Direct atmospheric pressure plasma jet (APPJ) synthesis of nanosized CuO_x–Ag composites for efficient electrochemical CO₂ reduction to multi-carbon products

Abstract

An atmospheric pressure plasma jet (APPJ) is an emerging technique capable of synthesising ligand-free nanosized composites with controllable configuration ratios. Here, an APPJ is used to synthesise and directly deposit CuO_x–Ag for the electrocatalytic CO₂ reduction reaction (CO₂RR). The morphology of CuO_x–Ag composites evolves from Janus-type to core–shell with higher Cu : Ag precursor ratios. When applied in CO₂RR catalysis, the nanoparticle configuration appears to matter more than the exact Cu : Ag composition. The core–shell arrangement is found to exhibit higher C₂₊ production than the Janus-type structure, attributed to its ability to retain a larger Cu–Ag interfacial area during the CO₂RR. Thorough pre- and post-catalysis electron microscopy investigations revealed that Cu–Ag pairing likely resulted in selective oxidation of Cu, resulting in a strained Cu₂O–Ag epitaxial relationship that can be retained in the reduced Cu–Ag during the CO₂RR. Using electrochemical methods as convenient “probes” to rationalise CO₂RR activity, we found ECSA and EIS measurements to be ineffective in predicting CO₂RR product selectivity. Instead, surface charge estimated using a modified pulse voltammetry technique is more suitable, where distinct behaviour between Ag- and Cu-containing catalysts can be observed. Due to severe reconstruction during catalysis, having a core–shell configuration is found to be more beneficial for catalytic performance than the initial composition. Beyond Cu–Ag, we believe that the findings are relevant to many other multi-component catalysts with immiscible constituents.

---

1 Introduction

Elevated global carbon dioxide (CO₂) concentrations have reached critically alarming levels, contributing to annual emissions exceeding 32 billion tonnes per year,¹ with the atmospheric CO₂ burden recognised as the primary driver of contemporary climate change.² Among the principal net-zero CO₂ strategies, carbon capture and utilisation (CCU) is arguably the more compelling strategy that allows valorisation of captured CO₂ into renewable feedstocks for the chemical industry.³ A key strategy for effective CCU lies in the design of advanced electrocatalysts that can efficiently and selectively convert CO₂ into valuable carbonaceous products.⁴ Electrochemical CO₂ reduction allows the integration of CO₂ splitting and hydrogenation within a streamlined continuous process.⁵ Furthermore, its compatibility with on-demand renewable energy sources that are becoming cost-effective (as low as US$ 0.03 kWh⁻¹ for solar⁶ and US$ 0.02 kWh⁻¹ for wind⁷) offers the potential for profitable, zero- or negative-emission industrial chemical production when deployed at scale.^8,9

Copper (Cu) is the most prominent single metal catalyst capable of reducing CO₂ into a myriad of multi-carbon products.^10,11 To enhance its CO₂RR activity, researchers have proposed to combine Cu with different metals to form Cu-containing bi- or multimetallics (including alloys or composites) that may exert synergistic effects. Amongst the different bimetallics, Cu–Ag stood out, with no less than 100 published literature studies demonstrating a significant boost in CO₂RR activity in the past decade. By itself, Ag is one of the most active catalysts for the CO₂RR to CO,¹² and heightened production of CO is commonly observed in the majority of Cu–Ag bimetallics.¹³ However, there are other bodies of work that do not show increased CO production, especially those with smaller fractions (<10 at%) of well mixed Ag.¹⁴

One of the most popular explanations for the enhanced CO₂RR on Cu–Ag alloys or mixtures is the “CO spillover effect”, ascribed to the utilisation of produced CO(g) or transfer of adsorbed *CO intermediates from Ag sites to the neighbouring Cu sites, which allows higher *CO coverage and enhanced C–C coupling activity.^15,16 Alternatively, Clark et al. proposed that compressive strain imposed on Cu atoms, especially on the surface of bulk Cu–Ag bimetallics, is responsible for the enhanced C₂₊ selectivity.¹⁷ A more recent study using in situ nanofocused X-ray absorption spectroscopy revealed the dynamic nature of the atomic arrangement and strain within Cu–Ag Janus-type particles.¹⁸

Several major complications of Cu–Ag for the CO₂RR are their immiscibility, very high mobility of both Cu and Ag atoms, and galvanic corrosion. Recalling that Cu–Ag is a eutectic system with two terminal solid solutions,¹⁹ Cu–Ag bimetallics are not likely to exist in an alloyed form at room temperature, but rather as segregated solid phases, even near either end of the phase diagram.^20,21 To gain some control over the segregation behaviour in Cu–Ag bimetallics, many have proposed engineered microstructures, such as core–shell structures,²² Janus particles²³ or extremely small (<2 nm) nanoparticles,²⁴ which typically mandates the addition of polymeric surfactants or other ligands. Synthesising Cu–Ag bimetallics with good composition and size control without any surfactant or ligand is significantly more challenging. Even with some initial control during synthesis, we expect that surface reconstruction of Cu–Ag bimetallics under the CO₂RR will be more severe compared to bare Cu due to the combined dynamic effects of galvanic corrosion^25,26 and surface reactions facilitated by water/hydroxyls.²⁷ This reconstruction process often results in irregular and much smaller particles compared to the initial form, especially those containing lower (<20%) Ag content.¹⁶ Conversely, Cu–Ag bimetallics with higher (>20%) Ag content tend to show a larger chunk of Ag-rich sections that are more resistant to reconstruction.^28–30

Besides complex high vacuum deposition and ablation techniques that require expensive high purity metal targets,^16,31 an atmospheric pressure plasma jet (APPJ) is a powerful and versatile alternative technique. An APPJ is capable of depositing ligand-free multi-component nanosized metals from aqueous metal salts without the need for any vacuum or extreme heating.^32–35 Here, we found that an APPJ can consistently synthesise CuO_x–Ag composite films consisting of conjoined (Janus) or core–shelled particles with a Cu : Ag ratio matching the initial precursor ratio. The galvanic pairing effect inevitably converts Cu to CuO_x, which serendipitously facilitates an epitaxial relationship between Cu and Ag. Films with higher Ag fractions (≥60%) were found to produce C₂₊ liquid products favourably. Extensive high resolution electron microscopy and elemental analyses, together with electrochemical characterisation, provided evidence that the majority of Ag tends to aggregate either as the inner core in core–shell arrangement or as a distinct section in Janus-type particles. We found that core–shell arrangements are more favourable than Janus-type arrangements as the former retains a higher Cu–Ag interfacial area during the CO₂RR.

2 Results and discussion

2.1 Cu–Ag composites with varying metal ratios as efficient electrocatalysts for CO₂ reduction

CO₂RR product distributions from electrodes synthesised on APPJ grown CuO_x–Ag composites with rising Cu content were investigated. Cu₀Ag₁₀₀ (pure Ag) was taken as a baseline, showing the expected CO as the main CO₂RR product (Fig. 1a). The addition of small amounts (5 at%) of Cu (Cu₅Ag₉₅) still yielded CO as the main CO₂RR product, but this time with extended H₂ evolution suppression towards more cathodic potentials up to −1.6 V (Fig. 1b). A striking observation is the appearance of liquid C₂ oxygenates (ethanol, FE_C2H5OH >3.7%) at −1 V that was accompanied by only trace amounts of the corresponding alkene (ethylene, FE_C2H4 <0.27%), representing an ethanol/ethylene ratio of >14 (Fig. 1g). Dominant ethanol production over ethylene was still observed with more cathodic potential on Cu₅Ag₉₅, albeit with a reduced ethanol/ethylene ratio of ≈2 at −1.6 V. With increasing Cu content, the production of ethanol and n-propanol increases significantly; at −1.2 V, FE_C2H5OH >12.7% and FE_C3H7OH >3.0% were recorded for Cu₂₀Ag₈₀ (Fig. 1c), while FE_C2H5OH >13.6% and FE_C3H7OH ≈ 3.9% were recorded for Cu₄₀Ag₆₀ (Fig. 1d). The total FE for 2- and 3-carbon liquids was ≈20% for Cu₂₀Ag₈₀ and Cu₄₀Ag₆₀. Increasing the Cu fraction of the CuO_x–Ag composites was found to be counterproductive, with diminishing FE_C2+ selectivity observed when the Cu fraction is >60% (Fig. 1e and f) and the overall product distribution beginning to resemble that of pure bulk Cu (Fig. S1b).

Fig. 1 Average of steady state CO₂RR faradaic efficiencies (FEs) for gaseous and liquid products against various applied potentials (V vs. RHE) during 1.5 h of electrolysis on (a) Cu₀Ag₁₀₀, (b) Cu₅Ag₉₅, (c) Cu₂₀Ag₈₀, (d) Cu₄₀Ag₆₀, (e) Cu₆₀Ag₈₀, and (f) Cu₈₀Ag₂₀. (g) FE_C2H5OH/FE_C2H4 ratio and (h) partial current density for multi-carbon products (j_C2+) on Cu₅Ag₉₅ to Cu₁₀₀Ag₀. Each point represents an average of three independent measurements, and the error bar represents the standard deviation.

Generally, our APPJ-grown films showed higher ethanol/ethylene ratios compared to pure Cu (Fig. 1g), with lower Cu fraction (5–40%) films showing more than twice the typical 0.4–0.5 ratios reported on bulk Cu or oxide-derived Cu catalysts.^11,36 In the literature, the CO₂RR pathway towards ethanol on bare Cu has been shown to branch out from C₂H₄ at *CH–COH,^37,38 where ethanol formation is more endothermic by 0.43 eV. Strategies to steer CO₂RR production to a higher ethanol/ethylene ratio has been reported by using higher KOH electrolyte concentration³⁹ or by modifying the Cu surface with N-containing^40,41 or S-containing⁴² organic ligands. Further, the majority of Cu–Ag bimetallics reported for CO₂RR typically show FE_C2H5OH/FE_C2H4 ratios less than 1,^13–18,43 with the exception of reports from Wang et al. (ratio: 1.17)⁴⁴ and Gao et al. (ratio: 8).⁴⁵ A more comprehensive comparison with other previously reported Cu–Ag catalysts is presented in Table S7. As no organic ligands were added to our APPJ-grown films, we posit that the favourable ethanol production should be attributed to weaker *O binding energy and the more difficult O–C bond cleavage for *OCHCH₂ on strained Cu–Ag compared to pure Cu.^18,46

Partial current densities for C₂₊ products (j_C2+) are plotted in Fig. 1h (C₁ products and H₂ in Fig. S2a and b) to estimate the CO₂RR turnover. The optimum potential to produce C₂₊ products was at −1.4 V vs. RHE for most catalysts except Cu₄₀Ag₆₀ that has a slightly higher turnover at −1.2 V. APPJ grown CuO_x–Ag composites generally outperform pure Cu up to Cu₆₀Ag₄₀, but a higher Cu : Ag ratio starts to follow the trend of pure Cu. Overall, Cu₄₀Ag₆₀ appears to exhibit the highest C₂₊ selectivity, followed by Cu₂₀Ag₈₀. Further characterisation will focus on these catalysts.

2.2 Characterisation of as-synthesised Cu₂₀Ag₈₀ and Cu₄₀Ag₆₀

We focused our investigations on Cu₂₀Ag₈₀ and Cu₄₀Ag₆₀ that show the best steady-state CO₂RR faradaic efficiencies (FE). SEM characterisation revealed that the higher Ag content Cu₂₀Ag₈₀ film consists of larger (ca. 100–200 nm) primary particles (Fig. 2a and b), whereas Cu₄₀Ag₆₀ consists of larger aggregates (Fig. 3a and b). High-angle annular dark-field (HAADF) imaging, scanning transmission electron microscope energy dispersive spectroscopy (STEM-EDS) mapping, and high-resolution TEM micrographs, however, tell a more interesting story. We discovered that the primary particles of lower Cu content Cu₂₀Ag₈₀ are Janus-type particles, with distinct segregation between Cu-rich and Ag-rich regions (Fig. 2d–g). On the other hand, Cu₄₀Ag₆₀ has core–shell-type primary particles, consisting of an Ag-rich core and Cu-rich shell (Fig. 3d–g). As there is no surface directing agent or linker used and an identical APPJ procedure was adopted, we expect that the factor determining the particle type is the molar ratio between Cu(NO₃)₂ and AgNO₃ precursors.

Fig. 2 Electron microscopy characterisation of the APPJ synthesised Cu₂₀Ag₈₀ bimetallic. (a) Low and (b) high magnification FE-SEM images of Cu₂₀Ag₈₀ deposited on a glassy carbon substrate. (c) Low magnification TEM image, (d) high angle annular dark field TEM image and electron dispersive spectroscopy map of (e) Ag, (f) Cu, and (g) O elements. Green arrows in the sub-figure (f) indicate traces of Cu deposited around the Ag-rich part. (h) High resolution TEM of representative APPJ grown Cu₂₀Ag₈₀ viewed along the [211̄] zone axis, showing a sharp Cu₂O–Ag boundary. The inset of the subfigure (h) shows the simulated electron diffraction of the particle generated by fast Fourier transform calculation, showing closely aligned Cu₂O and Ag lattices, with a compressed Cu₂O lattice along the [011] direction but relaxed along the [111] direction. (c–h) Images of samples directly grown on SiN_x TEM grids.

Fig. 3 Electron microscopy characterisation of the APPJ synthesised Cu₄₀Ag₆₀ composite. (a) Low and (b) high magnification field emission electron scanning microscopy image of Cu₄₀Ag₆₀ deposited on a glassy carbon substrate. (c) Low magnification transmission electron microscopy (TEM) image, (d) high angle annular dark field TEM image and electron dispersive spectroscopy map of (e) Ag, (f) Cu, and (g) O elements. (h) High resolution TEM of representative Cu₄₀Ag₆₀ showing less defined boundaries between Ag and Cu₂O regions and existence of multiple crystal orientations. Inset in (h) shows FFT generated electron diffraction, showing closely aligned Cu₂O [222] and Ag [222]. (c–h) Images of samples directly grown on SiN_x TEM grids.

The plasma used in APPJ is considered a non-equilibrium plasma, consisting of energetic electrons and “cold” neutral and ionic species.^34,47,48 Therefore, the majority of solvent droplets carrying Cu–Ag precursors should not evaporate immediately, allowing energetic surface electron irradiation to drive diffusion, transport and reaction (e.g., metal reduction).⁴⁹ Transitions from core–shell to Janus structures have been observed for uncapped Cu–Ag bimetallic nanoparticles, driven only by temperature and increasing Ag composition.⁵⁰ In our case, although the formation of bimetallics is driven by energetic electrons within the plasma, we posit that similar effects can still occur within the tiny vapour droplets carrying well mixed Cu–Ag precursors without any surface directing agents, resulting in a clear transition between Janus and core–shell when the Cu : Ag fraction is tuned. Although N₂ sheath gas was applied during APPJ synthesis, the EDS distribution map for our samples still detected significant amounts of oxygen, especially around the Cu-rich regions on both Cu₂₀Ag₈₀ and Cu₄₀Ag₆₀, pointing to possible CuO_x formation (Fig. 2g and 3g, and line EDS in Fig. S4). Although the deposition of single components Cu and Ag in metallic form has been demonstrated previously,³⁴ we believe that the selective oxidation of Cu in the Cu–Ag bimetallic configuration may be inevitable due to the galvanic corrosion effect.⁵¹

Closer scrutiny of the particle and the interface between Cu-rich and Ag-rich regions using HRTEM confirmed that the crystal structure of the Cu-rich region is consistent with the Cu₂O lattice. On the Cu-rich section of the Cu₂₀Ag₈₀ sample (Fig. 2h), the d-spacing perpendicular to the Cu₂O–Ag interface (i.e., Cu₂O d_(1–11)) was found to be ≈ 0.240 nm, which is close to the relaxed Cu₂O d₍₁₁₁₎ of 0.2464 nm.⁵² The Ag d₍₁₁₁₎ near the Cu–Ag boundary was found to be 0.230 nm, which is also close to the expected value for bare Ag (≈0.236 nm).⁵³ However, electron diffraction analysis (calculated using fast Fourier transform (FFT) from the HRTEM micrographs, inset of Fig. 2h) revealed closely matched Ag and Cu₂O lattices in the [011] direction, which indicates significant strain in the Cu₂O lattice. This supposition is also supported by the strain analysis based on the d-spacing variation (Fig. S7).

On Cu₄₀Ag₆₀, the observed Cu₂O d₍₁₁₁₎ of 0.242 and Ag d₍₁₁₁₎ of 0.230 nm (Fig. 3h) are also close to the respective bulk values. Like the Cu₂₀Ag₈₀ samples, a very close crystallographic relationship between Cu₂O and Ag was also observed on Cu₄₀Ag₆₀. Due to the core–shell nature of the particle that causes overlapping crystal orientations, the electron diffraction pattern can only reveal closely aligned Cu₂O(111) and Ag(111) planes perpendicular to the Ag–Cu₂O interface (inset in Fig. 3h), indicative of an epitaxial-like relationship. However, we believe that a similar strain along the Ag–Cu₂O interface also exists in the core–shell particle.

In the literature, an epitaxial relationship between metallic Cu–Ag is difficult to achieve due to the large lattice mismatch (12.6%).^52,53 To our knowledge, true epitaxial growth of metallic Cu–Ag can only be achieved via a surface limited redox replacement reaction of monolayer sacrificial metal (typically Pb)⁵⁴ or by introducing large amounts of twin boundaries on either Cu or Ag seed crystals in the presence of galvanic corrosion suppression agents.⁵¹ Serendipitously, oxidation of the Cu-rich region in our samples makes a stable epitaxial relationship between oxidised Cu₂O and Ag more facile. On the (111) plane especially, the lattice mismatch between Ag and Cu₂O (4.3%) is significantly smaller than that between Cu and Ag (>12%). It is notable that our APPJ synthetic approach can readily achieve a Cu₂O–Ag epitaxial-like relationship without adding any surface capping agents or other additives. As the lattices of both Ag and Cu₂O are almost fully relaxed, we believe that the epitaxial-like relationship in APPJ grown Cu₂O–Ag composites observed here can be maintained until it is ready for use as a CO₂RR catalyst.

Overall, the elemental analyses by TEM and SEM EDS demonstrate good control over Cu–Ag composition in the APPJ synthesised composites, where the actual Cu : Ag ratios are within 2% of the intended precursor values (Table 1 and S2). Curiously, the surface composition probed by XPS indicated a very high surface Cu : Ag ratio for both samples. While a high surface Cu content is expected for Cu₄₀Ag₆₀ that has a core–shell structure, the observation of a high surface Cu content on Cu₂₀Ag₈₀ suggests that Cu deposition may be spread more widely than expected. A closer inspection of the TEM EDS map of Cu₂₀Ag₈₀ shows traces of smaller Cu particles that likely surround the Ag-rich regions (green arrows in Fig. 2f).

Table 1 Elemental analyses of Cu₂₀Ag₈₀ and Cu₄₀Ag₆₀ tracking Cu and Ag before and after the CO₂ reduction reaction using TEM EDS (representing bulk composition) and XPS (representing surface composition)

	TEM EDS (at%)			XPS (at%)
	Cu	Ag	Std dev.	Cu	Ag	Std dev.
Cu20Ag80 as-grown	18.89	81.12	2.27	82.12	17.88	1.00
Cu20Ag80 post-CO2RR	8.26	91.75	2.40	86.99	13.01	0.78
Cu40Ag60 as-grown	41.40	58.60	3.37	84.31	15.69	1.24
Cu40Ag60 post-CO2RR	35.19	64.81	8.09	91.26	8.74	1.51

---

2.3 Characterisation of Cu₂₀Ag₈₀ and Cu₄₀Ag₆₀ post-CO₂RR

Changes in CuO_x–Ag composites after the CO₂RR can be discombobulating due to the compounding effects of cathodic potential, galvanic effect, surface reactions, and post-catalysis reoxidation that require comprehensive analysis using multiple analytical tools.^25,26

Starting with HAADF and STEM-EDS analyses, the most significant reconstruction is found on Cu₂₀Ag₈₀, where the Janus arrangements disappear after the CO₂RR. Instead, smaller particles with significantly more dispersed Cu and lower oxygen signals than the as-prepared samples are observed (Fig. 4a–d and S5a–c). A closer inspection of the Cu₂₀Ag₈₀ lattice using HRTEM reveals a practically unchanged Ag-rich region (Ag d₍₁₁₁₎ ≈ 0.230 nm) but a more spread-out Cu-rich region with lattice spacings that are difficult to discern (Fig. 4e). Statistical TEM particle analysis (Fig. S6a and Table S1) detected a slight reduction of the mean particle size of Cu₂₀Ag₈₀ from 28.4 nm to 24.3 nm after the CO₂RR, suggesting possible partial dissolution during catalysis.

Fig. 4 Post-CO₂RR characterisation using HAADF, TEM-EDS mapping, and HRTEM for (a–e) Cu₂₀Ag₈₀ and (f–i) Cu₄₀Ag₆₀, performed on scraped Cu–Ag samples on glassy carbon substrates, resuspended in isopropanol, and deposited on holey carbon TEM grids.

While copper oxides are generally expected to be reduced during the CO₂RR, re-oxidation within 60 seconds of voltage removal to a mixture of Cu₂O and other less well-defined Cu species has been observed.⁵⁵ To explain these findings, post-CO₂RR XPS is performed on the spent electrodes. A comparison of high-resolution Cu 2p spectra (Fig. S16a and c) clearly shows suppression of the Cu satellite peaks after catalysis, suggesting at least partial surface reduction to a possible mixture of Cu⁽⁰⁾ and Cu^(I) post-CO₂RR. Cu LMM Auger spectra show the presence of a mixed oxidation state after the CO₂RR (Fig. S16e), with dominant Cu^(I).56 On the other hand, a comparison of Ag 3d spectra (Fig. S16b and d) showed complete disappearance of Ag(I) peaks after the CO₂RR, suggesting significant re-oxidation resistance of Ag. Taken together, these observations indicate more severe changes in Cu compared to Ag in the Cu₂₀Ag₈₀ Janus structure, possibly leading to Cu redistribution throughout the electrode.

The reconstruction on the Cu₄₀Ag₆₀ sample looks rather different. Large sections (>100 nm) of Ag-rich particles are readily found post-CO₂RR (Fig. 4f, g, S5d and e). Most importantly, significant amounts of Cu are still observed to congregate around the Ag-rich core (Fig. 4h and S4f), suggesting that the core–shell morphology is relatively preserved. In contrast with Cu₂₀Ag₈₀, statistical TEM particle analysis of Cu₄₀Ag₆₀ (Fig. S6b and Table S1) indicates growing mean particle size from 46.2 to 60.1 nm after the CO₂RR, suggesting possible agglomeration after catalysis. Although we could not discern the lattice spacings for Cu₄₀Ag₆₀ HRTEM post-CO₂RR, the core–shell construction in the individual particles can still be observed, with the lighter contrast on the shell likely to be re-oxidised copper (Fig. 4j). Post-CO₂RR XPS analysis of Cu 2p and Ag 3d shows similar behaviour to Cu₂₀Ag₈₀, where the Cu satellite peaks are significantly reduced post-catalysis (Fig. S14a and c), and the Ag(I) shoulder is completely removed (Fig. S14b and d). A similar shift in Cu LMM peak kinetic energy is also observed (Fig. S14e), suggesting the formation of mixed Cu species after the CO₂RR with dominant Cu^(I).

The distinct modes of reconstruction on both samples are also corroborated using elemental analyses. The bulk Cu fraction of Cu₂₀Ag₈₀ (measured by TEM EDS) dropped from 18.9% to 8.3% post-CO₂RR, while a much smaller reduction from 41.4% to 35.2% was observed on Cu₄₀Ag₆₀ (Table 1). In contrast, a pronounced increase in surface Cu fraction (measured by XPS), from ≈84% to >91% was detected on Cu₄₀Ag₆₀, compared to 82% to 87% on Cu₂₀Ag₈₀.

While significant mobility of both Cu and Ag is expected during catalysis,⁵⁷ the different reconstruction modes observed on Cu₂₀Ag₈₀ and Cu₄₀Ag₆₀ suggest that the initial core–shell structure encourages better retention of the Cu–Ag interface during the CO₂RR than in Janus-type particles. To support this position, ICP-OES analysis of the spent catholytes was performed (Table S5 and Fig. S18a). Expectedly, a higher leached Cu absolute concentration was detected due to a higher absolute initial loading on samples with a higher Cu fraction. However, when normalised to the Cu fraction, a non-uniform U-like response was observed, suggesting a lower normalised Cu leaching rate on Cu₄₀Ag₆₀ compared to Cu₂₀Ag₈₀. We also note a generally more stable C₂H₄ turnover on the core shell Cu₄₀Ag₆₀ (≈3% reduction from t = 20 min to t = 80 min, Fig. S3b) compared to Cu₂₀Ag₈₀ (≈30% reduction, Fig. S3a).

2.4 Electrochemical characterisation

Electrochemical characterisation was also carried out to supplement physical characterisation data. First, the electrochemically active surface area (ECSA) was estimated using multiple CV runs at 14 different scan rates and double layer capacitance (C_DL) extracted from the current density and scan rate relationship (Fig. S4a and b). Overall, incorporation of Cu into Ag appears to increase the overall C_DL from 0.5 mF exhibited by Cu₀Ag₁₀₀ to about 0.8 mF for Cu₆₀Ag₄₀ (Fig. S4c). However, as there are mixed elements in the composite, it is difficult to ascertain if the small increments in the C_DL values are directly linked to a larger surface area.

We then turned to EIS to probe the samples further. Near the open circuit potential (OCP), films with larger Cu : Ag fractions showed lower charge transfer resistance (R₁) reflected roughly by the diameter of the semicircles (Fig. S9a). However, EIS measurements at OCP could not be relevant to the catalytic process, as the catalyst surface and Helmholtz plane composition will be substantially altered upon external voltage application beyond catalytic onset.⁵⁸ More representative and quantitative metrics can be extracted from EIS data at or near CO₂RR relevant potentials using a numerical fitting procedure.^40,58 Numerical fitting results using the Armstrong–Henderson equivalent circuit that considers electrochemically adsorbed intermediates (Fig. S9c)^59,60 are displayed in Fig. S10. The resistive components of R₁ and R₂ decreased exponentially with a more cathodic applied voltage, while the capacitive elements (CPE₁ and CPE₂) showed different responses. CPE₁ generally shows a relatively flat response towards the applied voltage (Fig. S10b), while exponential changes in CPE₂ were observed (Fig. S10d). While R₁ may be the preferred metric to explain simpler electrocatalytic reactions like water splitting,⁵⁸ we posit that the longer timescale CPE₂ is more relevant to multi-step electrocatalytic processes like the CO₂RR to C₂₊ that require intermediate accumulation and non-electrochemical C–C coupling steps.⁶¹ Unfortunately, attempts to find a correlation between electrochemically measured metrics, through either ECSA or EIS, and FEC₂₊ did not reveal any strong correlation (Pearson's |r| ≈ 0.02 to 0.28, Fig. S11). The weak correlation between R₁ and FEC₂₊ also suggests that the CO₂RR to C₂₊ on Cu–Ag in our experiment may not operate in the charge transfer limited region.

Additional analyses were carried out using modified pulse voltammetry (mPV), involving stepwise changes from open circuit potential (OCP) to progressively more cathodic voltage (Fig. S12). mPV has been shown to correlate well with the CO₂RR activity of organic-functionalised Cu.⁴⁰ Here, we believe that it may also be applicable to metal combinations and bimetallics. Specifically, the integrated anodic charge (Q_an) measured during potential reversal from applied cathodic potentials to near OCP is proposed to be proportional to the electrochemically absorbed species on the catalyst surface (see SI Section 6 for further information). Tabulation of Q_an values with respect to ΔV presented in Fig. 5a shows distinct characteristics of catalysts with and without Cu addition. Samples without Cu (i.e., Cu₀Ag₁₀₀) show relatively flat Q_an values with increasing ΔV, indicative of low accumulation of adsorbed species and consistent with Ag's inherently weak interaction with *CO (a crucial intermediate for the CO₂RR to C₂₊).⁶² When Cu is added in the bimetallic, we observe a significant “jump” in Q_an when ΔV >0.8 V. This voltage range coincides with the *CO accumulation event that preceded C₂₊ formation during the CO₂RR on Cu as reported previously.^11,61 Additionally, a strong correlation between Q_an and j_C2+ at −1.4 V vs. RHE on Cu–Ag bimetallics at comparable applied voltage is observed (Pearson's r = −0.85, Fig. 5b).

Fig. 5 (a) Integrated anodic charge (Q_an) of APPJ-deposited Cu–Ag bimetallics with increasing ΔV. The X axis “ΔV” represents the voltage difference between the applied pulse and the OCP, which varies slightly on each material. Roughly, ΔV = 1.6, 1.8, 2.0 and 2.2 correspond to catalytic potentials of −1.0, −1.2, −1.4 and −1.6 V vs. RHE. Data were obtained as the average of 3 independent replicates. (b) Correlation plot between average j_C2+ at −1.4 V vs. RHE and Q_an at ΔV = 2.0 V for APPJ deposited Cu–Ag bimetallics. Dotted line represents York linear fitting (Pearson's r = −0.85) that considers both X and Y errors, while grey shading marks the 95% confidence interval bands.

3 Discussion

There are two popular explanations in the literature for the enhanced CO₂RR to C₂₊ on Cu–Ag bimetallics. Early studies appear to favour the “CO spillover” explanation, where the enhanced C₂₊ products on Cu–Ag are always accompanied by a concomitant increase in CO(g) production.^15,16 The premise of this hypothesis is that some of the adsorbed *CO or the CO(g) produced in Ag sites can be transferred to the neighbouring Cu sites, enabling better *CO accumulation. Other researchers had different ideas, as the CO(g) production appears to be proportional to surface Ag content post-CO₂RR, and the enhanced C₂₊ production should instead be linked to the compressive strain in the Cu lattice.¹⁷ Although based on a smaller dataset, a relatively more linear j_CO relationship with the Cu : Ag fraction is also found in this work, but much less so for j_C2+ (Fig. S13). As such, we are inclined to concur that the enhanced C₂₊ production should be ascribed to a much more fundamental change in the active catalyst structure rather than a simple *CO spillover effect.

Extensive TEM experiments in this work unveiled that the Cu rich region in our APPJ grown Cu–Ag is readily oxidised to the Cu₂O structure, even though individual metallic Cu and Ag growth using the same technique has been demonstrated previously.³⁴ We believe that this is due to the galvanic effect and should be universal to pairings between Cu and more noble metals without any surface capping agent or linkers. Fortuitously, oxidised Cu₂O still has a matching (Fm3̄m) space group with Ag, with a significantly smaller lattice mismatch than the metallic Cu–Ag pair, allowing a stable Cu₂O–Ag epitaxial-like relationship. Additionally, Cu₂O should also experience a roughening phenomenon during the initial reduction phase and bring about the oxide-derived Cu benefit.⁵⁵

After the CO₂RR, a more severe reconstruction was observed on Cu₂₀Ag₈₀ catalyst. On this sample, Janus arrangements can no longer be found, accompanied by significant Cu redistribution away from Ag. On the other hand, the core–shell configuration on Cu₄₀Ag₆₀ can still be readily found, with a significant majority of Cu still surrounding the Ag-rich core. HRTEM investigation revealed post-catalytic conversion of Cu species to a less well-defined oxidised structure. XPS analyses show that both samples are generally more reduced after the CO₂RR to a Cu^(I) dominated mixture, accompanied by suppression of the satellite peaks consistent with Cu^(II).

Between Cu₂₀Ag₈₀ and Cu₄₀Ag₆₀, we believe that the morphology of the initial Cu–Ag bimetallics may be more dominant in determining CO₂RR activity than the absolute Cu : Ag ratio. As our CuO_x–Ag composites are not functionalised by any ligand, the morphology change is entirely driven by the Cu : Ag ratio and thermodynamics.⁵⁰ While we expect that the in situ epitaxial relationship between Cu₂O and Ag to be morphology agnostic, the core–shell microstructure may have a competitive edge due to its ability to retain more Cu in direct contact with the Ag core during the CO₂RR. The closer proximity of the CuO_x shell to the Ag core can also reduce Cu redistribution significantly during the initial phase of reduction. With a matching Fm3̄m space group, Ag is still a preferred Cu (re)nucleation site compared to random surfaces like glassy carbon due to lower activation free energy for nucleation.⁶³ This advantage is also reflected in a more stable C₂H₄ turnover on Cu₄₀Ag₆₀ (only about 3% degradation after 80 min) compared to Cu₂₀Ag₈₀ that shows a >30% decrease over the same period (Fig. S3). As Cu₂O is expected to reduce to metallic Cu during the CO₂RR, we believe the reduced Cu will be forced to maintain its tensilely strained state conforming to the Ag lattice under applied cathodic bias. Such voltage-dependent strain behaviour has been reported on Cu–Ag particles using nano-focused X-rays,¹⁸ where the larger tensile strain on the Cu side of the Janus can be retained under cathodic bias. Additionally, tensile strain in the Cu lattice has been shown using DFT to increase *CO binding strength and, more importantly, reduce ΔG_formation for C₂₊ intermediates on Cu.⁶⁴

We also acknowledge the possibility that true Cu–Ag alloying contributes to the enhanced CO₂RR activity. Although we cannot conclusively demonstrate the existence of Cu–Ag alloys in very small particles (<10 nm) after the CO₂RR, there are at least two papers,^24,65 where the α-phase (Ag rich) can exist up to a 28% Cu : Ag ratio in very small (5 nm) nanoparticles. However, due to severe reconstruction of the catalysts under reaction conditions, we posit that the precise initial composition of alloys, especially in immiscible metal pairings like Cu–Ag, may not matter too much as the metal pairs tend to segregate during the reaction. Instead, having an initial core–shell configuration, with a thin layer of the active species shell facing the reactant, is more critical as the active species will remain anchored to the stable core, while allowing the strain effect to kick in.

On the use of electrochemical methods as convenient “probes” to rationalise CO₂RR activity, we found ECSA and charge transfer resistance parameters extracted from numerical fittings of EIS measurements to be ineffective in predicting C₂₊ activity amongst CuO_x–Ag composites (Fig. S11). This is consistent with the literature as the CO₂RR to C₂₊ involves a non-electrochemical coupling step that is limited to the surface *CO coverage^61,66 instead of charge transfer kinetics. We found that a more suitable metric is the surface charge estimated using mPV, where distinct behaviours between Ag- and Cu-containing catalysts are observed (Fig. 5). Potentially, we believe that surface charge estimation using mPV can be applied to multi-metal catalysts, in addition to organic functionalised Cu demonstrated earlier.⁴⁰

4 Conclusion

This research provides critical information in the quest to understand the multi-step CO₂RR on bimetallic structures. Based on a comprehensive electron microscopy study that tracks the evolution of a ligand-free CuO_x–Ag composite from its growth using an APPJ to its reconstruction after the CO₂RR, we believe that the enhanced C₂₊ production on Cu–Ag should be ascribed to a much more fundamental change in the active catalyst structure rather than *CO spillover. One of such changes is the tensile strain in the Cu lattice that can be brought about by applied cathodic potential that can force reduced Cu to conform to the Ag lattice. The epitaxial relationship between Cu₂O and Ag is observed on APPJ grown films, accommodated by the initial oxidation of Cu due to galvanic corrosion upon pairing with the more noble Ag. Although this configuration inevitably evolved during the CO₂RR, we believe that the applied cathodic potential can provide the necessary driving force to maintain the Cu–Ag epitaxial relationship in the re-nucleated Cu. Evidence of smaller Cu nuclei surrounding larger Ag-rich regions that retain some epitaxial Cu–Ag relationship supports previous literature on the dynamic structure–strain relationship on Cu–Ag. Further, we believe that the morphology of the initial CuO_x–Ag composite may be more important in enhancing the CO₂RR activity than the absolute Cu : Ag ratio. In our APPJ grown films, a moderate Cu : Ag ratio of around 40 : 60 allows the formation of an Ag@Cu₂O core–shell arrangement that is beneficial as it exposes and better retains the more active species (strained Cu–Ag interface) during the CO₂RR. In addition to advancing green electrocatalyst development, this work underscores the potential of an APPJ in developing catalyst materials with controllable compositions and morphology for scalable, sustainable carbon capture and utilisation (CCU) technologies.

5 Methodology

5.1 Atmospheric pressure plasma jet (APPJ) deposition of Cu–Ag alloy thin films at various ratios

The synthesis and deposition of Cu–Ag bimetallics using an APPJ were performed according to our previous work with 20 W effective power.³⁵ In brief, 96% He gas mixed with 4% H₂ (NOX Singapore) was used as the plasma working gas, ensuring complete reduction of the samples. In brief, Cu–Ag thin films were prepared by nebulising aqueous precursor solutions containing only deionised (DI, 18.2 MΩ.m) water and varying molar ratios of Cu (NO₃)₂ and AgNO₃. The sample nomenclature Cu₂₀Ag₈₀ denotes a Cu : Ag precursor composition consisting of 20% (0.1 mM) Cu²⁺ and 80% (0.4 mM) Ag⁺. Unless stated otherwise, the total concentration of metal ions in the aqueous solutions was maintained at 0.5 mM for all samples, such that [Cu²⁺] + [Ag⁺] = 0.5 mM. All samples were deposited on glassy carbon disks (∅15 × 1 mm thickness, Goodfellow Cambridge Ltd).

5.2 Electrochemical methods

Glassy carbon disks were mechanically polished using 5 µm diamond slurry (Struers AG), followed by fine polishing to a mirror finish with a 0.3 µm alumina slurry (Buehler MicroPolish). The substrates were subsequently sonicated in concentrated KOH solution for 10 minutes to remove possible embedded alumina. Post-deposition, the samples were rinsed with deionised water, dried under a N₂ stream, and stored overnight in a dry chamber for 16–18 hours prior to electrochemical experiments. 99.999% purity Cu disks (∅15 mm × 1 mm thickness, Goodfellow Cambridge Ltd) were polished with 3 µm polycrystalline diamond slurry, followed by sonication in DI water for 1 min. The Cu disks were dried using inert N₂ gas.

5.3 Electrolyte preparation

EMSURE ACS grade KHCO₃ (99.7 to 100.5% assay, Supelco) was used for the electrolyte. To minimise dissolved metals in the electrolyte, pre-electrolysis electrolyte cleaning was performed by applying −1 mA for at least 30 min in a 2-electrode set-up with 2 graphite rods (TedPella, Spec-pure, ∅1/8” × 20 cm), while being purged with CO₂. The KHCO₃ 0.1 M electrolyte was initially bubbled and saturated at 20 sccm for 20 min with CO₂ gas (Linde Gas, 99.999%) using 2 mass flow controllers (MFC, Alicat Scientific MC series).

5.4 Electrochemical CO₂ reduction reaction (CO₂RR)

All electrocatalytic experiments were performed in a custom H-cell made of PEEK and PTFE, as reported previously.⁴⁰ Both anodic and cathodic compartments were filled with 8 ml of CO₂ saturated 0.1 M KHCO₃. These compartments were separated via an anion exchange membrane AMVN (SELEMION™, ACG Engineering, Japan). All electrochemical experiments were conducted using a calibrated Gamry potentiostat (Gamry 600+ or Gamry 3000). A 3-electrode system was used in all electrochemical experiments, consisting of glassy carbon disks (Goodfellow, ∅15 mm × 1 mm thick) as the working electrode. A leakless Ag/AgCl electrode (eDAQ) was used as the reference, while a graphite rod (TedPella, Spec-pure, ∅1/8” × 20 cm long) was used as the counter electrode.

Electrocatalysis was conducted under a constant CO₂ flow rate controlled via an MFC at 20 sccm in both the anodic and cathodic compartments. The electrolyte resistance was measured three times prior to each chronoamperometry run using high-frequency electrochemical impedance spectroscopy, and the average value was used for iR compensation via positive feedback correction (85%). Chronoamperometry was carried out for 100 min. All potentials reported in this thesis are with respect to the RHE, unless stated otherwise.

The CO₂ reduction catalytic performance was assessed across various Cu–Ag ratios to identify the optimal composition under four different potential ranges: −1.0, −1.2, −1.4, and −1.6 V vs. RHE. All experiments were conducted in triplicate, and the results are presented as average values with corresponding error bars in all figures, along with standard deviations for the recorded values. The catalytic performance of the glassy carbon disks was also evaluated (Fig. S1a), with predominant hydrogen selectivity. High-purity metallic Cu disks (99.999%) were assessed for their catalytic performance towards CO₂ in a custom-made electrocatalytic cell (Fig. S1b), and the results closely matched those reported in the literature.¹¹ Additional CO₂RR product evaluation methods are described in SI Section 1.

5.5 Electrochemical characterisation: electrochemically active surface area (ECSA), electrochemical impedance spectroscopy (EIS) and modified pulse voltammetry (mPV)

All the electrochemical characterisation studies were carried out in an open PTFE cell. A pre-reduction step was performed at −1.2 V vs. RHE for 10 min before electrochemical characterisation methods to reflect CO₂RR conditions.

Cyclic voltammetry (CV) measurements were conducted in 1 M KHCO₃ to improve the signals. CV scans were performed around the non-faradaic region near open circuit potential (OCP) between −0.6 V and −0.7 V vs. Ag/AgCl (equivalent to 0.0 to +0.1 V vs. RHE, Fig. S5a). CV scans were performed at 14 different scan rates within the capacitive regions, ranging from 20 mV s⁻¹ to 150 mV s⁻¹, with each subsequent scan rate increasing by 10 mV s⁻¹. Each scan rate was repeated three times, with only the 3rd scan used for data analysis. The exposed geometric area of the working electrode is 1.327 cm² and was calculated by subtracting the geometric surface area of the glassy carbon disk from that of the O-ring. The ECSA of each catalyst was determined based on their respective surface roughness factors, which correlate directly with the slope obtained from the linear relationship between current density and scan rates (Fig. S5b), based on previous literature.⁶⁷

EIS measurements were performed in 1 M KHCO₃ to improve the signals. Prior to the experiment, all samples were pre-reduced via chronoamperometry at 1.8 V vs. Ag/AgCl (−1.2 V vs. RHE) to reflect CO₂RR conditions. EIS measurements were conducted across a wide range of cathode potentials under CO₂RR conditions. EIS spectra were collected across 40 different potentials, ranging from −0.59 to −2.59 V vs. Ag/AgCl, in 0.05 V increments, with an oscillation applied to the AC voltage of ±10 mV rms. The spectra were collected from 0.2 Hz to 100 000 Hz at 10 points per decade (Fig. S6a and b). Gamry Echem Analyst (v7.9.0) was used to fit all Nyquist plots using a modified Armstrong–Henderson circuit model (Fig. S6c).

mPV experiments are performed in 0.1 M KHCO₃ under the same conditions as those of the CO₂RR without significant signal disturbance. mPV was tested with a square wave potential step, consisting of a fixed upper bound of 0 V vs. Ag/AgCl (near OCP) for 18 s, and increasingly cathodic lower bound voltages ranging from −0.5 V to −2.4 V vs. Ag/AgCl, incrementally decreased by 0.1 V at 18 s holding time, for a total of 41 steps, with a total duration of 738 s. The current response is recorded every 3 ms interval. The area under the anodic decay current response is integrated using scipy.intergrate.simps and fitted using scipy.optimize.curve_fit from t = 0 s to t = 2 s. The integrated current response is then plotted against ΔV, which is the difference between V_anodic and V_cathodic. An example plot of applied voltage and the corresponding current response is shown in Fig. S12a. More details are available in SI Section 6.

Unless stated otherwise, all potentials were converted into V vs. RHE for data analysis, as described previously. All experiments were performed in triplicate, and the results are presented as averages with their standard errors.

5.6 Additional characterisation techniques

X-ray photoelectron spectroscopy (XPS) spectra were recorded using a Kratos AXIS Supra + spectrometer equipped with an Al Kα X-ray excitation source (hν = 1486.7 eV). The spot size was determined to be 700 × 300 µm (slot mode). 50 scans were taken for Cu and Ag, while all other elements (O, N, and C) were averaged over 20 scans without Ar sputtering. Data were collected at 160 eV pass energy with a step size of 1 eV for survey scans, and at 40 eV pass energy with a step size of 0.1 eV for narrow scans. Charge compensation was achieved using low-energy neutralisation electrons. Casa XPS version 2.3.24PR1.0 was used to analyse the valence states from the XPS spectra and to calculate the elemental composition from the peak area, Thermo Avantage v5.9928 software was used. XPS data and detailed XPS fitting analyses are described in SI Section 8.

Transmission electron microscopy (TEM) analysis was conducted on a Thermo Scientific Talos F200X G2 (S)TEM (Thermo Fisher Scientific), operating at an accelerating voltage of 200 kV. The as-grown samples were directly deposited onto a 16-window Si grid (16 × 0.1 mm) coated with a 30 nm SiN_x (silicon nitride) membrane. Post-CO₂RR samples were obtained by mechanically transferring the films from glassy carbon disks, resuspended in isopropanol, and then dripped onto holey carbon films on nickel 200 mesh (50 µm) TEM grids. Gatan DigitalMicrograph software was used for image and data analysis. Elemental composition and distribution mapping were obtained using TEM-integrated EDS.

Inductively coupled plasma optical emission spectroscopy (ICP-OES) was performed on the catholyte before and after the CO2RR using an Avio 200 (PerkinElmer). Neat samples were analysed, with spectroscopy grade 0.1 M HNO₃ washing solution. More details and results are presented in SI Section 9.

Author contributions

Stefanos Agrotis: conceptualization, methodology, validation, investigation, visualization, formal analysis, data curation, writing – original draft, and writing – review & editing. Ming Lin: investigation: TEM experiments, TEM data acquisition, and analysis. Kallum H. Mehta: investigation: catalysis and writing – review & editing. Oliver S. J. Hagger: investigation and APPJ methodology. Riko I Made: writing – review & editing and supervision. Ivan P. Parkin: supervision, project administration, and funding acquisition. Daren J. Caruana: writing – review & editing, supervision, project administration, funding acquisition, and conceptualization. Albertus D. Handoko: writing – original draft, writing – review & editing, visualization, supervision, project administration, funding acquisition, formal analysis, data curation, and conceptualization.

Conflicts of interest

The authors declare the following financial interests/personal relationships that may be considered as potential competing interests: Daren J Caruana has a patent “PLASMA JET DEPOSITION PROCESS” pending to UCL Business Ltd.

Data availability

In addition to the main text, additional supporting data have been provided as part of the supplementary information (SI). Supplementary information: CO₂RR product evaluation methods, CO₂ reduction performance of baseline on Cu and blank glassy carbon, and additional electron microscopy analyses. Electrochemically active surface area (ECSA) estimation, electrochemical impedance spectroscopy (EIS) measurements and fitting, modified pulse voltammetry (mPV) measurements, surface Cu fraction post-CO₂RR correlation with C₂₊ selectivity, and XPS data and analysis. See DOI: https://doi.org/10.1039/d6ta02615j.

Acknowledgements

The authors acknowledge funding support from A*STAR through the Accelerated Catalyst Development Platform (Grant Reference No. A19E9a0103), the Advanced Manufacturing and Engineering Programmatic Fund (Grant Reference No. A1898b0043), and the HTCO Seed Funding (Grant Reference No. C231218004). This research used the TEM resources and facilities at the A*STAR Institute of Materials Research and Engineering (A*STAR IMRE). The authors acknowledge funding support from the EPSRC (Project number EP/T024836/1) and the UCL Department of Chemistry. SA acknowledges the A*STAR Research Attachment Program (Award number: HQ/R25-ARAP210201). The authors would also like to acknowledge Dr Debbie Seng Hwee Leng (A*STAR, IMRE) for performing the ex situ XPS experiments. The authors also acknowledge Mr Ng Fu Song and Mr Ong Wai Chung (A*STAR, ISCE²) for ICP-OES experiments.

References

1. F. Kesicki and A. Yanagisawa, Modelling the potential for industrial energy efficiency in IEA's World Energy Outlook, Energy Effic., 2015, 8(1), 155–169.
2. M. Hertzberg and H. Schreuder, Role of atmospheric carbon dioxide in climate change, Energy Environ., 2016, 27(6–7), 785–797.
3. S. Nitopi, E. Bertheussen, S. B. Scott, X. Liu, A. K. Engstfeld, S. Horch, B. Seger, I. E. L. Stephens, K. Chan, C. Hahn, J. K. Nørskov, T. F. Jaramillo and I. Chorkendorff, Progress and Perspectives of Electrochemical CO₂ Reduction on Copper in Aqueous Electrolyte, Chem. Rev., 2019, 119(12), 7610–7672.
4. A. D. Handoko, F. Wei, Jenndy, B. S. Yeo and Z. W. Seh, Understanding heterogeneous electrocatalytic carbon dioxide reduction through operando techniques, Nat. Catal., 2018, 1(12), 922–934.
5. M. Gotz, J. Lefebvre, F. Mors, A. M. Koch, F. Graf, S. Bajohr, R. Reimert and T. Kolb, Renewable Power-to-Gas: A technological and economic review, Renewable Energy, 2016, 85, 1371–1390.
6. B. L. Smith, M. Woodhouse, K. A. W. Horowitz, T. J. Silverman, J. Zuboy and R. M. Margolis, Photovoltaic (PV) Module Technologies: 2020 Benchmark Costs and Technology Evolution Framework Results, National Renewable Energy Laboratory (NREL), Golden, CO (United States), vol. 1, 2021.
7. R. H. Wiser, M. Bolinger, B. Hoen, D. Millstein, J. Rand, G. L. Barbose, N. R. Darghouth, W. Gorman, S. Jeong, A. D. Mills and B. Paulos, Wind Energy Technology Data Update: 2020 Edition; Energy Markets and Planning Department, Lawrence Berkeley National Laboratory, Berkeley, CA, USA, 2020.
8. M. Jouny, W. Luc and F. Jiao, General Techno-Economic Analysis of CO₂ Electrolysis Systems, Ind. Eng. Chem. Res., 2018, 57(6), 2165–2177.
9. A. D. Handoko, H. H. Khoo, I. Halim and K. Murugappan, Feasibility of CO₂ Capture and Utilization: From the LCA Perspective, in Life Cycle Assessment, ed. H. H. Khoo and R. B. H. Tan, World Scientific Publishing Co. Pte. Ltd, Singapore, 2022, pp. 39–53.
10. Y. Hori, H. Wakebe, T. Tsukamoto and O. Koga, Electrocatalytic process of CO selectivity in electrochemical reduction of CO₂ at metal electrodes in aqueous media, Electrochim. Acta, 1994, 39(11–12), 1833–1839.
11. K. P. Kuhl, E. R. Cave, D. N. Abram and T. F. Jaramillo, New insights into the electrochemical reduction of carbon dioxide on metallic copper surfaces, Energy Environ. Sci., 2012, 5(5), 7050–7059.
12. A. Salehi-Khojin, H.-R. M. Jhong, B. A. Rosen, W. Zhu, S. Ma, P. J. A. Kenis and R. I. Masel, Nanoparticle Silver Catalysts That Show Enhanced Activity for Carbon Dioxide Electrolysis, J. Phys. Chem. C, 2013, 117(4), 1627–1632.
13. L. R. L. Ting, O. Piqué, S. Y. Lim, M. Tanhaei, F. Calle-Vallejo and B. S. Yeo, Enhancing CO₂ Electroreduction to Ethanol on Copper–Silver Composites by Opening an Alternative Catalytic Pathway, ACS Catal., 2020, 10(7), 4059–4069.
14. K. Qi, Y. Zhang, N. Onofrio, E. Petit, X. Cui, J. Ma, J. Fan, H. Wu, W. Wang, J. Li, J. Liu, Y. Zhang, Y. Wang, G. Jia, J. Wu, L. Lajaunie, C. Salameh and D. Voiry, Unlocking direct CO₂ electrolysis to C₃ products via electrolyte supersaturation, Nat. Catal., 2023, 6(4), 319–331.
15. J. Gao, H. Zhang, X. Guo, J. Luo, S. M. Zakeeruddin, D. Ren and M. Grätzel, Selective C–C Coupling in Carbon Dioxide Electroreduction via Efficient Spillover of Intermediates As Supported by Operando Raman Spectroscopy, J. Am. Chem. Soc., 2019, 141(47), 18704–18714.
16. A. Herzog, A. Bergmann, H. S. Jeon, J. Timoshenko, S. Kühl, C. Rettenmaier, M. Lopez Luna, F. T. Haase and B. Roldan Cuenya, Operando Investigation of Ag-Decorated Cu₂O Nanocube Catalysts with Enhanced CO₂ Electroreduction toward Liquid Products, Angew. Chem., Int. Ed., 2021, 60(13), 7426–7435.
17. E. L. Clark, C. Hahn, T. F. Jaramillo and A. T. Bell, Electrochemical CO₂ Reduction over Compressively Strained CuAg Surface Alloys with Enhanced Multi-Carbon Oxygenate Selectivity, J. Am. Chem. Soc., 2017, 139(44), 15848–15857.
18. M. L. Frisch, L. Wu, C. Atlan, Z. Ren, M. Han, R. Tucoulou, L. Liang, J. Lu, A. Guo, H. N. Nong, A. Arinchtein, M. Sprung, J. Villanova, M.-I. Richard and P. Strasser, Unraveling the synergistic effects of Cu–Ag tandem catalysts during electrochemical CO₂ reduction using nanofocused X-ray probes, Nat. Commun., 2023, 14(1), 7833.
19. S. S. Balakrishna and A. K. Mallik, Synthesis of azeotropic and two terminal solid solution type binary diagrams, Calphad, 1979, 3(2), 109–118.
20. M. K. Bhat, P. T. Sukumar, L. Langenohl, J. P. Best and G. Dehm, Role of Ag segregation on microscale strengthening and slip transmission in an asymmetric Σ5 copper grain boundary, Acta Mater., 2023, 255, 119081.
21. F. Sansoz and X. Ke, Hall–Petch strengthening limit through partially active segregation in nanocrystalline Ag-Cu alloys, Acta Mater., 2022, 225, 117560.
22. G. A. Kamat, C. Yan, W. T. Osowiecki, I. A. Moreno-Hernandez, M. Ledendecker and A. P. Alivisatos, Self-Limiting Shell Formation in Cu@Ag Core–Shell Nanocrystals during Galvanic Replacement, J. Phys. Chem. Lett., 2020, 11(13), 5318–5323.
23. J. Zhang, Q. Shao, P. Wang, J. Guo and X. Huang, Catalytic Hydrogen Production by Janus CuAg Nanostructures, ChemNanoMat, 2018, 4(5), 477–481.
24. P.-C. Chen, M. Gao, C. A. Mccandler, C. Song, J. Jin, Y. Yang, A. L. Maulana, K. A. Persson and P. Yang, Complete miscibility of immiscible elements at the nanometre scale, Nat. Nanotechnol., 2024, 19(6), 775–781.
25. J.-L. Sang, S.-D. Zhang, Q. Liu, R. A. Geioushy, T. Mahmoud, Y. Zhang and M. Liu, Liquid-Liquid Interface -Driven Reconstruction of CuAg Nanocomposites for Selective CO₂ to C₂H₄ Electroreduction, Small, 2025, e08803.
26. I. Kim, G.-B. Lee, S. Kim, H. D. Jung, J.-Y. Kim, T. Lee, H. Choi, J. Jo, G. Kang, S.-H. Oh, W. Kwon, D. Hong, H. G. Kim, Y. Lee, U. Kim, H. Kim, M. Kim, S. Back, J. Park, Y.-C. Joo and D.-H. Nam, Unveiling the reconstruction of copper bimetallic catalysts during CO₂ electroreduction, Nat. Catal., 2025, 8(7), 697–713.
27. S. Mu, H. Lu, Q. Wu, L. Li, R. Zhao, C. Long and C. Cui, Hydroxyl radicals dominate reoxidation of oxide-derived Cu in electrochemical CO₂ reduction, Nat. Commun., 2022, 13(1), 3694.
28. Y. Ma, J. Yu, M. Sun, B. Chen, X. Zhou, C. Ye, Z. Guan, W. Guo, G. Wang, S. Lu, D. Xia, Y. Wang, Z. He, L. Zheng, Q. Yun, L. Wang, J. Zhou, P. Lu, J. Yin, Y. Zhao, Z. Luo, L. Zhai, L. Liao, Z. Zhu, R. Ye, Y. Chen, Y. Lu, S. Xi, B. Huang, C. S. Lee and Z. Fan, Confined Growth of Silver–Copper Janus Nanostructures with {100} Facets for Highly Selective Tandem Electrocatalytic Carbon Dioxide Reduction, Adv. Mater., 2022, 34(19), 2110607.
29. Y. Yu, D. Wang, Y. Hong, T. Zhang, C. Liu, J. Chen, G. Qin and S. Li, Bulk-immiscible CuAg alloy nanorods prepared by phase transition from oxides for electrochemical CO₂ reduction, Chem. Commun., 2022, 58(79), 11163–11166.
30. D. Wang, H. D. Jung, S. Liu, J. Chen, H. Yang, Q. He, S. Xi, S. Back and L. Wang, Revealing the structural evolution of CuAg composites during electrochemical carbon monoxide reduction, Nat. Commun., 2024, 15(1), 4692.
31. H. Dai, S. Dimitriadou, P. S. Sankara Rama Krishnan, A. D. Handoko, J. Recatala-Gomez, Y. Wang, D. V. M. Repaka, M. Thway, C. Zhang, M. Duchamp and K. Hippalgaonkar, Sub-10 nm Mixing and Alloying of Cu–Ag and Cu–Ni via Accelerated Solid Diffusion, ACS Appl. Mater. Interfaces, 2023, 15(23), 28398–28409.
32. F. Lockwood Estrin, O. S. J. Hagger, M. E. Sener and D. J. Caruana, Metal Painting by Plasma Jet, Adv. Mater. Interfaces, 2024, 11(25), 2400256.
33. O. S. J. Hagger, M. E. Sener, I. Khan, F. L. Estrin, S. Agrotis, A. D. Handoko, I. P. Parkin and D. J. Caruana, Rapid single step atmospheric pressure plasma jet deposition of a SERS active surface, Mater. Adv., 2023, 4(15), 3239–3245.
34. O. S. J. Hagger, M. A. Parkes, F. Lockwood Estrin, S. Agrotis, I. P. Parkin, A. D. Handoko and D. J. Caruana, Additive metal printing on multi materials using an atmospheric pressure plasma jet on a 5-Axis platform, Mater. Des., 2025, 251, 113681.
35. S. Agrotis, M. Emre Sener, O. S. J. Hagger, A. D. Handoko and D. J. Caruana, One-Step synthesis of nanosized Cu–Ag films using atmospheric pressure plasma jet, Appl. Mater. Today, 2024, 39, 102286.
36. A. D. Handoko, C. W. Ong, Y. Huang, Z. G. Lee, L. Lin, G. B. Panetti and B. S. Yeo, Mechanistic Insights into the Selective Electroreduction of Carbon Dioxide to Ethylene on Cu₂O-Derived Copper Catalysts, J. Phys. Chem. C, 2016, 120(36), 20058–20067.
37. T. Cheng, H. Xiao and W. A. Goddard, Full atomistic reaction mechanism with kinetics for CO reduction on Cu(100) from ab initio molecular dynamics free-energy calculations at 298 K, Proc. Natl. Acad. Sci. U. S. A., 2017, 114(8), 1795–1800.
38. R. Kortlever, J. Shen, K. J. P. Schouten, F. Calle-Vallejo and M. T. M. Koper, Catalysts and Reaction Pathways for the Electrochemical Reduction of Carbon Dioxide, J. Phys. Chem. Lett., 2015, 6(20), 4073–4082.
39. J. Zhang, G. Zeng, S. Zhu, H. Tao, Y. Pan, W. Lai, J. Bao, C. Lian, D. Su, M. Shao and H. Huang, Steering CO₂ electroreduction pathway toward ethanol via surface-bounded hydroxyl species-induced noncovalent interaction, Proc. Natl. Acad. Sci. U. S. A., 2023, 120(11), e2218987120.
40. C. Y. J. Lim, M. Yilmaz, J. M. Arce-Ramos, A. D. Handoko, W. J. Teh, Y. Zheng, Z. H. J. Khoo, M. Lin, M. Isaacs, T. L. D. Tam, Y. Bai, C. K. Ng, B. S. Yeo, G. Sankar, I. P. Parkin, K. Hippalgaonkar, M. B. Sullivan, J. Zhang and Y.-F. Lim, Surface charge as activity descriptors for electrochemical CO₂ reduction to multi-carbon products on organic-functionalised Cu, Nat. Commun., 2023, 14(1), 335.
41. Z. Liu, L. Song, X. Lv, M. Liu, Q. Wen, L. Qian, H. Wang, M. Wang, Q. Han and G. Zheng, Switching CO₂ Electroreduction toward Ethanol by Delocalization State-Tuned Bond Cleavage, J. Am. Chem. Soc., 2024, 146(20), 14260–14266.
42. Y. Lin, T. Wang, L. Zhang, G. Zhang, L. Li, Q. Chang, Z. Pang, H. Gao, K. Huang, P. Zhang, Z.-J. Zhao, C. Pei and J. Gong, Tunable CO₂ electroreduction to ethanol and ethylene with controllable interfacial wettability, Nat. Commun., 2023, 14(1), 3575.
43. H. Tang, Y. Liu, Y. Zhou, Y. Qian and B.-L. Lin, Boosting the Electroreduction of CO₂ to Ethanol via the Synergistic Effect of Cu–Ag Bimetallic Catalysts, ACS Appl. Energy Mater., 2022, 5(11), 14045–14052.
44. P. Wang, H. Yang, C. Tang, Y. Wu, Y. Zheng, T. Cheng, K. Davey, X. Huang and S.-Z. Qiao, Boosting electrocatalytic CO₂–to–ethanol production via asymmetric C–C coupling, Nat. Commun., 2022, 13(1), 3754.
45. X. Gao, Y. Jiang, J. Liu, G. Shi, C. Yang, Q. Xu, Y. Yun, Y. Shen, M. Chang, C. Zhu, T. Lu, Y. Wang, G. Du, S. Li, S. Dai and L. Zhang, Intermediate-regulated dynamic restructuring at Ag-Cu biphasic interface enables selective CO₂ electroreduction to C₂₊ fuels, Nat. Commun., 2024, 15(1), 10331.
46. M. Mavrikakis, B. Hammer and J. K. Nørskov, Effect of Strain on the Reactivity of Metal Surfaces, Phys. Rev. Lett., 1998, 81(13), 2819–2822.
47. P. J. Bruggeman, F. Iza and R. Brandenburg, Foundations of atmospheric pressure non-equilibrium plasmas, Plasma Sources Sci. Technol., 2017, 26(12), 123002.
48. P. D. Maguire, C. M. O. Mahony, C. P. Kelsey, A. J. Bingham, E. P. Montgomery, E. D. Bennet, H. E. Potts, D. C. E. Rutherford, D. A. McDowell, D. A. Diver and D. Mariotti, Controlled microdroplet transport in an atmospheric pressure microplasma, Appl. Phys. Lett., 2015, 106(22), 224101.
49. P. Maguire, D. Rutherford, M. Macias-Montero, C. Mahony, C. Kelsey, M. Tweedie, F. Pérez-Martin, H. McQuaid, D. Diver and D. Mariotti, Continuous In-Flight Synthesis for On-Demand Delivery of Ligand-Free Colloidal Gold Nanoparticles, Nano Lett., 2017, 17(3), 1336–1343.
50. C. Langlois, Z. L. Li, J. Yuan, D. Alloyeau, J. Nelayah, D. Bochicchio, R. Ferrando and C. Ricolleau, Transition from core–shell to Janus chemical configuration for bimetallic nanoparticles, Nanoscale, 2012, 4(11), 3381.
51. W.-L. Weng, C.-Y. Hsu, J.-S. Lee, H.-H. Fan and C.-N. Liao, Twin-mediated epitaxial growth of highly lattice-mismatched Cu/Ag core–shell nanowires, Nanoscale, 2018, 10(21), 9862–9866.
52. A. Kirfel and K. Eichhorn, Accurate structure analysis with synchrotron radiation. The electron density in Al₂O₃ and Cu₂O, Acta Crystallogr., Sect. A, 1990, 46(4), 271–284.
53. R. W. G. Wyckoff, Crystal Structures, Interscience Publishers, New York, 1963, vol. 1.
54. L. T. Viyannalage, R. Vasilic and N. Dimitrov, Epitaxial Growth of Cu on Au(111) and Ag(111) by Surface Limited Redox ReplacementAn Electrochemical and STM Study, J. Phys. Chem. C, 2007, 111(10), 4036–4041.
55. D. Ren, Y. Deng, A. D. Handoko, C. S. Chen, S. Malkhandi and B. S. Yeo, Selective Electrochemical Reduction of Carbon Dioxide to Ethylene and Ethanol on Copper(I) Oxide Catalysts, ACS Catal., 2015, 5(5), 2814–2821.
56. M. C. Biesinger, Advanced analysis of copper X-ray photoelectron spectra, Surf. Interface Anal., 2017, 49(13), 1325–1334.
57. B. Tomc, M. Bele, M. Plut, M. Kostelec, S. Popović, M. A. Nazrulla, F. Ruiz-Zepeda, A. R. Kamšek, M. Šala, A. Elbataioui, L. D. Rafailović, Y. B. Pissolitto, F. Trivinho-Strixino, W. J. Stępniowski, L. Suhadolnik and N. Hodnik, Recognizing the Universality of Copper Reconstruction Via Dissolution–Redeposition at the Onset of CO₂ Reduction, J. Phys. Chem. Lett., 2025, 16(36), 9553–9560.
58. S. Anantharaj and S. Noda, Appropriate Use of Electrochemical Impedance Spectroscopy in Water Splitting Electrocatalysis, ChemElectroChem, 2020, 7(10), 2297–2308.
59. R. D. Armstrong and M. Henderson, Impedance plane display of a reaction with an adsorbed intermediate, J. Electroanal. Chem. Interfacial Electrochem., 1972, 39(1), 81–90.
60. D. A. Harrington and B. E. Conway, ac Impedance of Faradaic reactions involving electrosorbed intermediates—I. Kinetic theory, Electrochim. Acta, 1987, 32(12), 1703–1712.
61. Y. Huang, A. D. Handoko, P. Hirunsit and B. S. Yeo, Electrochemical Reduction of CO₂ Using Copper Single-Crystal Surfaces: Effects of CO* Coverage on the Selective Formation of Ethylene, ACS Catal., 2017, 7(3), 1749–1756.
62. I. Tezsevin, M. C. M. van de Sanden and S. Er, Surface charging activated mechanism change: A computational study of O, CO, and CO₂ interactions on Ag electrodes, J. Energy Chem., 2020, 50, 307–313.
63. J. L. Kenty and J. P. Hirth, Epitaxy and heterogeneous nucleation theory, Surf. Sci., 1969, 15(3), 403–424.
64. Q. Lei, L. Huang, J. Yin, B. Davaasuren, Y. Yuan, X. Dong, Z.-P. Wu, X. Wang, K. X. Yao, X. Lu and Y. Han, Structural evolution and strain generation of derived-Cu catalysts during CO₂ electroreduction, Nat. Commun., 2022, 13(1), 4857.
65. S. I. Bogatyrenko, A. P. Kryshtal and A. Kruk, Effect of Size on the Formation of Solid Solutions in Ag–Cu Nanoparticles, J. Phys. Chem. C, 2023, 127(5), 2569–2580.
66. J. H. Montoya, A. A. Peterson and J. K. Nørskov, Insights into C-C Coupling in CO₂ Electroreduction on Copper Electrodes, ChemCatChem, 2013, 5(3), 737–742.
67. C. W. Li, J. Ciston and M. W. Kanan, Electroreduction of carbon monoxide to liquid fuel on oxide-derived nanocrystalline copper, Nature, 2014, 508(7497), 504–507.

---