cu-Ag particles, respectively. Cu foil-Ag has more C 2+ products than Cu foil at most potentials. In contrast, Cu-Ag particles only have higher FE for C 2+ than Cu particles at −1.0 V. We select the best C 2+ products performance obtained by each electrode and compared Cu foil to Cu foil-Ag, Cu particles to Cu-Ag particles, and Cu nanoneedle to Cu needle-Ag respectively. As shown in Figure 3d, the C 2+ products of Cu foil-Ag are 19.4% higher than that of Cu foil. Compared to Cu particles, the C 2+ products of Cu-Ag particles are improved by 6%. Most importantly, compared to Cu nanoneedle, Cu needle-Ag shows up to 57.2% improvement.
To exclude the effect of the number of active sites, doublelayer capacitance measurements of the Cu foil, Cu foil-Ag, Cu particles, and Cu-Ag particles were used to determine ECSA of the electrodes (as shown in Figure S25a-e, Supporting Information). The ECSA value of Cu foil-Ag and Cu-Ag particles are 5.07 and 8.14, respectively ( Figure S25f, Supporting Information). To examine the increase in ECSA upon addition of the optimal loading of Ag particles, we have introduced a factor that is referred to as ECSA enhancement percentage. This reflects the increased ECSA as compared to the corresponding Cu-only catalyst. Hence, we have calculated the ECSA enhancement percentage for each of the three tandem electrode designs as compared with their corresponding bare Cu electrode design. Interestingly, the Cu-Ag particle sample exhibited the highest increase to ECSA upon adding Ag, but the Cu-Ag particle sample did not have the highest enhancement in faradaic efficiency of C 2+ products ( Figure S26a, Supporting Information). Figure S26b (Supporting Information) shows ECSA-normalized partial current densities of total C 2+ products over different electrodes. In this context, the Cu needle-Ag electrode still shows the highest ECSA-normalized C 2+ partial current density. Thus, the higher performance of the Cu needle-Ag structures should be ascribed to its unique structure.
In summary, the Cu-Ag particles electrode and the Cu needle-Ag electrode have the same mass ratio of Cu to Ag, but the Cu needle-Ag shows significantly better performance. These results confirm that the electrode structure has a significant effect on the performance of Cu-Ag tandem catalysts. On the tandem electrodes, the CO usually generated from Ag spills over to Cu and is further reduced to C 2+ products via CC coupling. In the Cu needle-Ag electrode, Ag NPs are deposited at the bottom of the Cu needles, and the needle structure subjects CO to a long diffusion path length and the coverage of *CO, which allows more CO to be subjected to further reduction. Thus, this structure allows the Cu needle-Ag electrode to obtain more C 2+ product.
To gain deeper insight into the role of structure on the C−C coupling mechanism, we conducted in situ Raman spectroscopy for two Cu-Ag catalysts with different structures. At open circuit potential, Cu 2 O at ≈526 and 628 cm −1 are measured for Cu-Ag particles and Cu needle-Ag (Figure 3e,f). [54] The Raman band at ≈2070 cm −1 is associated with the CO stretching vibrational mode of absorbed *CO on a Cu top site (*CO atop ). Importantly, Cu needle-Ag exhibited a higher *CO atop peak intensity compared to that of the Cu-Ag particles, indicating the higher coverage of *CO atop on the surface of Cu. The intensity of *CO signal decreases as the potential becomes more negative because the intermediate reacts rapidly to generate hydrocarbon products. [55,56] The additional peak at ≈1980 cm −1 on the Cu needles-Ag electrode indicates interaction between the absorbed CO* intermediates, which has been shown conducive to the formation of C 2+ molecules. [57,58] In contrast, Cu-Ag particles only show a single peak at ≈2070 cm −1 , which means that it is less favorable for CC coupling. In addition, oxygencontaining species are also present in the Raman spectra of Cu needle-Ag (dotted green lines at ≈320, ≈392, and ≈620 cm −1 ; these species do not have a direct correlation with CO 2 RR activity. [59,60] Figure S27 (Supporting Information) shows in situ Raman spectra of Cu particles and Cu needle. No prominent peaks of reaction intermediates were observed in their spectra, indicating that their CC coupling ability is relatively weak. In summary, the in situ Raman spectroscopic analysis indicates that Cu needle-Ag electrode promotes the adsorption of *CO intermediate and C 2+ product formation, in good agreement with the experimental results in an H-cell and a flow cell.
To investigate the effect of different structures on the CO 2 RR product distribution, we further apply finite-element method (FEM) calculations to track the critical species related to CO 2 RR over models consisting of Cu needle, Cu needle-Ag, and Cu-Ag particles. For Cu needle, a Cu needle with a length of 2 µm and a diameter of 160 nm was used to represent the Cu needle structure immersed in the electrolyte. For Cu needle-Ag, although the average diameter of the Ag NPs prepared in this work is ≈52 nm, the Ag NPs cannot be completely dispersed as single Ag NPs distributed at the bottom of the Cu needles when preparing Cu needle-Ag electrode. The Ag NPs deposited on the bottom of the Cu needle are Ag clusters with a diameter of ≈200 nm, and there are two Ag clusters on average in proximity to the bottom of each Cu needle (as shown in Figure S5, Supporting Information). Therefore, in the calculations, a Cu needle (same as Cu needle sample) with a length of 2 µm and two Ag particles with diameters 200 nm at the bottom of the Cu needle were used to represent Cu needle-Ag. In addition, randomly distributed Cu and Ag particles (number ratio 1:2) were used to represent Cu-Ag particles. The simulation of all samples shows the product distribution after the reaction reaches a stable state at t = 600 s.
The calculation results show that, for the Cu needle catalyst, the concentrations of C 1 and C 2 products are the highest on the surface of the Cu needles, and the product concentrations are lower in regions farther away from the Cu needles (Figure 4a,b). It should be noted that in the steady state, the product concentration of the Cu needle catalyst differs very little in space, so the values on the color bars are very similar. For the Cu needle-Ag catalyst, C 1 products are concentrated on the surface of Ag particles and diffused from the surface of Ag particles to Cu needles (Figure 4c), and they are subsequently reduced to C 2 products (Figure 4d). It can be observed that Cu needle-Ag has a greater C 2 product concentration in the whole region than the Cu needle. Figure 4e,f presents the distribution of C 1 and C 2 products of Cu-Ag particles catalyst, respectively. Specifically, the C 2 /C 1 ratio of Cu needle-Ag is 2.61 and 2.81 times higher than that of Cu needle and Cu-Ag particles, respectively (as shown in Figure S28, Supporting Information). We used the array model to study the concentration distribution of C 1 and C 2 products as a function of Cu needle length ( Figure S29, Supporting Information). The simulation results suggest that the ratio of C 2 /C 1 reaches a maximum value at a Cu needle length of 1.8 µm in this model system ( Figure S30, Supporting Information). Excessively long Cu needles reduce the ratio of C 2 /C 1 because the CO diffused from the Ag surface cannot fully diffuse to the entire surface of the Cu needles. In this case, the longer the Cu needles, the more similar the performance of Cu needle-Ag to Cu needle. In this work, the distance between the top of the Cu needle and the Ag particle is between 1.5 and 2 µm ( Figure S11, Supporting Information), which is consistent with the simulation results. Altogether, the simulation results suggest that depositing Ag NPs on the bottom of the Cu needle can effectively promote the further reduction of C 1 species, thereby improving the selectivity of C 2 species. In addition, these results provide insight into the design principles of tandem catalysts and reveal the importance of controlling the length of the C 2+ selective catalyst material.

Conclusion
In conclusion, we have successfully synthesized a 3D tandem catalyst electrode by a straightforward method. Compared to the same structure without Ag NPs at the bottom, the obtained Cu needle-Ag demonstrated significantly improved C 2+ products with an FE up to 64% in H-cell and 70% in flow cell with commercially relevant current densities. Compared with other tandem structure catalysts, as-prepared Cu needle-Ag shows a superior tandem improvement effect. Experimental studies combined with FEM calculations revealed that increasing the local concentration and CO diffusion path length contributes to the occurrence of CC coupling. Moreover, the improvement effect can be observed in various reactors. It is believed that the work reported here not only provides a new method for preparing tandem catalysts, but it also provides new guidance for designing tandem catalysts for CO 2 RR.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.