Synergistic regulation of hydrophobicity and basicity for copper hydroxide-derived copper to promote the CO2 electroreduction reaction

Abstract

A simple method was proposed to activate alkaline Cu(OH)₂ with an acidic ionomer, Nafion, to regulate its surface microenvironment, including hydrophobicity and local basicity. In particular, the direct complete neutralization reaction between Cu(OH)₂ and Nafion in aqueous solution induces the exposing of vast anions which can exclude the in-situ-formed hydroxides and raise the local basicity. Remarkably, the optimal Nafion-activated Cu(OH)₂-derived Cu can efficiently suppress the hydrogen evolution reaction (HER) and improve the selectivity for multi-carbon products in the CO₂ electroreduction reaction (eCO₂RR). The H₂ Faradaic efficiency (FE) decreased to 11% at a current density of 300 mA/cm² (−0.76 V vs. RHE) in a flow cell, while the bare one with H₂ had an FE of 40%. The total eCO₂RR FE reaches as high as 83%, along with an evidently increased C₂H₄ FE of 44% as compared with the bare one (24%), and good stability (8000 s), surpassing that of most of the reported Cu(OH)₂-derived Cu. The experimental and theoretical results both show that the strong hydrophobicity and high local basicity jointly boosted the eCO₂RR as acquired by felicitously introducing ionomer on the Cu(OH)₂-derived Cu surface.

INTRODUCTION

With the rapid growth of the global economy and excessive consumption of traditional fossil fuels, CO₂ gas has been increasingly accumulating in the atmosphere, resulting in the intractable global greenhouse effect.^1–3 Practically, CO₂ is an economic, safe, and sustainable chemical feedstock, which has great potential to be transformed into high-value-added fuels and chemicals.^4–6 Nevertheless, CO₂ is a completely oxidized and thermodynamically stable molecule;⁷ suitable technologies are particularly required for CO₂ transformation to maximize energy efficiency and decrease the cost.⁸ It is noteworthy that the CO₂ electroreduction reaction (eCO₂RR) represents one of the optimal pathways to convert CO₂ into hydrocarbons, oxygenates, and so forth, with suitable electrocatalysts.^9–12

To date, copper (Cu)-based catalysts are regarded as the most favorable for achieving multi-carbon products (ethylene, ethanol, acetic acid, etc.).¹³ In particular, CuO, Cu₂O, and Cu(OH)₂-derived Cu have been verified to have excellent selectivity for multi-carbon products because of their abundant active sites for facilitating C–C coupling, such as grain boundaries,^14,15 and oxygen vacancies.^16,17 Among them, Cu(OH)₂ is a promising positively charged Cu precursor for eCO₂RR due to its low price, simple synthesis method, and easily controllable structure.^18–20 However, since they only work in H-type cells due to their hydrophilic property, most of the Cu(OH)₂ electrodes yield unremarkable activity toward eCO₂RR, along with a relatively high Faradaic efficiency (FE) for the hydrogen evolution reaction (HER).^19,21–23 Therefore, it is imperative to further boost the catalytic performance of eCO₂RR for Cu(OH)₂-derived Cu.

Recently, owing to their ability to overcome the mass transfer limitation and high reaction rate, flow cells have been extensively used to boost electrocatalytic performance by using advanced electrocatalysts^24,25 and appropriate reaction microenvironments.^26,27 To smoothly run flow cells and suppress HER, the hydrophobic catalytic surface is a prerequisite for the construction of a stable three-phase interface. The hydrophobic property of a catalyst can be achieved by modification with some hydrophobic macromolecules,^28,29 such as Nafion-60-modified CuO,³⁰ PTFE-modified Cu/C,³¹ and 1-octadecanethiol-treated Cu dendrites.³² It is noteworthy that ionomers, such as Nafion, Sustainion, and so forth, both with hydrophobic and hydrophilic functional groups, have been commonly used to bind electrocatalysts and obtain a hydrophobic surface,^30,33 while less attention has been paid to determining the effect of their acidity or basicity on the local pH. As is well known, Nafion is the most extensively used binder in the electrocatalytic field, but it contains sulfonic acid groups (Figure S1A), which can confer the catalysts with a local acidic microenvironment and is beneficial for promoting HER.^34,35 Therefore, it is very important to pre-activate Nafion-modified electrocatalysts by neutralizing their acidic group with suitable alkaline substances to suppress HER.

Herein, to activate the hydrophilic and alkaline Cu(OH)₂ available in a flow cell, the ionomer, Nafion, was chosen to improve the hydrophobicity and simultaneously increase the local basicity, which is realized through direct and thorough deprotonation of Nafion by Cu(OH)₂ in an aqueous solution. The obtained optimal Cu(OH)₂-derived Cu, modified with 28.4 wt% of Nafion (denoted [email protected] wt%), successfully suppressed HER with H₂ FE of 11% at a current density of 300 mA/cm² (−0.76 V vs. RHE), while the bare AN-Cu(OH)₂-derived Cu electrode had H₂ FE of 40%. Also, the total eCO₂RR FE reached as high as 83%, along with an evidently increased C₂H₄ FE of 44% as compared to the bare one (24%), and good stability (8000 s). Moreover, the Nafion-activated strategy could be extended to other types of Cu(OH)₂ precursors prepared using other methods, while the substitution of Nafion with a kind of neutral ion exchange ionomer, Sustainion (Figure S1B), could not yield a better result.

RESULTS AND DISCUSSION Characterization of AN-Cu@Nafion

AN-Cu(OH)₂ was prepared using a simple one-step wet chemical method by using ammonia and copper nitrate as precursors.³⁶ The obtained AN-Cu(OH)₂ yields a nanorod morphology with various nanopores on their surface as observed in the transmission electron microscopy (TEM) image (Figure S2A). The high phase purity of AN-Cu(OH)₂ was confirmed from the high-resolution transmission electron microscopy (HRTEM) image (Figure S2B) and X-ray diffraction (XRD) analysis (Figure S2C). From the Brunauer–Emmett–Teller (BET) and Barrett–Joyner–Halenda (BJH) pore size distribution measurements of AN-Cu(OH)₂, the surface area was estimated to be 29.7 m² g⁻¹ and the pore size distribution ranged from mesopores to macropores, with an average pore size of ∼3.1 nm, further confirming the existence of abundant nanopores (Figure S2D).³⁷ Then, the AN-Cu(OH)@Nafion samples were prepared by mixing AN-Cu(OH)₂ with different contents of Nafion. The contact angle measurements show that the AN-Cu(OH)₂ surface changes from hydrophilic to hydrophobic on changing the Nafion content from 0 wt% to 41.6 wt% (Figure 1A). Also, the contact angle basically remained at ~140° when the Nafion content was above 28.4 wt%. Therefore, we chose the AN-Cu(OH)₂@Nafion-28.4 wt% sample for the following characterization. AN-Cu(OH)₂@Nafion-28.4 wt% retains the nanorod morphology, as shown in the TEM image (Figure S3A). The atomic structure of AN-Cu(OH)₂@Nafion-28.4 wt% was further examined using HRTEM (Figure S3B). Both AN-Cu(OH)₂ and Nafion domains could be clearly observed, in which the amorphous Nafion layer was unevenly attached to the AN-Cu(OH)₂ nanorods' edge surface, revealing the successful introduction of Nafion.^32,38

Figure 1. (A) Photographs of contact angle measurements on AN-Cu(OH)2 with different percentages of Nafion. (B) SEM, (C) TEM, (D) HRTEM, (E) HAADF-STEM, and (F) corresponding elemental mapping images of AN-Cu(OH)[email protected] wt% after electroreduction at 10 mA/cm2. The inset in (B) shows a contact angle photograph of AN-Cu(OH)[email protected] wt% after electroreduction at 10 mA/cm2.

Before catalysis of eCO₂RR, AN-Cu(OH)₂@Nafion-28.4 wt% was electroreduced to form hydroxide-derived Cu, that is, [email protected] wt%, at a current density of 10 mA/cm² for 10 min. The nanorod morphology with nanopores on the surface is preserved (Figure 1B,C), while the nanorods cross-linked to form a network on the gas diffusion layer (GDL), which might be owing to the bubble effect due to the gaseous products of eCO₂RR. The contact angle remained 116.6° (inset in Figure 1B), indicating that [email protected] wt% still has an excellent hydrophobic surface, which is very important for constructing a stable three-phase interface to promote eCO₂RR.³¹ The HRTEM image shows that the Cu(111) lattice fringe was highly distributed on the sample,^39,40 along with some CuO species on the subsurface due to the oxidation of Cu in the atmosphere (Figure 1D).³⁶ The Nafion layer was randomly distributed along the edge surface, which is benefiting for exposing the inside Cu active sites. The HAADF-STEM images and the corresponding elemental mapping images (Figure 1E,F) show the even distribution of Cu, F, O, and S throughout the entire [email protected] wt% sample, suggesting that Nafion not only was dispersed on the surface but also entered the nanopores in AN-Cu. This further proves that Nafion was securely retained in the AN-Cu sample after the electroreduction treatment. The XRD analysis also shows the main peak corresponding to Cu(111) (Figure 2A), indicating that Cu mainly exists in the metallic form in the sample.

Figure 2. (A) XRD pattern of [email protected] wt%. (B) Cu 2p, (C) O 1s, and (D) S 1s XPS spectra of AN-Cu (above row) and [email protected]% (bottom row).

The valence state of [email protected] wt% was further analyzed by X-ray photoelectron spectroscopy (XPS). As expected, the survey XPS spectrum of [email protected] wt% shows the coexistence of Cu, F, O, and S, while AN-Cu lacks the signal of S (Figure S4A). The fine XPS spectra of Cu 2p show the presence of mainly metallic Cu at 932.5 eV and 952.5 eV for [email protected] wt% and AN-Cu (Figure 2B).⁴¹ The O 1s XPS spectra of both [email protected] wt% and AN-Cu show a peak centered at 531.8 eV, which can be assigned to the oxygen defect due to the electroreduction treatment.^42,43 [email protected] wt% also shows an extra band centered at 535.4 eV, which could be ascribed to the oxygen atoms in -SO₃ (Figure 2C).³⁰ Furthermore, the S 1s XPS spectrum of [email protected] wt% shows the typical signals from the sulfonate group of Nafion (Figure 2D),⁴⁴ clearly confirming the introduction of Nafion. A similar phenomenon was observed in the fine spectra of F 1s XPS spectra for these two samples (Figure S4B), which could have resulted from the presence of the F element of poly(tetrafluoroethylene) from the GDL substrate.⁴⁵

eCO₂RR activity of AN-Cu@Nafion in a flow cell

The CO₂ electroreduction activities of AN-Cu@Nafion were evaluated using a three-compartment flow electrolyzer as shown in Figure S5. The catalyst loading was first screened and 0.2 mg/cm² of [email protected] wt% on GDL was found to show the best performance for eCO₂RR (Figure S6). Then, the influence of Nafion content on eCO₂RR performance was investigated through modifying AN-Cu with 0, 0.07, 19.2, 28.4, and 41.6 wt%. Though the R_ct (charge-transfer resistance) increases slightly with increasing Nafion content, the adhesion of AN-Cu@Nafion samples on GDL could accelerate electron transfer as compared to the bare GDL on the electrochemical impedance spectra (EIS, Figure 3A).⁴⁶ The electrochemical surface areas (ECSAs) of those electrodes were thereafter estimated by double-layer capacitance (C_dl), which is linearly proportional to the ECSA.¹ As expected, the ECSA evidently decreases after modifying of Nafion to the AN-Cu (Figure S7), suggesting a decrease in the total electrocatalytic intrinsic activity. To make the AN-Cu work available in the flow cell, it is accepted to partly sacrifice some active sites. The eCO₂RR performance of these electrodes was determined at a current density of 300 mA/cm², as shown in Figure 3B. The competing HER can be significantly suppressed from FE 40% (Nafion: 0 wt%) to 11% (Nafion: 28.4 wt%). On continually increasing the mass percentage of Nafion to 41.6 wt%, the production of H₂ begins to recover, which could be due to the excessive loss of active sites as mentioned above in the ECSA analysis (Figure S7). Among all the screening Nafion percentage electrodes, the [email protected] wt% electrode shows the highest total eCO₂RR FE of 83% for monoxide (CO), methane (CH₄), ethylene (C₂H₄), ethanol (EtOH), acetate (AcO⁻), n-propanol (PrOH), and formate (HCOO⁻). Also, C₂H₄ FE reaches 44%, which surpasses the value of most of the Cu(OH)₂-derived Cu and is comparable to that of the state-of-the-art electrocatalysts (Table S1).^{18–20,22,25,47–59} Therefore, the Nafion mass percentage of 28.4 wt% was chosen for the next eCO₂RR investigation.

Figure 3. (A) Nyquist plots of AN-Cu@Nafion catalysts with different Nafion percentages constructed at −0.1 V versus RHE from 0.01 Hz to 100 kHz and (B) product distribution FEs of AN-Cu@Nafion catalysts with different percentages of Nafion at 300 mA/cm2. (C) Chronopotentiometry curves at various current densities for CO2 reduction of [email protected] wt% in a 1 M KOH electrolyte, and (D) corresponding product distribution FEs, (E) voltage profile and C2H4 FEs, and (F) product distribution FEs at a constant current density of 300 mA/cm2 in an 8000 s test.

To examine the stability of the [email protected] wt% electrode, the eCO₂RR activity was first tested at different current densities on one electrode. The chronopotentiometry curves could be easily and stably recorded from 50 to 500 mA/cm² without the problem of flooding, even with some fluctuation at 500 mA/cm² due to the gas accumulation at a high reaction rate (Figure 3C). The product distribution in Figure 3D shows that the FE of HER was suppressed to ~10% at all these current densities. On increasing the current densities, the FE of the total products surpasses 70%, with the highest value of 83% at 300 mA/cm² (−0.76 V vs. RHE), CO decreases from 54.5% to 10.2%, and FE of C₂H₄ shows a volcano trend with the highest FE of 44% at 300 mA/cm², indicating the smooth conversion of CO₂. Under a constant current density of 300 mA/cm² over 8000 s, the potential profile is very stable without any obvious change, the C₂H₄ FE remains steady at ~45% (Figure 3E), and the product distributions are almost identical (Figure 3F). The morphology, composition, and crystallinity of the [email protected] wt% electrode were preserved, as determined from the postreaction SEM, TEM, elemental mapping, XRD, and XPS analyses (Figures S8 and S9).

Furthermore, the same phenomena of suppression of HER and boosting of eCO₂RR are also found for the other two kinds of Cu(OH)₂ samples prepared using different methods (Figure S10), but not for the other common Cu-based electrocatalysts (Figure 4A), including CuO (Figure S11A,B), commercial Cu(Ac)₂ (Figure S11C,D), Cu NPs (Figure S12A,B), and Cu₂(OH)₂CO₃ (Figure S12C,D). These results indicate that the activation effect of ionomer  for eCO₂RR is selectively working on Cu(OH)₂-derived Cu, excluding the case that Nafion itself drops the HER intrinsic activity. Besides, substitution of Nafion with another neutral anion ionomer, Sustainion, to activate AN-Cu(OH)₂ did not yield a better performance for the eCO₂RR (Figure 4B). To trace the origin of these results, some control experiments were performed, including contact angle and bulk pH measurements. As shown in Figures 1A and S13, the modification of Nafion with 28.4 wt% on all the tested Cu-based samples led to the development of a hydrophobic surface, which was also found for the AN-Cu(OH)₂@Sustainion-28.4 wt% electrode (Figure S14). It follows that a hydrophobic surface alone is not enough to promote eCO₂RR. Intriguingly, the bulk pH of the mixture of alkaline Cu(OH)₂ and acidic Nafion can be sensitively tuned by adjusting their mass proportion (Figure 4C and Table S2), indicating the occurrence of direct acid–base neutralization between them and the thorough removal of Nafion's proton. And, the electrodes with enhancement of eCO₂RR performance are with their bulk ink pH value higher than 7 for the AN-Cu(OH)₂ with different percentage of Nafion (Figure 4C) and the other two kinds of Cu(OH)₂ samples with 28.4 wt% of Nafion (Figure 4D). Hence, this can be assigned to the negative background charge in Nafion causing the phenomenon of Donnan exclusion to increase the local basicity at the catalyst surface^33,60 and thus suppress their HER.

Figure 4. (A) Product distribution FEs of various catalysts@Nafion for eCO2RR and (B) AN-Cu(OH)2@different ionomers. The pH of bulk (C) AN-Cu(OH)2@Nafion ink with different percentages of Nafion, (D) various [email protected] wt% ink, and (E) AN-Cu(OH)2@different ionomer ink. AN-Cu(OH)2: Cu(OH)2 was produced on adding NH3·H2O and NaOH; N-Cu(OH)2: Cu(OH)2 was prepared by eliminating NH3·H2O; AS-Cu(OH)2: Cu(OH)2 was obtained by replacing Cu(NO3)2 with CuSO4.

Furthermore, as a typical anion-conducting ionomer, Sustainion, a polystyrene vinylbenzyl methylimidazolium (Figure S1B), can efficiently remove the local OH⁻ on the catalyst surface and decrease the local pH.³³ As it can also be used to improve the hydrophobicity, Sustainion was chosen as a control ionomer to study the effect of local basicity on eCO₂RR. As expected, though the bulk pH of Sustanion modification is similar to that of Nafion on AN-Cu(OH)₂ (Figure 4E), the HER could not be suppressed to a high degree as that with Nafion. This confirms that the high local basicity caused by the deprotonated Nafion layer on Cu boosted the performance of eCO₂RR.

To understand further the influence of Nafion on Cu(OH)₂-derived Cu(111) surfaces (Figures 1D and 2A), density functional theory (DFT) calculations were conducted and the calculation methods are described in the Supporting Information. The introduction of deprotonated Nafion on the surface of Cu(111) could stabilize the in-situ-formed OH⁻, which could be due to the formation of hydrogen bonds with oxygen from Nafion²⁶ and the Donnan exclusion (Figure 5A).³³ Consequently, the adsorption of in-situ-formed OH⁻ on the surface of Nafion-Cu(111) (−1.9578 eV) becomes more energetically favorable than that on Cu(111) (−0.8969 eV) (Figure 5B), which helps to suppress HER through resisting water molecules.²⁶

Figure 5. DFT calculation of the eCO2RR pathway to C2H4 on [email protected] wt%-derived Cu surfaces. Adsorption energy of OH− on the (A) Nafion-Cu(111) surface and (B) the Cu(111) surface. (C) DFT-calculated Gibbs free energy of the intermediates on Nafion-Cu(111) (red) and Cu(111) (black) surfaces.

Since the formation of C₂H₄ is significantly promoted by the addition of Nafion from the experimental result (Figure 3B), the pathway for achieving C₂H₄ was recorded and is shown in Figure 5C. According to the general mechanism of COCO* coupling,^61,62 the primary path went through in the step of CO₂ → CO₂* → COOH* → CO* → COCO* → COCOH* → COC* → COHC* → CC* → CCH* → CHCH* → CHCH₂* → CH₂CH₂* → CH₂CH₂ (Figures S15 and S16). The free-energy change of every step for the C₂H₄ production on Nafion-Cu(111) is lower than that of Cu(111), indicating that the electroreduction of CO₂ to C₂H₄ on the surface of Nafion-Cu(111) is overall thermodynamically favorable. It is noteworthy that the energy barrier of *CO dimerization as the rate-limiting step changes by 0.75 eV for Nafion-Cu(111), which is much lower than that of Cu(111) (ΔG = 1.40 eV), suggesting that the introduction of deprotonated Nafion could facilitate the C–C coupling process. Therefore, the subsequent reactions proceeded smoothly, including carbon hydrogenation, deoxidation, and dehydration, and ultimately C₂H₄ desorption with enhanced selectivity, which is in good agreement with the experimental results.

CONCLUSION

In summary, we proposed a simple strategy for realizing Cu(OH)₂-derived Cu as an efficient eCO₂RR catalyst in a flow cell by activating it with an acidic ionomer of Nafion. The optimal eCO₂RR performance of [email protected] wt% can be achieved in a flow cell at  300 mA/cm² with a C₂H₄ FE of 44% and total CO₂ reduction FE of 83%. This strategy is also demonstrated with good compatibility for the scope of Cu(OH)₂-based materials. The construction of a hydrophobic microenvironment and increasing of local basicity jointly work for suppressing HER and promoting the C₂H₄ formation by exposing vast of deprotonated Nafion on Cu. This work synergistically regulates two kind of catalyst microenvironment factors to boost eCO₂RR, which will spur interests toward more efficient catalysts for a wide range of electrolysis technologies.

ACKNOWLEDGMENTS

This work was financially supported by the National Natural Science Foundation of China (51872209, 52072273, 21972126, 52201227), the Zhejiang Provincial Special Support Program for High-level Talents (2019R52042), and the Key Project of Zhejiang Provincial Natural Science Foundation (LZ20B030001).

CONFLICTS OF INTEREST

The authors declare no conflicts of interest.