Plasmon-Enhanced C₂H₄ Generation in the CO₂ Electroreduction Reaction on a CuPd Tandem Catalyst

Abstract

Electrocatalysis enables the conversion of CO₂ into value-added fuels and chemicals, offering a sustainable solution for greenhouse gas mitigation. However, achieving high selectivity for C2 products like ethylene (C₂H₄) remains challenging due to competing C1 pathways and complex multielectron processes. Here, we demonstrate that plasmon resonances can selectively enhance the electroreduction of CO₂ to C₂H₄ by 27.0% on a CuPd catalyst under LED illumination (625 nm) at −1.3 V_RHE. Photocurrent response, in situ FTIR spectroscopy, and COMSOL simulations reveal that plasmon-derived hot electrons and heating greatly facilitate *CO formation at the CuPd interface, which diffuses to the Cu surface for subsequent C–C coupling. DFT calculations show that the increased *CO coverage on the Cu sites reduces the energy barrier for C–C coupling, ultimately enhancing C₂H₄ generation. This work offers valuable mechanistic insights into plasmon-mediated electrocatalysis, guiding the development of more efficient plasmonic tandem electrocatalysts for future carbon recycling technologies.

Introduction

Plasmon-enhanced electrocatalysis attracts much attention due to its ability to enhance the activity of electrocatalytic processes such as CO₂ electroreduction reaction (CO₂RR), − hydrogen evolution reaction (HER), − oxygen reduction reaction, , and more. − Plasmonic nanomaterials such as Cu, Ag, and Au exhibit localized surface plasmon resonances (LSPRs) in the optical (i.e., sunlight) regime. , The interaction between light and the free metal electrons concentrates light at the nanoscale, amplifying local electric fields by over 3 orders of magnitude. This phenomenon leads to absorption cross sections much larger than the physical size of the nanostructures, enabling efficient light absorption. Upon LSPR excitation, their energy can decay either through the scattering of resonant photons or by generating energetic charge carriers. These charge carriers can transfer energy to adsorbed molecules or dissipate by heating the metal lattice. The local electric field enhancement, hot carrier generation, and photothermal effect have been utilized in electrochemistry to accelerate chemical reactions (enhancing efficiency) or influence reaction pathways (improving selectivity). − Despite the potential of plasmonic effects in electrocatalysis, the field remains in its early stages, especially for CO₂RR. , Maximizing the use of all of the energy pathways derived from the plasmon decay is one of the greatest challenges in plasmonic catalysis.

To produce C2 compounds from CO₂RR, the reaction involves two key processes: the first is the reduction of CO₂ to the *CO intermediate (* represents intermediates absorbed on the catalyst surface), , and the second is C–C coupling. Copper (Cu), a plasmonic metal in the visible range of the electromagnetic spectrum, is capable of catalyzing C–C coupling reactions, − offering a promising avenue for plasmon-driven CO₂RR to generate valuable C2 chemicals and fuels. Additionally, a second metal (cocatalyst), which is favorable for *CO generation, such as Ag, − Au, − and Pd, − is often introduced to develop Cu-based bimetallic catalysts. For example, phase-separated CuPd nanoparticles (NPs) have been shown to enhance C₂ product formation, with *CO generated at Pd and CuPd interface sites and Cu sites catalyzing subsequent C–C coupling steps. The performance of such tandem catalysts for the CO₂RR upon plasmon excitation is a frontier topic.

In this study, tandem catalyst CuPd NPs were used to explore the role of plasmon resonances in enhancing the C₂H₄ formation during CO₂RR. Our results show that under LSPR excitation (625 nm LED) at −1.3 V_RHE, C₂H₄ production on CuPd NPs increased by 27.0% and CO production decreased by 11.4%, compared to the dark conditions. Through photocurrent response, COMSOL simulations, density functional theory (DFT) calculations, and in situ FTIR spectroscopy, we identified the mechanisms driving the selective C₂H₄ generation under plasmon excitation. The increase in C₂H₄ production is mainly attributed to plasmon-induced hot electrons and heating promoting *CO formation on CuPd interface sites. Subsequently, *CO diffusion from the CuPd interface to Cu sites results in increased *CO coverage on Cu sites, which decreases the energy barrier of C–C coupling and promotes C₂H₄ generation. Plasmon excitation also increased the H₂ production on CuPd NPs by 26.6%, primarily due to the photothermal effect. This study highlights the potential of plasmon resonances to improve the selectivity of high-value C2 products from the CO₂RR, presenting significant opportunities for exploration and innovation in the emerging field of tandem plasmonic catalysts.

Results and Discussion

Catalyst Preparation and Characterization

The CuPd, Cu, and Pd NPs were prepared by a wet chemical reduction method (see Materials and Methods in the Supporting Information). The synthesized CuPd, Cu, and Pd samples exhibit quasi-spherical morphology with average diameters of 15 and 55 nm, respectively (Figures S1 and S2). HAADF-STEM and EDS mapping reveal that the CuPd NPs consist of Pd NPs with multiple adsorbed Cu NPs (Figure a). The high-resolution TEM (HRTEM) images of CuPd NPs in Figure S2 show clear CuPd(111) and CuPd(100) interfaces, marked by the red lines.

1.

(a) HAADF-STEM images combined with EDS mapping of the CuPd sample. (b) LSV curves of CuPd, Cu, and Pd catalysts in dark and illumination conditions. (c) CO, (d) C₂H₄, and (e) H₂ enhancement of CuPd and Cu catalysts at different applied potentials under 625 nm LED illumination, compared to dark conditions.

The XRD patterns of CuPd and Cu NPs show characteristic peaks of Cu₂O before electrochemical reduction, which disappear after 1 h of reduction (Figure S3). Additionally, the XRD and XPS analyses reveal that CuPd NPs consist of independent Cu and Pd phases rather than a CuPd alloy (Figure S4). The CO₂-temperature-programmed desorption spectrum of CuPd NPs in Figure S5 shows three distinct desorption peaks, which can be assigned to Cu sites, CuPd interfacial sites, and Pd sites.

The UV–vis spectra in Figure S6 show that the synthesized CuPd NPs exhibit two distinct LSPRs at 280 and 605 nm, corresponding to Pd NPs and Cu NPs, respectively. For the subsequent light-illuminated CO₂RR measurements, the CuPd and Cu catalysts were illuminated with a 625 nm LED, to excite the LSPRs of Cu.

Plasmon-Enhanced CO₂RR Performance

The linear sweep voltammogram (LSV) curves (Figure b) and Tafel slope (Figure S8) show increased reaction kinetics for CuPd and Cu NPs under 625 nm illumination, while the reaction kinetics on Pd NPs remain unchanged due to their low light absorption cross-section. Electrochemical CO₂RR experiments were performed sequentially under dark and light illumination conditions. To highlight the plasmon-induced contributions to the CO₂RR on CuPd and Cu NPs, we calculated the enhancement in the production yield between the illumination and dark conditions for each product using the following formula:

$$Enhancement(\%)=\frac{{Light}_{(ppm)}-{Dark}_{(ppm)}}{{Dark}_{(ppm)}}$$

As a key intermediate for C2 generation, *CO can either be released from the catalyst surface to form CO gas or undergo C–C coupling to produce C2 products. , As shown in Figures c and d, CO gas production increased by 9.8% and that of C₂H₄ increased by 8.6% on the Cu catalyst at −1.3 V_RHE under illumination, suggesting that plasmon excitation promoted *CO generation on Cu NPs. Some of the *CO desorbed from the Cu surface, while some participated in the C–C coupling step, generating C2 products. In contrast, on the CuPd catalyst at −1.3 V_RHE under illumination, the CO gas generation decreased by 11.4% and C₂H₄ generation increased by 27.0%, indicating that the plasmon excitation promoted *CO generation and selectively facilitated its consumption through C–C coupling rather than desorption from the CuPd surface.

We also monitored the HER, an important competitive reaction for CO₂RR. The H₂ yield increased by 51.4% on Cu NPs and by 26.6% on CuPd NPs at −1.3 V_RHE under illumination, compared to dark conditions (Figure e).

The enhancement of other products, along with variations in total current density, Faradaic efficiency for CuPd, Cu, and Pd catalysts, and stability, as well as the EDS mapping images after reaction, are presented in Figures S9–S14. These results indicate that the CuPd catalyst favors plasmon-driven C₂H₄ generation together with an increase in H₂ generation. In the following section, we explore the mechanisms of plasmon-driven C₂H₄ and H₂ generation on CuPd NPs.

Plasmon-Induced Hot Electrons and Heating

As mentioned in the Introduction, plasmonic excitation generates hot carriers and heats the metal lattice of the nanostructures. The photocurrent response experiments have been reported to quantify the contributions of hot electrons and the photothermal effect. As shown schematically in Figure a, the photocurrent response exhibits a two-stage increase: a rapid increase in the initial 0.05 s due to hot electrons, followed by a slower increase from 0.05 to 10 s due to the photothermal effect. Here, I _0.05 denotes the photocurrent in the initial 0.05 s, I _0.05–10 represents the photocurrent from 0.05 to 10 s, and I ₁₀ corresponds to the total photocurrent over 10 s. The contributions of plasmon-induced hot electrons (C _{hot electrons} ) and the photothermal effect (C _photothermal ) can be quantified using the following equations:

$$C_{hotelectrons}=\frac{I_{0.05}}{I_{10}}$$

$$C_{photothermal}=\frac{I_{0.05-10}}{I_{10}}$$

2.

(a) Schematic of evaluating the contributions of hot electrons and photothermal effects based on the photocurrent response. (b) Photocurrent responses of Ar-saturated and CO₂-saturated 0.1 M KHCO₃ upon 625 nm light chopped in every 10 s on the CuPd NP electrode at −1.3 V_RHE. (c) The yields of H₂, CO, and C₂H₄ at different applied potentials on the CuPd NP electrode at 16, 26, and 36 °C. (d) The thermodynamic energy barrier of C–C coupling on Cu sites with different CO* coverage.

The photocurrent response experiment was carried out on CuPd and Cu NPs under chopped light every 10 s at −1.3 V_RHE in Ar- and CO₂-saturated electrolyte (Figures S15, S16, and b). In an Ar-saturated electrolyte at −1.3 V_RHE where only HER occurs, the photothermal effect accounts for 80% of the total photocurrent on CuPd NPs and 76% on Cu NPs, suggesting that the photothermal effect mainly drives H₂ generation under light illumination. In the CO₂-saturated electrolyte at −1.3 V_RHE where both the CO₂RR and HER occur (CO₂RR dominates), the contribution of hot electrons to the photocurrent increased significantly to 42% on CuPd NPs and 28% on Cu NPs, suggesting that both hot electrons and the photothermal effect contribute to the enhancement of the CO₂RR on CuPd NPs. In contrast, for the CO₂RR on Cu NPs, the photothermal effect plays a dominant role compared to hot electrons. This is also supported by the greater increase in H₂ production on Cu NPs compared to CuPd NPs under illumination (Figure e).

Plasmon-induced hot electrons are closely associated with local electric field enhancement on the surface of plasmonic nanomaterials, where stronger near-fields facilitate hot electron generation. , Therefore, COMSOL simulations were employed to model the electric field distributions on CuPd, Cu, and Pd NPs under 625 nm illumination (Figure S17). The simulations reveal that the Cu–Pd interface exhibits the strongest local electric field enhancement.

Consistent with the electric field distribution, the plasmon-induced hot electrons are also mainly localized at the CuPd interface, as shown in the COMSOL simulations (Figure S18). This is also reasonable from the perspective of the work function of the two metals (Figure S19): the larger work function of Pd compared to Cu drives the diffusion of plasmon-excited hot electrons from the Cu side to the CuPd interface. Notably, hot electrons were reported to dominate the reduction of CO₂ to CO, significantly increasing CO generation. , Taken together with the experimental results from Figures and , these results suggest that hot electrons localized at the CuPd interface enhance *CO generation over H₂ generation.

Following the generation of hot electrons at the CuPd interface, these high-energy carriers dissipate their energy through scattering processes in approximately 1–10 ps, resulting in localized heating and rapid diffusion across the entire metal surface. Photocurrent measurements confirm that the photothermal effect enhances both H₂ evolution and CO₂RR. Plasmon-induced heating has been reported to facilitate CO₂ activation and *CO generation. ,, External heating from 16 to 36 °C (Figure c), closely matching the temperature rise reported for bimetallic plasmonic systems under illumination, led to increased H₂ and CO production but reduced C₂H₄ formation at −1.3 V_RHE for CuPd NPs. This is likely due to reduced CO solubility at higher temperature of the electrolyte, which promotes CO release and hinders C–C coupling. In contrast, the unique advantage of plasmonic heating is its confinement to the catalyst surface without affecting the bulk electrolyte temperature. Indeed, we did not detect changes in the electrolyte temperature under illumination. The increased surface temperature under illumination enhances *CO formation on the surface of CuPd NPs, thereby offering a greater potential for subsequent C–C coupling steps.

The results above demonstrate that the role of plasmon excitation for CO₂RR is 2-fold: hot electrons facilitate *CO formation on CuPd interfacial sites and the photothermal effect promotes *CO formation on the whole surface of CuPd NPs. DFT calculations show that increased *CO coverage on CuPd interfacial sites and Pd sites weakens *CO adsorption (Figures S20–S23), thereby facilitating its diffusion to the Cu sites. The accumulation of *CO on Cu sites lowers the energy barrier of C–C coupling (Figures d and S24–S26), ultimately leading to significantly increased C₂H₄ generation on CuPd NPs.

ATR-FTIR Measurements

The in situ ATR-FTIR measurements, which can track the surface dynamics of the reaction intermediate, were carried out to track *COOH and *CO intermediates on CuPd, Cu, and Pd NPs under dark and illumination conditions (Figures a–d and S27–S29). The tentative band assignment was summarized in Table S1. The vibrational bands in the range 1285–1250 cm^–1 are attributed to the functional groups of *COOH on CuPd, Cu, and Pd NPs (Figures a–b and S28). − We observed a rapid increase of the band area of *COOH on CuPd NPs with 625 nm light excitation compared to dark conditions (Figures a, b, and S28, and S29), suggesting plasmon-enhanced CO₂ activation.

3.

in situ FTIR spectra on the CuPd NP electrode in CO₂-saturated 0.1 M KHCO₃ electrolyte at −1.3 V_RHE under dark (a and c) illumination (b and d) for 10 min.

Next, we analyzed the *CO vibration bands (2100–2000 cm^–1) on the CuPd, Cu, and Pd catalyst under dark and illumination conditions (Figures c, d, and S28). In dark and illumination conditions, the *CO stretching bands on Cu and Pd NPs appear near 2071 cm^–1 and 2094 cm^–1, respectively (Figure S28), in line with previous reports. − On CuPd NPs in dark conditions, the *CO stretching band appears near 2024 cm^–1, which deviates from the *CO stretching band on both Cu (2071 cm^–1) and Pd NPs (2094 cm^–1). The band near 2024 cm^–1 could be attributed to the *CO being adsorbed on CuPd interface sites, similar to previous reports. Within ∼2 min of light illumination, the vibrational bands at 2017 cm^–1, attributed to *CO adsorbed on CuPd interfacial sites, gradually decreased, while a new band emerged at 2075 cm^–1 (Figure d). This band closely matches the linear *CO stretching band on the Cu surface, indicating the diffusion of *CO from the CuPd interface sites to the Cu surface.

Mechanism of Plasmon-Enhanced CO₂RR

Considering all the experimental and theoretical results, we can paint the complete image of CO₂RR on CuPd NPs in the dark, with external heating, and with light excitation (Figure ). In the dark, CO₂ is converted to *CO at the Pd and CuPd interface sites, and the generated *CO diffuses to the Cu sites, where it undergoes C–C coupling to form C₂H₄.

4.

Schematic diagram of CO₂RR on the CuPd catalyst under dark, with external heating, and with illumination.

Under external heating, the CO solubility decreases with an increase in the electrolyte temperature, thereby yielding more CO gas but less C₂H₄.

Under light excitation of CuPd NPs, the plasmon-generated hot electrons, which are mainly localized at the CuPd interface, accelerate the formation of *CO. Meanwhile, plasmon-induced heating accelerates the CO₂RR and facilitates the *CO formation on CuPd NPs. Subsequently, the *CO generated on Pd and CuPd interfacial sites diffuses to Cu sites. The increased *CO coverage on the Cu sites lowers the energy barrier of C–C coupling, ultimately promoting C₂H₄ generation. Simultaneously, plasmon-induced heating also enhances H₂ production.

Conclusion

In summary, we synthesized CuPd NPs featuring distinct Cu, Pd, and Cu–Pd interfacial sites and demonstrated that the synergy among the Cu–Pd interface, plasmon-induced hot carriers, and photothermal effects plays a pivotal role in boosting C₂H₄ production under light illumination. Under 625 nm LED illumination at −1.3 V_RHE, the CuPd NP electrode exhibited a 27.0% increase in C₂H₄ production, accompanied by an 11.4% decrease in CO and a 26.6% increase in H₂ generation compared to dark conditions. In contrast, Cu NPs under the same illumination showed only an 8.6% enhancement in C₂H₄ production, alongside more pronounced increases in H₂ (51.4%) and CO (9.8%), highlighting the unique selectivity of the CuPd interface.

Photocurrent measurements and COMSOL simulations revealed that plasmon decay at the Cu–Pd interface facilitates *CO formation via hot electron injection and photothermal heating. DFT calculations further showed that the elevated *CO coverage at the Cu–Pd interface weakens *CO adsorption, enhancing its diffusion to neighboring Cu sites. This increased *CO availability on the Cu sites lowers the energy barrier for C–C coupling, thereby promoting ethylene formation. These mechanistic insights are corroborated by in situ FTIR spectroscopy. Furthermore, our analysis indicates that photothermal heating is the dominant driver of the enhanced H₂ evolution under illumination, as supported by photocurrent response data.

Overall, this work highlights the powerful role of plasmonic effectsnot only in improving reaction rates but also in steering product selectivity in CO₂ electroreduction. Our findings underscore the importance of rational catalyst design in harnessing plasmonic excitation to selectively control reaction pathways, opening new avenues for developing next-generation tandem plasmonic electrocatalysts.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacs.5c10517.

• Additional experimental details, materials, and methods, including schemes of the experimental setups and computational details; TEM, XRD, XPS, TPD, and UV–vis characterization for all catalysts; additional CO₂ electroreduction efficiencies for all catalysts; photocurrent response measurements, COMSOL simulations, DFT calculations, and in situ FTIR spectra for all catalysts (PDF)

The authors declare no competing financial interest.