This article is licensed under CC-BY 4.

pubs.acs.org/acscatalysis Research Article

Synthesis and Catalytic Performance of Bimetallic Oxide-Derived

CuO−ZnO Electrocatalysts for CO2 Reduction
Matt L. J. Peerlings, Kai Han, Alessandro Longo, Kristiaan H. Helfferich, Mahnaz Ghiasi,
Petra E. de Jongh,* and Peter Ngene*
Cite This: ACS Catal. 2024, 14, 10701−10711 Read Online

ACCESS Metrics & More Article Recommendations *

sı Supporting Information
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

ABSTRACT: Steering the selectivity of electrocatalysts toward the

desired product is crucial in the electrochemical reduction of CO2.
A promising approach is the electronic modification of the catalyst’s
Downloaded via Nikhil Bhoumik on November 13, 2024 at 04:11:05 (UTC).

active phase. In this work, we report on the electronic modification

effects on CuO−ZnO-derived electrocatalysts synthesized via
hydrothermal synthesis. Although the synthesis method yields
spatially separated ZnO nanorods and distinct CuO particles,
strong restructuring and intimate atomic mixing occur under the
reaction conditions. This leads to interactions that have a profound
effect on the catalytic performance. Specifically, all of the bimetallic
electrodes outperformed the monometallic ones (ZnO and CuO)
in terms of activity for CO production. Surprisingly, on the other
hand, the presence of ZnO suppresses the formation of ethylene on
Cu, while the presence of Cu improves CO production of ZnO. In situ X-ray absorption spectroscopy studies revealed that this
catalytic effect is due to enhanced reducibility of ZnO by Cu and stabilization of cationic Cu species by the intimate contact with
partially reduced ZnO. This suppresses ethylene formation while favoring the production of H2 and CO on Cu. These results show
that using mixed metal oxides with different reducibilities is a promising approach to alter the electronic properties of electrocatalysts
(via stabilization of cationic species), thereby tuning the electrocatalytic CO2 reduction reaction performance.
KEYWORDS: electrochemical CO2 reduction, bimetallic Cu−Zn catalyst, phase separation, oxidation state, electronic modification,
stability

■ INTRODUCTION
A promising strategy to close the carbon cycle is the
overpotential, and process stability are essential for scaling up
this process.7
Although Ag and Au have shown promising CO2RR
electrochemical CO2 reduction reaction (CO2RR) to chemicals
performance, using these precious metals might not be the
and fuels using renewable electricity.1,2 This is necessary to meet best option for making affordable electrolyzers. Instead, metal−
the Paris accord on climate change, which aims to keep the nitrogen-doped carbon materials might be more suitable due to
global temperature rise well below 2 °C above preindustrial their use of earth-abundant materials, low CO2RR to CO
levels.3 To date, a wide variety of electrochemical CO2 reduction overpotentials, and high current densities.8,9 A different cost-
products can be generated depending on the catalyst material affordable option is the use of Zn because it is an earth-abundant
used. Different metals are reported to be active CO2RR and cheap metal.4 Despite good CO2RR to CO selectivity, Zn-
electrocatalysts, and they have been categorized based on their based CO2RR electrodes are however less attractive due to high
main products. These products include CO (on Au, Ag, Zn), required overpotentials for selective CO production, low
formate (on Sn, Pb, Bi), and hydrocarbons (on Cu).4 In the catalyst activity, and stability issues.10 Therefore, it is important
latter case, CO is identified as the key intermediate for further that approaches are developed to increase their catalytic
hydrocarbon production.5 Furthermore, carbon monoxide performance in order to make them a viable alternative to
(CO) is widely used as an important feedstock for the
production of methanol and other valuable chemicals. There- Received: March 13, 2024
fore, the electrochemical conversion of CO2 to CO is currently Revised: June 24, 2024
being scaled up for industrial applications using Ag-based Accepted: June 25, 2024
electrodes.6 However, the process still suffers from selectivity Published: July 2, 2024
issues, mainly due to the competing hydrogen evolution reaction
(HER). Furthermore, improvements in terms of current density,
© 2024 The Authors. Published by
American Chemical Society https://doi.org/10.1021/acscatal.4c01575
10701 ACS Catal. 2024, 14, 10701−10711
ACS Catalysis pubs.acs.org/acscatalysis Research Article

other catalyst materials. On the other hand, although Cu makes These studies demonstrate that combining oxide-derived Zn
CO and other hydrocarbon products at moderate potentials, it and Cu is a promising approach to tune the catalyst selectivity.
suffers from poor selectivity toward any one of the desired Furthermore, the catalyst structure plays an important role,
products.11,12 although much is still unclear about the impact of restructuring
Thus, far, several approaches have been made to tune the on the electrocatalytic performance. In this study, we focus on
binding strength of reaction intermediates and, hence, the the catalytic performance of oxide-derived CuO−ZnO catalysts
electrocatalytic performance of Zn- and Cu-based electrodes. prepared via a hydrothermal synthesis method. Remarkably,
Two of these approaches include tailoring the catalyst structure although the synthesis method led to structurally distinct and
and using oxide-derived catalysts.13 For instance, by manipulat- spatially separated CuO and ZnO, the oxides undergo massive
ing the structure of the Zn catalyst into a hierarchical hexagonal restructuring and intermixing during CO2RR. The resulting
shape, a high CO2 to CO Faradaic Efficiency (FE) of 95% (at close contact between copper and zinc affects their reducibility
−1.05 V in 0.5 M KCl) could be achieved, as well as a high and thereby leads to a significant change in the product
stability (5% activity loss in 30 h).14 A study by Luo et al.10 found selectivity due to the stabilization of cationic Cu species by
that when using oxide-derived Zn, regardless of the initial partially reduced ZnO, and enhanced reducibility of the ZnO by
structure, the ZnO undergoes severe restructuring to porous Cu. Therefore, our work reveals that the stabilization of cationic
structures composed of hexagonal Zn crystals. These oxide- metal species by the presence of another metal oxide with less
derived reconstructed catalysts displayed a high CO FE above reducibility is a promising approach to tune the catalytic
90% and stability of more than 18 h, as well as a similar intrinsic performance of CO2 reduction electrocatalysts.
CO activity after normalization for their electrochemical surface
area (ECSA).10 In a different study, Geng et al. showed that the
introduction of oxygen vacancies via a H2 plasma treatment
■ EXPERIMENTAL SECTION
Cu1−xZnxO Preparation. The bimetallic Cu1−xZnxO
could increase both the CO FE from 44 to 83%, as well as the catalysts were grown on carbon paper (TGP-H-060) using a
CO partial current density by a factor of 5.1, compared to modified hydrothermal method based on previous work for
starting from pristine ZnO nanosheets.15 More recently, we ZnO.16,24 Typically, a piece of carbon paper disc (d = 2.5 cm)
studied nanorod-shaped ZnO catalysts and demonstrated that was immersed into an aqueous solution containing 0.05 M
the morphology of the resulting Zn catalyst could be tuned via M(NO3)2 (M = Zn or Cu) and 0.05 M hexamethylenetetramine
different reduction procedures. It was found that the final (HMT). The solution was transferred to a Teflon liner and
structure of the active Zn catalysts significantly affected both the sealed in an autoclave reactor. The autoclaves were subsequently
activity and product selectivity for CO2 reduction.16 kept at 100 °C for 5 h. Afterward, the carbon paper was washed
A different approach to improve the performance of CO2RR with Milli-Q water and dried under ambient conditions. By
electrocatalysts is to combine two different metals.17 For change of the ratio of copper to zinc nitrate, catalysts with
example, Feng et al. demonstrated the principle of tandem different atomic compositions were prepared.
catalysis, in which Zn is added to Cu in order to increase the Structural Characterization. XRD measurements were
*CO coverage on Cu and hence promote the formation of performed with a Bruker D2 Phaser, equipped with a Co Kα X-
ethylene.18 Interestingly, however, Ren et al. found that adding ray source (λ = 1.79026 Å). The XRD measurements were
Zn to Cu shifts the selectivity to ethanol instead of ethylene conducted at a 2θ range from 35 to 80° using 1 s as the
integration time. SEM-EDX micrographs and elemental maps
formation.19 Likely, catalyst restructuring and oxidation state
were made on a Zeiss EVO 15 instrument equipped with a
play roles in these seemingly contradicting results. In an effort to
secondary electron detector, operating at 400 pA and 10.0 kV for
disentangle these contributions, da Silva et al. studied copper-
imaging and at 20.0 kV for elemental maps. High-resolution
rich bimetallic CuZn-based catalysts. They showed that the
SEM images were made on a Zeiss Gemini instrument equipped
catalyst morphology had the strongest effect on the selectivity to with an InLens detector and operated at 500 pA and 10.0 kV.
C 2+ products, although other factors such as atomic Bright field transmission electron microscopy (BF-TEM) and
composition, oxidation state, and surface roughness were also high angle annular dark field scanning transmission electron
found to play a role in the overall catalyst performance.20 microscopy (HAADF-STEM) were performed on the Talos
However, C2+ products are not always formed on CuZn-based F200x operated at 200 kV. Energy dispersive X-ray spectroscopy
catalysts, as in the work of Jeon et al.21 They studied well-defined (EDX) was employed to identify and map the present elements.
Cu−Zn nanoparticles of ca. 5 nm with a wide range of atomic Samples were prepared by drop casting onto a gold-coated
compositions. The main products formed were CH4, H2, and carbon grid after the dispersion of the catalysts by ultra-
CO, depending on the atomic composition. Increasing the Zn sonication in ethanol. Inductively coupled plasma (ICP)
content decreased the CH4 and increased syngas (CO/H2) analysis was carried out using a PerkinElmer AAS Model
selectivity, corresponding to the formation of CuZn alloys Analyst 200 at MIKROLAB (Mikroanalytisches Labor Kolbe, c/
instead of nonfully reduced Cu-ZnO.21 Consequently, the o Fraunhofer Institut UMSICHT, Germany).
catalysts also exhibit time-dependent selectivity because at Electrochemical Measurements. The electrochemical
longer reaction times, more cationic Zn species are reduced and experiments were performed in an H-type cell (Figure S1)
CuZn alloy is formed, favoring H2 over CH4 production.21,22 using a PARSTAT MC potentiostat. In the cell, both the
However, the effect of restructuring on time-dependent cathode and anode compartments are filled with 15 mL of 0.1 M
selectivity was not studied in detail. Interesting in this regard KHCO3 electrolyte solution, separated by a Fumasep membrane
is the study by Wan et al.,23 which demonstrates bimetallic Cu− (FAA-3-PK-130, Fumatech BWT, GmbH). The cathode and
Zn catalysts selective toward CO. The CO activity and stability anode compartments were flushed for 30 min before and during
were found to be higher in the case of a phase-separated sample the experiments using 20 mL/min CO2 and Ar, respectively. In
than a core−shell one due to zinc enrichment at the catalyst the cell configuration, the carbon paper-supported catalyst was
surface in the latter case. pressed onto a piece of glassy carbon current collector. A Ag/
10702 https://doi.org/10.1021/acscatal.4c01575
ACS Catal. 2024, 14, 10701−10711
ACS Catalysis pubs.acs.org/acscatalysis Research Article

AgCl (3 M KCl) reference electrode and commercial iridium given by the sum of the n-body contributions g2, g3, ..., gn which
oxide-based counter electrode (Dioxide Materials) were used. takes into account all of the possible single and multiple
The exposed geometric area of both the working and counter scattering (MS) paths between the n atoms. The modeling of
electrodes is 3.8 cm2. c(k) to the experimental EXAFS signal allowed us to refine the
All electrodes were tested at five increasingly cathodic relevant structural parameters of the different coordination
potentials for two cycles, of which the second cycle was used shells. The quality of the model is also checked by comparison of
to determine the selectivity. Subsequently, the electrodes were the experimental EXAFS signal Fourier transform (FT) with the
kept at the maximum cathodic potential for long-term testing. FT of the calculated c(k) function. The coordination numbers
All potential values were converted to potentials with respect to and the global fit parameters that were allowed to vary during the
the reversible hydrogen electrode potential (RHE) and iR- fitting procedure were the following: the distance R (Å),
corrected using a typical resistance of around 28 Ω in 0.1 M Debye−Waller factor (s2), and the angles of the gn contributions.
KHCO3 solution as determined from electrochemical impe- The edge energy E0 was fixed at the Cu K-edge (8979 eV) and
dance spectroscopy (EIS) measurements. Typically, 80% iR Zn K-edge (9659 eV) corresponding values.
correction was applied during each measurement, and the
remainder was corrected afterward. All results shown are the
average of three independent measurements. Error bars display
■ RESULTS AND DISCUSSION
Structural Characterization of the Cu1−xZnxO Cata-
the standard deviation. lysts. To assess the elemental composition of the Cu1−xZnxO
Gas Product Detection. Gaseous products were analyzed electrodes prepared using hydrothermal synthesis, inductively
using a Global Analysis Solutions Microcompact GC 4.0. The coupled plasma (ICP) measurements were carried out. The
GC system was equipped with 3 detector channels (2 FID and 1 results are summarized in Table 1. From these results, it follows
TCD). The first channel has a Rt-Q Bond (10 m × 0.32 mm,
Agilent) packed column and an FID detector for the detection of Table 1. ICP Results of the Electrodesa
CH4, C2H4, and C2H6; the second channel has a Molecular Sieve
5A (10 m × 0.53 mm, Restek) packed column that separates Cu:Zn atom %
small gaseous molecules such as CO and CH4. This channel has in on
an FID detector with a methanizer to increase the detection sample solution electrode metal loading on C paper (wt %)
sensitivity of CO. The third channel has a Carboxen 1010 (8 m ZnO 0 0 2.9
× 0.32 mm, Agilent) packed column which separates H2 and Cu0.14Zn0.86O 50 14 1.8
CO2 with a TCD. High purity nitrogen (N2, 99.999%) was used Cu0.43Zn0.57O 67 43 1.5
as carrier gas. Cu0.75Zn0.25O 75 75 1.1
Liquid Product Detection. Products in the liquid phase CuO 100 100 0.8
a
were analyzed by taking 1 mL of electrolyte samples after each Indicating the atom % of Cu with respect to Zn in the hydrothermal
potential step and replacing this with fresh 0.1 M KHCO3. 100 synthesis precursor solution and on the fabricated electrodes;
μL amount of internal standard solution containing 10 mM additionally, the metal oxide weight loadings on the electrodes are
DMSO and 50 mM phenol in D2O was added to each 500 μL reported.
electrolyte sample and analyzed using 1H NMR with solvent
suppression on a 400 MHz VNMRS-400 Varian NMR. Products that a series of electrodes with systematically varying Cu:Zn
were quantified by comparing the product and internal standard ratios have been prepared. Because some copper precipitation
peak areas and correcting for sampling volume. was already observed upon mixing the reagents, the obtained
In Situ XAS Measurements. X-ray absorption measure- Cu:Zn ratio is lower than the precursor ratio in solution, and the
ments were performed at the LISA beamline (BM 08) of the total catalyst loading decreases slightly with addition of a higher
European Synchrotron Radiation Facility (ESRF) in Grenoble, relative amount of copper.
France. For the operando XAS measurements, a homemade XAS To further evaluate the structure of the Cu1−xZnxO catalysts,
electrochemical cell was used (Figure S2). A CO2-saturated 0.1 scanning electron microscopy (SEM) images were made as
M KHCO3 electrolyte was flown through the cell with a flow rate shown in Figures 1 and S12. For ZnO particles, a rod-like shape
of 3 mL/min using a peristaltic pump. The catalysts were used as is observed, in line with previous results.16,24 In contrast, CuO is
the working electrode, a platinum wire as counter electrode, and present in a less well-defined shape. These results agree with the
a Ag/AgCl 3 M KCl electrode as reference electrode. A BioLogic XRD patterns in Figure S3, which show high crystallinity and a
SP-50 potentiostat was used for all in situ XAS experiments. No preferential growth direction for ZnO, in contrast to weak
iR correction was applied. diffraction lines for CuO. In the bimetallic samples, both rod-
All in situ XAS experiments were performed in fluorescence shaped particles and less faceted particles are observed. In
mode with an angle of 45° between the incoming X-rays and the agreement with the findings for the monometallic electrodes, the
sample. Spectra were recorded at the Cu K-edge (8979 eV) and elemental maps confirm that the rod-shaped particles contain
Zn K-edge (9659 eV), respectively. The time to acquire a mainly zinc, whereas the less faceted particles contain mostly
spectrum was 80 min for recording both the Cu and Zn spectra copper (Figure 2). Furthermore, the images indicate that copper
at each potential. Reference spectra of ZnO, Zn, CuO, and Cu and zinc are not mixed homogeneously, but rather they are
were acquired in transmission mode. Both Athena and GNXAS present in separate structures. Indeed, it is known that bulk CuO
data processing software were used for the analysis of data. The and ZnO do not form mixed phases,18 unlike their metallic
spectra (two scans per sample) were energy-calibrated, counterparts which can coexist in different compositions.20
averaged, and further analyzed using GNXAS.25 In this Electrochemical Performance of the Cu1−xZnxO Cata-
approach, the local atomic arrangement around the absorbing lysts. All electrodes were tested for two cycles at increasingly
atom is decomposed into model atomic configurations cathodic potentials, between −0.5 and −1.1 V vs RHE. Figure
containing 2, ..., n atoms. The theoretical EXAFS signal c(k) is 3A shows a typical chronoamperometry measurement of the
10703 https://doi.org/10.1021/acscatal.4c01575
ACS Catal. 2024, 14, 10701−10711
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Figure 1. SEM images of freshly prepared Cu1−xZnxO electrodes. The ZnO particles have a rod-like shape, whereas CuO particles are more
amorphous.

Figure 2. SEM-EDX elemental maps of freshly prepared Cu1−xZnxO-based electrodes showing the Cu (orange) and Zn (blue) elemental distributions.
Zn is located in the rod-shaped particles, whereas Cu is mainly present in the more amorphous ones.

Cu0.14Zn0.86O electrode. The current is stable at each potential formed mainly on the CuO and Cu0.75Zn0.25O electrodes as
within the 30 min testing time. Only at the start of the shown in Figure S5 and summarized in Table S2.
measurement an increased cathodic current is observed. This is From Figure 3C, it follows that the activity for H2 production
likely related to the reduction of copper and zinc oxide species, depends strongly on the Cu:Zn ratio, with a higher relative
which is thermodynamically favorable in this potential range. copper content generally resulting in more H2 production. The
Since this feature is much less pronounced in the ZnO sample only exception is the pure CuO electrode, which shows a lower
(Figure S4), the main contribution here arises from copper H2 partial current density at more cathodic potentials than the
oxide reduction. This implies that copper oxide reduction takes Cu0.75Zn0.25O electrode. These results explain to a large extent
place on much shorter time scales than zinc oxide reduction. why the total cathodic current in Figure 3B increases with a
Figure 3B shows a comparison of the current densities of the higher relative Cu content, whereas the total current is similar
different catalysts at a similar applied potential of −0.9 V vs for the CuO and Cu0.75Zn0.25O electrodes.
RHE. All current densities are normalized by the mass of copper On the other hand, the activity for CO production as shown in
and zinc metal present on the electrodes to account for the Figure 3D follows a different trend. Upon addition of copper to
slightly different loadings summarized in Table 1. A clear trend ZnO, the CO activity increases, with the Cu0.43Zn0.57O electrode
can be observed, with a higher relative Cu to Zn content showing an optimum production of CO. This effect does not
resulting in a higher cathodic current density. This shows that persist for higher Cu:Zn ratios, which show lower CO partial
copper is a more active reduction catalyst than zinc. The current densities with increasing relative copper content.
geometric current densities without normalization by metal Overall, all bimetallic electrodes show a higher CO activity
weight are given in Figures S6−S8. Note that the main difference than the monometallic CuO and ZnO electrodes when
upon normalization is for the CuO electrode, which shifts the normalized by catalyst metal weight loading. The optimum
most since it has the lowest metal weight loading. CO activity for the Cu0.43Zn0.57O electrode regardless of
Although the total current density is a good descriptor of the normalization is remarkable, and indicates the presence of a
catalyst activity, it provides no information on the type of synergistic interaction between copper and zinc.
products formed. For all electrodes, the main reaction products The selectivity of the different catalysts at −0.95 V as a
are H2 and CO. Figure 3C,3D show their respective partial function of Cu:Zn ratio are summarized in Figure 4. As
current densities as a function of potential. Other reaction previously discussed, H2 and CO are the main products
products include CH4, C2H4, and a number of liquid products observed for all of the catalysts. The only exception is the pure
like formate, ethanol, acetate, and 1-propanol, all of which are CuO catalyst, which has a low CO FE but instead produces
10704 https://doi.org/10.1021/acscatal.4c01575
ACS Catal. 2024, 14, 10701−10711
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Figure 3. Electrochemical evaluation of Cu1−xZnxO catalysts. (A) Chronoamperometry data of Cu0.14Zn0.86O at five consecutive applied potentials.
The indicated lines are an average of three independent measurements, with the shaded areas indicating the standard deviation. (B) Current density of
all electrodes at −0.9 V vs RHE. Partial current densities to H2 (C) and CO (D) of all electrodes vs iR-corrected potential. All currents are normalized
by geometric surface area and metal weight.

hydrocarbon products such as CH4 and C2H4 to a significant

extent. It is also clear from this graph that the higher the relative
Zn content, the higher the selectivity to CO.
Interestingly, at this potential, the FE for CH4, C2H4, and
other C2+ products does not amount to more than 5% for the
bimetallic catalysts, whereas it is significantly higher for the pure
CuO electrode. Most notable is that the presence of ZnO
suppresses the formation of C2H4. This can also be observed in
Table S2, which shows that the onset potential for C2H4
production is more cathodic for the bimetallic Cu0.75Zn0.25O
catalyst (−0.85 V) than that for CuO (−0.73 V). Additionally,
Figure S5B shows that the formation of C2H4 is larger on CuO
than that on the bimetallic catalysts at similar potentials.
Surprisingly, the selectivity to C2+ oxygenates like acetate,
ethanol and 1-propanol is also higher on CuO than on the
bimetallic catalysts, in contrast to other studies.19,20
This is remarkable, as this finding contradicts what is expected
based on CO-spillover effects. It also indicates that although the
electrodes consisted of spatially separated CuO and ZnO before
Figure 4. Faradaic efficiency as a function of Cu:Zn ratio at −0.95 V. catalysis, both metals are able to interact during the CO2RR and
Other C2+ products include ethanol, acetate, and 1-propanol. are hence in intimate contact, implying mobility of Zn and/or
Cu species during catalysis. These effects will be discussed in
further detail in the following sections.
10705 https://doi.org/10.1021/acscatal.4c01575
ACS Catal. 2024, 14, 10701−10711
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Figure 5. Stability tests of all electrodes at −1.0 V, showing the evolution of the CO (A) and CH4 (B) FE over time.

Figure 6. SEM images (top) and corresponding EDX maps (bottom) of the three bimetallic CuO−ZnO catalysts after testing, with Cu indicated in
orange and Zn in blue.

Catalyst Stability. It is well-known that both Cu- and Zn- on ZnO. After reduction, a more stable catalyst active state is
based electrocatalysts are not structurally stable during the formed, resulting in the observed stable CO selectivity.
CO2RR. Therefore, it is important to also investigate the Looking at the evolution of the CH4 FE over time, it is clear
changes in both the structure and catalytic performance as a that the CuO-rich catalysts exhibit a less stable catalytic
function of time. To this end, stability studies were performed at performance in terms of their product selectivity toward CH4
a fixed potential of −1.0 V. The CO FE of all catalysts are shown than CO. Especially for the CuO electrode, a ca. 3-fold increase
in Figure 5 as a function of time for 15 h. All catalysts show stable of the CH4 FE is observed during 15 h of testing. On the other
CO selectivity over the testing duration. A slight drop is hand, for the Cu0.75Zn0.25 electrode, the CH4 FE is relatively
observed in both the CO and H2 FE of the Cu0.43Zn0.57O more stable over time, whereas the other catalysts show no
electrode in between 6 and 7 h. This might be explained by a significant CH4 production. For CuO, the increase in CH4 FE is
temporary issue with the gas flow, as no change in the accompanied by both a decrease in H2 FE and increase in C2H4
potentiostat data is observed at this time, and the CO:H2 ratio FE in the first hours of testing as shown in Figure S9, and thus an
remains constant. Therefore, this is not an effect of the electrode. improvement in CO2RR efficiency. However, at longer testing
Most notably, ZnO actually shows an increase in the CO FE in times, the increasing FE of CH4 goes at the cost of C2H4. These
the first few hours. This effect is also present for the Zn-rich changes indicate that the CuO catalyst is not stable, and that it is
bimetallic catalysts, although less pronounced. Given that the important to take long-term stability tests into account when
stability studies were carried out after testing at the different evaluating electrocatalyst performance.
cathodic potentials, we think that the most significant changes The structural stability of the catalysts was investigated by ICP
on these catalysts happen relatively quickly at the start of all and SEM-EDX measurements on the electrodes after catalysis.
measurements. These changes are likely related to the reduction The ICP results in Table S3 show no significant change in Cu:Zn
of copper and zinc oxides, accompanied by catalyst restructur- ratio after catalytic testing. However, some loss of metal is
ing. Zinc oxide reduction is slower than copper oxide reduction, observed. This is likely due to particle detachment from the
explaining the longer time required for achieving a stable CO FE binder-free electrodes. We believe metal dissolution is minimal,
10706 https://doi.org/10.1021/acscatal.4c01575
ACS Catal. 2024, 14, 10701−10711
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Figure 7. Cu K-edge (a) and Zn K-edge (b) normalized XANES spectra of fresh CuO−ZnO catalysts and the corresponding normalized XANES
spectra of reference Cu, CuO, Zn, and ZnO materials. Shoulder peak positions of Cu and Zn reference spectra are indicated by yellow dotted lines, of
CuO and ZnO by green ones.

Figure 8. Normalized in situ (A) Cu K-edge and (B) Zn K-edge XANES spectra of Cu0.14Zn0.86O and Cu, CuO, Zn, and ZnO references in CO2-
saturated 0.1 M KHCO3 under CO2RR conditions at potentials of +0.6 to −0.6 V vs RHE. (C) Results of linear combination analysis on Cu K-edge
XANES results of Cu0.14Zn0.86O catalyst for determination of the CuO fraction at different applied potentials.

as the different solubilities of copper and zinc would have based structures, which show that fragmentation and subsequent
affected the Cu:Zn ratio in this case. agglomeration readily take place under CO2RR conditions. This
Figures 6, S11, and S12 show the SEM-EDX images of the likely happens through a dissolution-redeposition mechanism.26
CuO−ZnO catalysts after testing. The results show that both the Recently, it has been argued that the soluble transient copper
CuO and ZnO particles have lost their initial shape. Instead, species that drive electro-reconstruction are copper carbonyls
smaller fragmented particles as well as larger amorphous and oxalates with copper in a +1 oxidation state.28
agglomerates are observed for all electrodes. Additionally, Before catalytic testing, the electrodes consisted of spatially
dendritic structures are observed on the copper-rich separated CuO and ZnO structures. However, because of
Cu0.75Zn0.25O sample. These findings are in line with other significant reconstruction, such a spatial separation is no longer
studies on electro-reconstruction of copper26,27 and zinc10,16- visible in the SEM-EDX maps after catalysis. It should be noted
10707 https://doi.org/10.1021/acscatal.4c01575
ACS Catal. 2024, 14, 10701−10711
ACS Catalysis pubs.acs.org/acscatalysis Research Article

Figure 9. Fourier-transformed (FT) EXAFS data at different potentials for the (A) Cu K-edge and (B) Zn K-edge of the Cu0.14Zn0.86O catalyst under
CO2RR conditions. Note that these data are not phase corrected, meaning that the FT-EXAFS peak positions do not correspond to the true RDF peak
positions and bond lengths. Black lines show the experimental data; red lines show the corresponding fits.

that the spatial resolution of SEM-EDX measurements is limited In Situ XAS Studies on the Bimetallic CuO−ZnO
to around 1−2 μm, making it impossible to assess the proximity Catalysts. To gain further insight into the chemical state and
of copper and zinc on smaller length scales using this technique. structure of the catalysts under operating conditions, in situ X-
For this reason, TEM-EDX maps were acquired, as shown in ray absorption spectroscopy (XAS) measurements were
Figure S13. Interestingly, nanoscale mixing is observed in these performed at the LISA beamline at the ESRF in Grenoble.
images, especially in the cases of the Cu0.14Zn0.86O and First, the ex-situ spectra of the fresh catalysts will be discussed.
Cu0.43Zn0.57O catalysts. These images show the formation of Figure 7a,7b shows the normalized Cu and Zn K-edge X-ray
intimately mixed Cu−Zn phases, which suggests that alloying absorption near edge structures (XANES) spectra for the fresh
occurs under CO2RR reaction conditions, even though the bimetallic CuO−ZnO catalysts with different Cu:Zn ratios
initial electrodes consisted of spatially separated particles. together with those of the Cu, CuO, Zn, and ZnO references.
However, this mixing is not entirely homogeneous, as for both The Zn K-edge absorption peaks of the fresh catalysts match
the Cu0.14Zn0.86O and Cu0.75Zn0.25O catalysts, more Zn-rich and those of the ZnO reference. This is in agreement with the XRD
more Cu-rich regions are observed. We think that the extent of data, which shows the presence of ZnO phases. Additionally,
mixing is related to the oxidation state of both metals under the detailed analysis of the Zn K-edge by comparison to reference
CO2RR conditions, with intimate mixing requiring full spectra (Figure S14) confirms that Zn is in a + 2 oxidic state and
reduction of copper and zinc oxides. Conversely, the presence hexagonal in structure. The normalized Cu K-edge XANES
of segregated domains indicates incomplete reduction of the spectra for the CuO−ZnO catalysts do not resemble closely that
of CuO. This is possibly due to a more amorphous nature of the
initial copper and/or zinc oxides.
copper species, the presence of copper hydroxides, or an
The fact that the bimetallic electrodes showed higher CO
interaction with nearby ZnO.
partial current densities than the monometallic ones and that the
Both the Cu and Zn K-edge XANES spectra in Figures 8, S16,
addition of zinc to copper increased the cathodic potentials
and S18 show little differences between the three different
required for production of hydrocarbons strongly suggests that catalysts, even though they differ in terms of elemental
copper and zinc interact. This is remarkable because it is composition. Hence, the following discussions focus only on
generally assumed that intimate mixing (at the atomic-scale) of the Cu0.14Zn0.86O catalyst.
the starting materials is required for synergistic effects in In situ XAS measurements were carried out at different
bimetallic electrocatalysts. In contrast, the electrodes used in this potentials to obtain information on the changes in the chemical
work consisted of spatially separated CuO and ZnO particles state and structure of the Cu0.14Zn0.86O catalyst under CO2RR
initially. Importantly, the EDX maps show that intimate mixing conditions. Figure 8A shows the evolution of the Cu K-edge
is not a prerequisite for the starting materials because it takes spectra in the potential range of 0.6 to −0.6 V. It should be noted
place under reaction conditions. As such, these results underline here that these potentials are more anodic than the observed
the importance of catalyst characterization after testing because onset potential for the CO2RR (around −0.7 V) because of
significant restructuring can occur during CO2RR, which in turn experimental difficulties with, for example, bubble formation at
could strongly affect the electrocatalytic CO2RR performance. more cathodic potentials. Upon applying increasingly negative
To better understand the complex interplay between catalyst potentials, the Cu K-edge absorption shifts toward a lower
oxidation state, structure, and catalytic performance, in situ XAS energy, and the white line intensity decreases. Whereas the
measurements were performed. spectra initially resemble the CuO reference, the spectra at the
10708 https://doi.org/10.1021/acscatal.4c01575
ACS Catal. 2024, 14, 10701−10711
ACS Catalysis pubs.acs.org/acscatalysis Research Article

most negative potentials recorded here resemble that of the Cu and Cu−Zn because they are neighbors in the periodic table
reference most closely. No large changes in the XANES spectra and, hence, have similar scattering functions.
are observed for the potentials below 0.1 V, suggesting that Table S3 provides the quantitative analysis results of the Cu
copper reduction is complete. Interestingly, however, Cu K-edge EXAFS data. At 0.1 V and more negative potentials, the
reduction does not seem complete even at −0.6 V, since no second-shell Cu−O−M contribution disappears, while the
complete agreement with the Cu reference spectrum is obtained. coordination number corresponding to the Cu−O bond
To quantify the extent of copper oxide reduction, a linear decreases to below 1. This is in line with the previously
combination analysis is applied to compare the obtained Cu K- discussed XANES data, which showed a strong but incomplete
edge XANES spectra at different potentials to the Cu and CuO reduction of copper oxide, even at −0.6 V.
reference spectra. The results of this analysis are shown in Figure Changes in the FT-EXAFS spectra of the Zn K-edge are less
8C. Figure 8C shows that copper oxide reduction takes place pronounced than those of the Cu K-edge. A clear decrease in
mainly between 0.2 and 0 V. Interestingly, this linear intensity for both the Zn−O and Zn−O−M contributions is
combination analysis indicates that not all copper on our observed only at −0.6 V, as well as the appearance of a new peak
electrode is fully reduced, and some oxidized species persist even where the first-shell Zn−M bond is expected, based on the Zn
at −0.6 V. A possible explanation for this behavior is close foil reference. This indicates that ZnO reduction partially takes
contact and interaction with ZnO. Such an electronic interaction place at −0.6 V. This remarkable finding is in line with the
has also been suggested by others for electrodeposited CuZn previously discussed XANES data, again indicating that the
electrodes, showing that this interaction is relevant regardless of presence of nearby reduced Cu species enables the (partial)
the difference in initial oxidation state.29 reduction of ZnO.
Figure 8B shows the Zn K-edge EXAFS spectra obtained at Based on the findings of the in situ XANES and EXAFS
potentials ranging from 0.6 to −0.6 V vs RHE. In contrast to the analysis, combined with the ex-situ SEM and TEM analyses, we
Cu K-edge, the changes in these spectra are much less deduce that copper and zinc clearly show electronic interactions
pronounced, and all spectra strongly resemble that of the ZnO despite being spatially separated before catalysis. This indicates
the high mobility of both copper and zinc species under CO2RR
reference. This is logical given the thermodynamic reduction
conditions. Furthermore, the presence of copper enables zinc
potential of zinc oxide being more cathodic, namely −0.83 V at
reduction above its thermodynamic reduction potential, where-
pH 6.8. Interestingly, however, the −0.6 V spectrum features a
as the presence of zinc oxide prevents the full reduction of
decrease in the white line intensity, although no shift in
copper. As such, combining metals with different reduction
absorption energy is observed. This indicates partial reduction of potentials is a promising strategy for tuning the electrocatalyst
ZnO, which is likely enabled by its electronic interaction with oxidation state under the CO2RR conditions.
nearby reduced Cu species. Not only does this affect the The electronic interactions between copper and zinc play an
electronic structure of the ZnO, but also a structural effect is important role in their electrocatalytic performance. Specifically,
observed, with zinc partially changing from a tetrahedral to all bimetallic CuO−ZnO electrodes showed improved activity
octahedral configuration as shown in Figure S15. for CO as compared to the pure CuO and ZnO ones. Based on
Figure 9 shows the Fourier-transformed (FT) Cu K-edge and our findings, we suspect that the addition of copper increased
Zn K-edge EXAFS spectra of the Cu0.14Zn0.86O catalyst over a the amount of catalytically active Zn species that benefit the CO
range of applied potentials under CO2RR conditions. The fitting activity. Even though the in situ XAS measurements were
results corresponding to each spectrum are also shown, with the performed at a potential range where CO2RR activity is less
quantitative data summarized in Tables S3 and S4. The FT- pronounced, a detailed look at Figure 3 indicates a significant
EXAFS spectra of the fresh catalysts show contributions of both decrease in HER activity upon addition of ZnO to CuO at −0.6
Cu−O and Zn−O bonds at 1.5 Å for both the Cu K-edge and Zn V, more than would be expected based on their respective
K-edge (not corrected for phase shift). The phase corrected amounts. Additionally, we found that the bimetallic catalysts
Cu−O and Zn−O bond lengths are given in Tables S3 and S4, showed surprisingly little ethylene formation. Based on the in
being 1.99 and 1.93 Å, respectively. They correspond to the situ XAS and ex-situ microscopy results, we rationalize these
expected values for bulk CuO and ZnO of 1.94 and 1.97 Å.30 findings by the interaction of copper with nearby zinc oxide,
Furthermore, contributions of second-shell Cu−O−M and Zn− which prevents the full reduction of copper. Indeed, it has been
O−M bonds can be observed in the spectra of the fresh catalysts reported for copper-only catalysts using pulse experiments that
at 2.6 and 2.9 Å, respectively. The corresponding phase the presence of cationic copper species lowers the selectivity
corrected bonds lengths are given in Tables S3 and S4, being toward C2H4.31 However, the effect of the oxidation state has
3.02 and 3.23 Å, respectively. Although the second-shell Zn−M been a topic of strong debate, with some studies also arguing that
bond is in line with the expected value of 3.25 Å for bulk ZnO, the copper oxidation state does not play an important role in
the second-shell Cu−M bond is slightly higher than the expected electrodeposited CuZn-based systems.29 We think that by
2.84 Å.30 This value lies between the values of the CuO and ZnO starting from CuO and ZnO rather than electrodeposited CuZn
reference materials, indicating intimacy between the copper and in which the majority of species is already in a reduced form
zinc atoms. before catalytic testing, the effect of the copper and zinc
Upon application of increasing cathodic potentials, clear oxidation state is more pronounced in our catalytic data.
changes in the Cu K-edge FT-EXAFS spectra can be observed.
The contributions of both the first-shell Cu−O and second-shell
Cu−O−M bonds decrease in intensity and can no longer be
■ CONCLUSIONS
We studied the electrocatalytic reduction of CO2 on oxide-
observed at 0 V and more cathodic potentials. At the same time, derived CuO−ZnO electrodes with tunable compositions. The
a new peak appeared at 2.2 Å. This can be assigned to a first-shell use of hydrothermally synthesized CuO−ZnO electrodes
Cu−M bond, similar to that observed for the Cu foil. It should enabled us to study the effects of the initial catalyst structure,
be noted that it is impossible to discriminate between Cu−Cu composition, and oxidation state on electrocatalytic CO2RR
10709 https://doi.org/10.1021/acscatal.4c01575
ACS Catal. 2024, 14, 10701−10711
ACS Catalysis pubs.acs.org/acscatalysis Research Article

performance. Whereas the as-prepared electrodes consisted of Alessandro Longo − European Synchrotron Radiation Facility
spatially separated ZnO and CuO particles on carbon paper, (ESRF), Grenoble F-38000, France; Istituto per lo Studio dei
strong restructuring and atomic mixing took place under Materiali Nanostrutturati (ISMN)-CNR, UOS Palermo,
CO2RR conditions. This leads to an intimate contact between Palermo 90146, Italy; orcid.org/0000-0002-5381-7810
copper and zinc (oxide) and thereby a change in the catalyst Kristiaan H. Helfferich − Materials Chemistry and Catalysis,
activity and selectivity. Specifically, the higher the copper Debye Institute for Nanomaterials Science, Utrecht University,
content, the higher the overall activity. Interestingly, all of the 3584 CG Utrecht, The Netherlands
bimetallic electrodes showed higher CO production when Mahnaz Ghiasi − Materials Chemistry and Catalysis, Debye
corrected for the weight of metal than the monometallic Institute for Nanomaterials Science, Utrecht University, 3584
electrodes, with an optimum for the Cu0.43Zn0.57O electrode. CG Utrecht, The Netherlands
Furthermore, addition of zinc to copper decreased the C2H4 Complete contact information is available at:
selectivity, in contrast to what would be expected based on CO- https://pubs.acs.org/10.1021/acscatal.4c01575
spillover effects.
Especially remarkable is the interplay between CuO and ZnO Author Contributions
proximity and oxidation state during the CO2RR and the
The manuscript was written through contributions of all
consequent effect on the electrocatalytic performance. Using in
authors. All authors have given approval to the final version of
situ XAS measurements, we observe that CuO is readily reduced
the manuscript.
at potentials more negative than 0.1 V. Surprisingly, however,
the reduction is not complete even at −0.6 V because of the Notes
interaction with nearby ZnO. On the other hand, ZnO is The authors declare no competing financial interest.
partially reduced at −0.6 V which is below its thermodynamic
reduction potential. This shows that copper facilitates the
reduction of ZnO, leading to more catalytically active species for
■ ACKNOWLEDGMENTS
This publication is part of the Reversible Large-Scale Energy
CO, while stabilization of the cationic copper species by nearby Storage (RELEASE) consortium with Project Number 17621,
partially reduced ZnO suppresses the ability of Cu to make which is financed by the Dutch Research Council (NWO).
ethylene. These results clearly demonstrate that the initial Additionally, this work was supported by the European Research
structure of electrocatalysts is of minor importance to the Council; Project Number ERC-2014-CoG 648991. The authors
catalytic performance due to profound structural modifications also acknowledge the European Synchrotron Radiation Facility
under the reaction conditions. Our work also reveals that (ESRF) for providing beamtime at the BM8 and ID20 beamline.
electronic modification via interaction with metal oxide species Marisol Tapia Rosales, Francesco Mattarozzi, Maaike Vink-van
with different reducibility in oxide-derived bimetallic catalysts is Ittersum, Lisanne Blom, and Jan Willem de Rijk are acknowl-
a promising approach to design electrocatalysts with improved edged for useful discussions on the electrochemical setup and
activity and selectivity for electrochemical CO2 reduction. measurements.

■
*
ASSOCIATED CONTENT
sı Supporting Information
■ REFERENCES
(1) Spurgeon, J. M.; Kumar, B. A Comparative Technoeconomic
The Supporting Information is available free of charge at Analysis of Pathways for Commercial Electrochemical CO2 Reduction
https://pubs.acs.org/doi/10.1021/acscatal.4c01575. to Liquid Products. Energy Environ. Sci. 2018, 11 (6), 1536−1551.
(2) Nitopi, S.; Bertheussen, E.; Scott, S. B.; Liu, X.; Engstfeld, A. K.;
Electrochemical cell schematics, XRD patterns, electro- Horch, S.; Seger, B.; Stephens, I. E. L.; Chan, K.; Hahn, C.; Nørskov, J.
chemical data, SEM and TEM images and TEM-EDX K.; Jaramillo, T. F.; Chorkendorff, I. Progress and Perspectives of
maps after catalytic testing, in situ XANES and EXAFS Electrochemical CO2 Reduction on Copper in Aqueous Electrolyte.
data (PDF) Chem. Rev. 2019, 119 (12), 7610−7672.
(3) Rogelj, J.; Den Elzen, M.; Höhne, N.; Fransen, T.; Fekete, H.;

■ AUTHOR INFORMATION
Corresponding Authors
Winkler, H.; Schaeffer, R.; Sha, F.; Riahi, K.; Meinshausen, M. Paris
Agreement Climate Proposals Need a Boost to Keep Warming Well
below 2 °C. Nature 2016, 534 (7609), 631−639.
(4) Hori, Y. Electrochemical CO2 Reduction on Metal Electrodes.
Petra E. de Jongh − Materials Chemistry and Catalysis, Debye Mod. Asp. Electrochem. 2008, 42 (42), 89−189.
Institute for Nanomaterials Science, Utrecht University, 3584 (5) Birdja, Y. Y.; Pérez-Gallent, E.; Figueiredo, M. C.; Göttle, A. J.;
CG Utrecht, The Netherlands; orcid.org/0000-0002- Calle-Vallejo, F.; Koper, M. T. M. Advances and Challenges in
2216-2620; Email: [email protected] Understanding the Electrocatalytic Conversion of Carbon Dioxide to
Peter Ngene − Materials Chemistry and Catalysis, Debye Fuels. Nat. Energy 2019, 4 (9), 732−745.
Institute for Nanomaterials Science, Utrecht University, 3584 (6) Krause, R.; Reinisch, D.; Reller, C.; Eckert, H.; Hartmann, D.;
CG Utrecht, The Netherlands; orcid.org/0000-0003- Taroata, D.; Wiesner-Fleischer, K.; Bulan, A.; Lueken, A.; Schmid, G.
3691-0623; Email: [email protected] Industrial Application Aspects of the Electrochemical Reduction of
CO2 to CO in Aqueous Electrolyte. Chem. Ing. Tech. 2020, 92 (1−2),
Authors 53−61.
(7) Whipple, D. T.; Kenis, P. J. A. Prospects of CO2 Utilization via
Matt L. J. Peerlings − Materials Chemistry and Catalysis, Direct Heterogeneous Electrochemical Reduction. J. Phys. Chem. Lett.
Debye Institute for Nanomaterials Science, Utrecht University, 2010, 1 (24), 3451−3458.
3584 CG Utrecht, The Netherlands (8) Varela, A. S.; Ranjbar Sahraie, N.; Steinberg, J.; Ju, W.; Oh, H. S.;
Kai Han − Materials Chemistry and Catalysis, Debye Institute Strasser, P. Metal-Doped Nitrogenated Carbon as an Efficient Catalyst
for Nanomaterials Science, Utrecht University, 3584 CG for Direct CO2 Electroreduction to CO and Hydrocarbons. Angew.
Utrecht, The Netherlands Chem., Int. Ed. 2015, 54 (37), 10758−10762.

10710 https://doi.org/10.1021/acscatal.4c01575
ACS Catal. 2024, 14, 10701−10711
ACS Catalysis pubs.acs.org/acscatalysis Research Article

(9) Varela, A. S.; Ju, W.; Bagger, A.; Franco, P.; Rossmeisl, J.; Strasser, (27) Arán-Ais, R. M.; Rizo, R.; Grosse, P.; Algara-Siller, G.; Dembélé,
P. Electrochemical Reduction of CO2 on Metal-Nitrogen-Doped K.; Plodinec, M.; Lunkenbein, T.; Chee, S. W.; Roldan Cuenya, B.
Carbon Catalysts. ACS Catal. 2019, 9 (8), 7270−7284. Imaging Electrochemically Synthesized Cu2O Cubes and Their
(10) Luo, W.; Zhang, Q.; Zhang, J.; Moioli, E.; Zhao, K.; Züttel, A. Morphological Evolution under Conditions Relevant to CO2 Electro-
Electrochemical Reconstruction of ZnO for Selective Reduction of CO2 reduction. Nat. Commun. 2020, 11 (1), No. 3489.
to CO. Appl. Catal., B 2020, 273, No. 119060. (28) Vavra, J.; Ramona, G. P. L.; Dattila, F.; Kormányos, A.;
(11) Kortlever, R.; Shen, J.; Schouten, K. J. P.; Calle-Vallejo, F.; Koper, Priamushko, T.; Albertini, P. P.; Loiudice, A.; Cherevko, S.; Lopéz, N.;
M. T. M. Catalysts and Reaction Pathways for the Electrochemical Buonsanti, R. Solution-Based Cu+ Transient Species Mediate the
Reduction of Carbon Dioxide. J. Phys. Chem. Lett. 2015, 6 (20), 4073− Reconstruction of Copper Electrocatalysts for CO2 Reduction. Nat.
4082. Catal. 2024, 7 (1), 89−97.
(12) Clark, E. L.; Resasco, J.; Landers, A.; Lin, J.; Chung, L. T.; (29) Velasco-Vélez, J.; Poon, J.; Gao, D.; Chuang, C. H.; Bergmann,
Walton, A.; Hahn, C.; Jaramillo, T. F.; Bell, A. T. Standards and A.; Jones, T. E.; Haw, S. C.; Chen, J. M.; Carbonio, E.; Mom, R. V.;
Protocols for Data Acquisition and Reporting for Studies of the Ivanov, D.; Arrigo, R.; Cuenya, B. R.; Knop-Gericke, A.; Schlögl, R.
Electrochemical Reduction of Carbon Dioxide. ACS Catal. 2018, 8 (7), Cationic Copper Species Stabilized by Zinc during the Electrocatalytic
6560−6570. Reduction of CO2 Revealed by In Situ X-Ray Spectroscopy. Adv.
(13) Ross, M. B.; De Luna, P.; Li, Y.; Dinh, C. T.; Kim, D.; Yang, P.; Sustainable Syst. 2023, 7 (5), 2200453.
Sargent, E. H. Designing Materials for Electrochemical Carbon Dioxide (30) Tohji, K.; Udagawa, Y.; Mizushima, T.; Ueno, A. The Structure
Recycling. Nat. Catal. 2019, 2 (8), 648−658. of the Cu/ZnO Catalyst by an in-Situ EXAFS Study. J. Phys. Chem. A
(14) Won, D. H.; Shin, H.; Koh, J.; Chung, J.; Lee, H. S.; Kim, H.; 1985, 89 (26), 5671−5676.
Woo, S. I. Highly Efficient, Selective, and Stable CO2 Electroreduction (31) Timoshenko, J.; Bergmann, A.; Rettenmaier, C.; Herzog, A.;
Arán-Ais, R. M.; Jeon, H. S.; Haase, F. T.; Hejral, U.; Grosse, P.; Kühl,
on a Hexagonal Zn Catalyst. Angew. Chem., Int. Ed. 2016, 55 (32),
S.; Davis, E. M.; Tian, J.; Magnussen, O.; Roldan Cuenya, B. Steering
9297−9300.
the Structure and Selectivity of CO2 Electroreduction Catalysts by
(15) Geng, Z.; Kong, X.; Chen, W.; Su, H.; Liu, Y.; Cai, F.; Wang, G.;
Potential Pulses. Nat. Catal. 2022, 5 (4), 259−267.
Zeng, J. Oxygen Vacancies in ZnO Nanosheets Enhance CO2
Electrochemical Reduction to CO. Angew. Chem. 2018, 130 (21),
6162−6167.
(16) Han, K.; Ngene, P.; de Jongh, P. Structure Dependent Product
Selectivity for CO2 Electroreduction on ZnO Derived Catalysts.
ChemCatChem 2021, 13 (8), 1998−2004.
(17) He, J.; Johnson, N. J. J.; Huang, A.; Berlinguette, C. P.
Electrocatalytic Alloys for CO2 Reduction. ChemSusChem 2018, 11 (1),
48−57.
(18) Feng, Y.; Li, Z.; Liu, H.; Dong, C.; Wang, J.; Kulinich, S. A.; Du,
X. Laser-Prepared CuZn Alloy Catalyst for Selective Electrochemical
Reduction of CO2 to Ethylene. Langmuir 2018, 34 (45), 13544−13549.
(19) Ren, D.; Ang, B. S. H.; Yeo, B. S. Tuning the Selectivity of Carbon
Dioxide Electroreduction toward Ethanol on Oxide-Derived CuxZn
Catalysts. ACS Catal. 2016, 6 (12), 8239−8247.
(20) da Silva, A. H. M.; Raaijman, S. J.; Santana, C. S.; Assaf, J. M.;
Gomes, J. F.; Koper, M. T. M. Electrocatalytic CO2 Reduction to C2+
Products on Cu and CuxZny Electrodes: Effects of Chemical
Composition and Surface Morphology. J. Electroanal. Chem. 2021,
880, No. 114750.
(21) Jeon, H. S.; Timoshenko, J.; Scholten, F.; Sinev, I.; Herzog, A.;
Haase, F. T.; Cuenya, B. R. Operando Insight into the Correlation
between the Structure and Composition of CuZn Nanoparticles and
Their Selectivity for the Electrochemical CO2 Reduction. J. Am. Chem.
Soc. 2019, 141 (50), 19879−19887.
(22) Timoshenko, J.; Jeon, H. S.; Sinev, I.; Haase, F. T.; Herzog, A.;
Roldan Cuenya, B. Linking the Evolution of Catalytic Properties and
Structural Changes in Copper-Zinc Nanocatalysts Using Operando
EXAFS and Neural-Networks. Chem. Sci. 2020, 11 (14), 3727−3736.
(23) Wan, L.; Zhang, X.; Cheng, J.; Chen, R.; Wu, L.; Shi, J.; Luo, J.
Bimetallic Cu-Zn Catalysts for Electrochemical CO2 Reduction: Phase-
Separated versus Core-Shell Distribution. ACS Catal. 2022, 12 (5),
2741−2748.
(24) Law, M.; Greene, L. E.; Johnson, J. C.; Saykally, R.; Yang, P.
Nanowire Dye-Sensitized Solar Cells. Nat. Mater. 2005, 4 (6), 455−
459.
(25) Firet, N. J.; Burdyny, T.; Nesbitt, N. T.; Chandrashekar, S.;
Longo, A.; Smith, W. A. Copper and Silver Gas Diffusion Electrodes
Performing CO2 Reduction Studied through Operando X-Ray
Absorption Spectroscopy. Catal. Sci. Technol. 2020, 10 (17), 5870−
5885.
(26) Huang, J.; Hörmann, N.; Oveisi, E.; Loiudice, A.; De Gregorio,
G. L.; Andreussi, O.; Marzari, N.; Buonsanti, R. Potential-Induced
Nanoclustering of Metallic Catalysts during Electrochemical CO2
Reduction. Nat. Commun. 2018, 9 (1), No. 3117.

10711 https://doi.org/10.1021/acscatal.4c01575
ACS Catal. 2024, 14, 10701−10711