Controlling the Surface Oxidation of Cu Nanowires Improves Their Catalytic Selectivity and Stability toward C2+ Products in CO2 Reduction

Article abstract of DOI:10.1002/anie.202011956

Copper nanostructures are promising catalysts for the electrochemical reduction of CO₂ because of their unique ability to produce a large proportion of multi-carbon products. Despite great progress, the selectivity and stability of such catalysts still need to be substantially improved. Here, we demonstrate that controlling the surface oxidation of Cu nanowires (CuNWs) can greatly improve their C₂₊ selectivity and stability. Specifically, we achieve a faradaic efficiency as high as 57.7 and 52.0 % for ethylene when the CuNWs are oxidized by the O₂ from air and aqueous H₂O₂, respectively, and both of them show hydrogen selectivity below 12 %. The high yields of C₂₊ products can be mainly attributed to the increase in surface roughness and the generation of defects and cavities during the electrochemical reduction of the oxide layer. Our results also indicate that the formation of a relatively thick, smooth oxide sheath can improve the catalytic stability by mitigating the fragmentation issue.

Full text of DOI:10.1002/anie.202011956

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Controlling the Surface Oxidation of Cu Nanowires Improves
Their Catalytic Selectivity and Stability toward C2+ Products in
CO2 Reduction
Authors: Zhiheng Lyu, Shangqian Zhu, Minghao Xie, Yu Zhang, Zitao
Chen, Ruhui Chen, Mengkun Tian, Miaofang Chi, Minhua
Shao, and Younan Xia
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202011956
Link to VoR: https://doi.org/10.1002/anie.202011956

10.1002/anie.202011956
Angewandte Chemie International Edition
Controlling the Surface Oxidation of Cu Nanowires Improves Their
Catalytic Selectivity and Stability toward C2+ Products in CO2 Reduction
[b],
[a]
Zhiheng Lyu,[a], Shangqian Zhu,
Minghao Xie, Yu Zhang,[b,c] Zitao Chen,[c] Ruhui Chen,[a]
+
+
[d]
[b]
[a,c]
[e]
Mengkun Tian, Miaofang Chi, Minhua Shao,* Younan Xia*
[a] Z. Lyu, M. Xie, R. Chen, Prof. Dr. Y. Xia
School of Chemistry and Biochemistry, Georgia Institute of Technology
Atlanta, Georgia 30332 (USA)
Email: younan.xia@bme.gatech.edu (for material synthesis and characterization)
[b] S. Zhu, Dr. Y. Zhang, Prof. Dr. M. Shao
Department of Chemical and Biological Engineering, The Hong Kong University of Science and
Technology
Clear Water Bay, Kowloon, Hong Kong (P. R. China)
Email: kemshao@ust.hk (for electrocatalytic measurements)
[c] Dr. Y. Zhang, Z. Chen, Dr. Y. Xia
The Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and
Emory University
Atlanta, Georgia 30332 (USA)
[d] Dr. M. Tian
Institute for Electronics and Nanotechnology, Georgia Institute of Technology
Atlanta, Georgia 30332 (USA)
[e] Dr. M. Chi
Center for Nanophase Materials Sciences, Oak Ridge National Laboratory
Oak Ridge, Tennessee 37831 (USA)
[+] These authors contributed equally to this work.
Supporting information for this article is given via a link at the end of the document.
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
Abstract
Copper nanostructures are promising catalysts for the electrochemical reduction of CO2
because of their unique ability to produce a large proportion of multi-carbon products. Despite
major progress, the selectivity and stability of such catalysts still need to be substantially
improved. Here we demonstrate that controlling the surface oxidation of Cu nanowires (CuNWs)
can greatly improve their C2+ selectivity and stability. Specifically, we achieve a faradaic
efficiency as high as 57.7 and 52.0% for ethylene when the CuNWs are oxidized by the O2 from
air and aqueous H2O2, respectively, and both of them show hydrogen selectivity below 12%. The
high yields of C2+ products can be mainly attributed to the increase in surface roughness and the
generation of defects and cavities during the electrochemical reduction of oxide layer. Our
results also indicate that the formation of a relatively thick, smooth oxide sheath can improve the
catalytic stability by mitigating the fragmentation issue.
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
Introduction
The large amount of CO2 produced from transportation and industry has raised increasing
concerns in recent years because of the major environmental burdens from this gas and the
necessary increase in mitigation investment.[1,2] In addition to photosynthesis, another promising
route is to electrochemically convert CO2 to chemical compounds for use as fuels or feedstock.
Typically, a variety of products are generated in this process, with the most common ones being
CO and formate. Compared to C1 species, multi-carbon (C2+) products such as ethylene and
ethanol are more desirable considering their higher values, larger energy densities, and broader
applications. Despite the remarkable progress, it remains a grand challenge to electrocatalytically
promote C-C coupling and thereby achieve high selectivity toward C2+ products during the CO2
reduction reaction (CO2RR).[1-3]
Among all the metals under consideration for CO2RR, Cu easily stands out for its ability to
produce a substantial amount of oxygenates and hydrocarbons. However, for those catalysts
based on pure Cu (e.g., Cu foils), they tend to suffer from a number of shortcomings, including
low selectivity, poor competition with the hydrogen evolution reaction (HER), high overpotential,
and compromised stability. To address these issues, the Cu nanostructures have been engineered
in multiple ways, such as control of their shapes and thus surface structures,[4-7] doping with
other metals or non-metals,[8-10] and complete or partial oxidation of Cu.[11-13] When switching to
Cu dendrites with a rough, defective surface, for example, a faradaic efficiency (FE) as high as
57% was achieved for ethylene at a potential more negative than -1.4 VRHE (RHE: reversible
hydrogen electrode) in 0.1 M aqueous KBr. The excellent performance was attributed to the
presence of high-index facets.[5] In another study, star-shaped Cu decahedra featuring a penta-
twinned structure were demonstrated with a good selectivity of 52.4% toward ethylene at -1.0
VRHE in 0.1 M aqueous KHCO3. The tensile surface strain caused by the twin boundaries and the
induced surface defects were proposed as a main contribution to the high C2+ yields.[6]
By partially oxidizing Cu nanostructures or directly using nanostructures made of CuxO,
improvement in catalytic performance was also achieved, including the observations of lower
overpotentials and higher selectivity toward C2+ products. The improvement was typically
ascribed to an increase in surface roughness and thus a higher density of surface defects, induced
changes to the local pH, and the possible existence of residual Cu(I) species that might serve as
active sites.[1] In one study, it was reported that when Cu2O nanoparticles were used as catalysts,
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
a high selectivity approaching 60% was achieved for ethylene after 6 h of electrolysis at -1.1
VRHE in 0.1 M aqueous KHCO3. The 20-nm Cu2O particles were gradually broken down into 2-4
nm particles under the negative potential, leading to an increase in the selectivity toward
ethylene that finally reached 60%. The compact arrangement of small particles and the high
density of grain boundaries among them were believed to be responsible for the high yield of C2+
products because the defective structures helped promote C-C coupling.[11] In another report,
plasma-activated Cu nanocubes with an oxygen content of 30% (in terms of atomic percentage)
showed a higher FE toward ethylene (up to 45% at -1.0 VRHE in 0.1 M aqueous KHCO3) relative
to those with lower oxygen contents. The key parameters were believed to be the defects induced
by plasma oxidation and the surface/subsurface oxygen species, both capable of enhancing the
binding of CO to the Cu surface.[13]
In an attempt to leverage the benefits arising from both shape control and surface oxidation,
here we synthesize Cu nanowires (CuNWs) and then partially oxidize their surface under two
different conditions to generate Cu@CuxO core-sheath nanostructures. When the oxide sheath is
reduced to elemental Cu, the increase in surface roughness and the penta-twinned structure
intrinsic to the nanowires are able to work synergistically in promoting the generation of C2+
products during CO2RR. As a result, FEs as high as 57.7 and 52.0% for ethylene are obtained at -
1.0 VRHE in 0.1 M aqueous KHCO3 when the CuNWs oxidized by the O2 from air (A-CuNWs)
and aqueous H2O2 (H-CuNWs) serve as catalysts, respectively. The high selectivity toward
ethylene, together with C2+ yields approaching 78.4 and 71.9%, is among the highest when
benchmarked against all the Cu-based nanostructures reported in literature (Table S1). In
addition to the high selectivity, we also demonstrate that a thick and smooth oxide sheath on the
CuNWs can substantially improve their catalytic stability, with H-CuNWs showing a much
better stability than A-CuNWs, due to the mitigation of fragmentation and thus preservation of
one-dimensional (1-D) morphology. This work represents the first attempt to bring a tight control
to the oxidation process of CuNWs for the demonstration of its importance in determining the
activity, selectivity, and stability of Cu catalysts toward CO2RR.
Results and Discussion
The preparation of a catalyst typically involved two major steps: synthesis of CuNWs in the
presence of hexadecylamine and glucose, followed by surface oxidation at room temperature
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
under different conditions. Figure S1 shows transmission electron microscopy (TEM) images of
the as-prepared CuNWs with an average diameter of 22.4 ± 4.2 nm and length up to several
micrometers. Based on our previous report, the nanowires had a penta-twinned structure and
their side surface was covered by {100} facets.[14] After oxidization with the O2 from air, the 1-D
morphology was retained while the surface was roughened due to the formation of an oxide
sheath (Figure 1a, b). As shown in Figure 1b, lattice fringes with a spacing of 0.21 nm,
corresponding to {111} planes of Cu, could be observed at the central region of the nanowire,
indicating the presence of metallic Cu in the core. The fringes could hardly be observed when
moving toward the surface of the nanowire, along with a decrease in contrast, suggesting the
formation of oxides on the surface. At a few spots where lattice fringes could be resolved, the
spacing was measured to be 0.25 nm, corresponding to the {111} planes of Cu2O. Taken
together, it can be concluded that the oxide sheath was comprised of a mixture of both
amorphous and crystalline copper oxides. The presence of surface oxides was also supported by
the electron energy loss spectroscopy (EELS) mapping data shown in Figure 1c, where a thin
layer containing oxygen could be resolved on the core made of metallic Cu. Interestingly, in the
dark-field scanning transmission electron microscopy (DF-STEM) image, we observed the
formation of an oxide sheath with varying thickness at different locations. Based on previous
reports, it was demonstrated that the formation of copper oxides on Cu(100) facets followed an
island-growth mode and it was found to be more obvious under a slow oxidation kinetics.[15,16]
Taken together, the non-uniform oxidation occurring on the surface of a CuNW was believed to
be responsible for the formation of a CuxO sheath with uneven thickness.
In the X-ray photoelectron spectroscopy (XPS) spectrum recorded from A-CuNWs (Figure
1e), two peaks were resolved at 932.1 and 951.9 eV, which could be assigned to Cu(0) or Cu(I)
2p peaks. Additionally, we observed two peaks at 933.9 and 953.7 eV, together with satellite
peaks around 941, 943, and 962 eV, and all of them could be assigned to Cu(II) 2p peaks.[17]
These results indicate that the oxide sheath was made of a mix of Cu2O and CuO. In the X-ray
diffraction (XRD) pattern, however, no peaks of copper oxides were observed (Figure 1d). We
only resolved three peaks at 43.4°, 50.5°, and 74.2°, corresponding to the diffraction from
Cu(111), (200), and (220) planes. The absence of CuxO peaks in the XRD pattern confirmed the
dominance of amorphous phase, in agreement with our conclusion drawn from STEM imaging.
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
Different from the slow oxidation by the O2 from air, CuNWs covered by a relatively
thicker CuxO sheath (H-CuNWs) were obtained when the sample was subjected to oxidation by
aqueous H2O2 (Figure 2a). The H-CuNWs had a metallic Cu core with an average diameter of
18.5 ± 3.4 nm and an oxide sheath of 6.9 ± 0.7 nm in thickness. As shown by the DF-STEM
image in Figure 2b, both the oxide layer and the Cu core had smooth surfaces and this outcome
can be attributed to the fast oxidation kinetics. Multiple nucleation sites for copper oxides were
generated at the same time across the surface of a nanowire, leading to the formation of an oxide
sheath uniform in thickness.[15] The existence of CuxO layer was also verified by EELS mapping
(Figure 2c). A higher intensity of oxygen was observed relative to that of A-CuNWs, consistent
with the fact that H-CuNWs had a thicker layer of oxides. In the DF-STEM image, the clearly
resolved lattice fringes in the core confirmed its composition as metallic Cu, while nearly no
fringes were observed in the sheath. From the XRD pattern (Figure 2d), the intensity of Cu(0)
was also the strongest, with three peaks positioned at 43.4°, 50.5°, and 74.2°. In contrast, the
peaks of both Cu2O and CuO were very weak, with that at 28.9° corresponding to Cu2O and
those at 39.0°, 47.0°, and 48.0° to CuO. A plausible reason for the weak intensity of copper
oxides in the XRD pattern can be ascribed to the dominance of amorphous structure, which was
further proven by nano-beam electron diffraction (NBED). As shown in Figure S2, rings along
with irregularly distributed diffraction spots were observed in the NBED patterns recorded from
selected areas of the CuxO sheath, revealing the amorphous phase containing randomly oriented
nanocrystallites. In contrast, discrete spots located at nearly the same positions could be clearly
observed in the NBED patterns taken from the Cu core, with no accompanied rings,
demonstrating the continuity and crystalline structure of the metallic core.[18] The composition of
the surface oxides was also analyzed using XPS (Figure 2e). Different from A-CuNWs, the
intensity of Cu(II) accounted for the vast majority and the peaks of Cu(0) or Cu(I) were barely
detected, suggesting that the surface of H-CuNWs was covered by a relatively thick sheath made
of CuO. In a typical process of Cu oxidation, the metallic Cu was first oxidized to Cu2O, which
could be further converted to CuO in the presence of adequate oxidants. The relatively high
concentration of H2O2 and its fast oxidation kinetics enabled the fully oxidation of Cu, leading to
the formation of a CuO-based sheath.[19,20]
In addition to microscopy evidence, the presence of oxide layer also agreed with the
ultraviolet-visible (UV-vis) spectra (Figure S3). Due to the localized surface plasma resonance
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
(LSPR) property, Cu nanostructures exhibit absorption in the visible region, and the peak
position is highly sensitive to the oxidation state of Cu on the surface because of the change in
dielectric constant.[21] For the freshly prepared CuNWs, A-CuNWs, and H-CuNWs, we obtained
peaks located at 564, 568, and 568 nm, respectively. Compared to the as-obtained CuNWs, the
peaks of both A-CuNWs and H-CuNWs were slightly red-shifted and the shift can be attributed
to the surface oxidation of Cu nanostructures.[22,23]
We then compared the selectivity and activity of the CuNWs with different surface
oxidation toward CO2RR in the potential range of -0.8 to -1.2 VRHE in 0.1 M aqueous KHCO3
(Figure 3, with details provided in the Supporting Information). Multiple products were detected,
including CO, CH4, ethylene, and ethanol, and we paid special attention to the C2+ species, in
particular, ethylene. For A-CuNWs, a FE approaching 57.7% was obtained at -1.03 VRHE for
ethylene, together with a total efficiency reaching 78.4% for C2+ products. These values were
among the highest selectivity for CO2RR in a KHCO3 solution, even compelling when compared
to the performance measured in a flow cell containing a strong alkaline electrolyte (Figure 3a
and Table S1). Compared with the high yield of C2+ products at -1.03 VRHE, the production of
CO and CH4 only accounted for 0.65 and 1.9%, respectively. The FE of H2 was as low as 8.5%,
indicating the effective suppression of HER. Normalized by geometric area, a total current
density of 34.0 mA·cm-2 was achieved along with the high selectivity toward ethylene, whose
partial current density reached 19.6 mA·cm-2 (Figure 3c, d). When normalized to the mass of
catalyst, we obtained a mass activity of 384.1 mA·mg-1 for the selective production of ethylene
(Figure S4). The high current density and mass activity of A-CuNWs demonstrate their
prominent performance in electrochemical CO2RR.[11]
In comparison with A-CuNWs, H-CuNWs with a thicker oxide sheath and smoother surface
showed a slightly lower FE toward ethylene: 52.0% at -1.02 VRHE (Figure 3b). The total yield for
C2+ products reached 71.9% while the proportion of H2 was at 11.4%, together with low yields
for both CO and CH4 (2.0 and 6.1%, respectively). At -1.02 VRHE, a geometric current density of
24.1 mA·cm-2 was achieved for all the products, with a partial current density of 12.5 mA·cm-2
for ethylene (Figure 3c, d). Although slightly lower than A-CuNWs, a mass activity approaching
246.2 mA·mg-1 was obtained for ethylene production when H-CuNWs served as the catalyst
(Figure S4). When normalized to the electrochemically active surface area (ECSA), we obtained
partial current densities as high as 4.1 and 3.5 mA·cm-2 toward ethylene for A-CuNWs and H-
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
CuNWs, respectively, demonstrating their outstanding performance in CO2RR (Figure S5 and S6,
the current densities of comparable catalysts can be found in Table S1). It should be noted that
the high (approaching 30 mA·cm-2) geometric current densities achieved by CuNWs could be
attributed to the involvement of 3-D mass transport, which raised the limitation of current
density.[24]
To look into the mechanism underlying the high selectivity and activity of CuNW-based
catalysts and the reason why they behaved differently, we characterized the structures of both
catalysts after 1 h of electrolysis at -1.02 VRHE. As shown in Figure 4a, fragmentation was
observed for A-CuNWs, where small pseudo-spherical nanoparticles and nanorods resulted from
disintegration of nanowires could be found in the TEM image. For the nanowires whose 1-D
structure was preserved, their surface became fairly rough. A close examination of the nanowire
illustrated the existence of large holes and cavities on the surface, which could be attributed to
the reduction of surface oxides (Figure 4b). As shown by the bright-field STEM (BF-STEM)
image in Figure 4c, the lattice spacing of 0.21 and 0.18 nm could be assigned to the {111} and
{200} planes of Cu, respectively, confirming that the majority of surface CuxO was reduced to
metallic Cu. The conversion of oxide sheath was also revealed by XPS. Compared to the
relatively strong peaks of copper oxides in the as-prepared A-CuNWs (Figure 1e), no satellite
peaks were observed for the same sample after 1 h of electrolysis (Figure 4d), confirming the
reduction of Cu(II) to Cu(0) or Cu(I). Considering the larger lattice spacing of copper oxides and
possible surface reconstruction, voids would be generated for the formation of cavities on the
surface when CuxO was reduced to metallic Cu.[25] The non-uniform thickness of oxide layer
caused by the slow oxidation kinetics of O2 from air resulted in the formation of voids with
different sizes and thus the presence of deep cavities, increasing the chance of breaking up the
nanowires. In addition to the reduction of copper oxides, detachment of small nanoparticles from
the CuNWs during electrolysis might also facilitate their fragmentation.[26] It was reported that
the adsorption of either H or CO, intermediates in HER and CO2RR, on the surface of Cu
nanocubes would lead to degradation to the shapes when a low potential was applied. An
increased number of pinholes on the surface and a decrease in the size of Cu nanocubes were
observed, along with an increased number of nanoparticles emerging around the cubes. [27] This
was similar to what we observed on A-CuNWs.
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
In contrast to the A-CuNWs, the 1-D morphology of the H-CuNWs was largely retained
during electrolysis. As shown in Figure 4e, we found nearly no fragment around the nanowires,
let alone their disintegration. Compared with the smooth surface of H-CuNWs before electrolysis,
the surface roughness was increased after being held at -1.02 VRHE for 1 h (Figure 2a). As shown
by the BF-STEM image in Figure 4f, small, shallow voids were formed on the surface. The
magnified, atomic-resolution image revealed the presence of metallic Cu on the surface, where
the lattice spacing of 0.21 and 0.18 nm could be indexed to the Cu {111} and {100} planes
(Figure 4g). Some copper oxides were also observed outside the metallic Cu core, but they are
highly possible to be generated during the preparation of sample for STEM imaging. Similar to
the A-CuNWs, the XPS peaks of CuO disappeared after electrolysis, confirming the reduction of
Cu(II) to Cu(0) or Cu(I) (Figure 4h). To further verify the oxidation state of Cu on the surface,
we measured the Auger Cu LMM spectra of the nanowires before and after electrolysis (Figure
S7). The samples were protected by Ar throughout the transfer process. Before electrolysis, the
surface oxides of both types of nanowires were mainly composed of CuO, as revealed by the
peaks located at 568.9 eV. After 1 h of electrolysis, we observed both Cu(0) and Cu(I) peaks
located at 568.0 and 569.8 eV, respectively, suggesting the presence of Cu(I) ions in the catalysts.
It is worth pointing out that the retention of Cu(I) species during electrolysis was also observed
in recent studies involving both ex situ and in situ techniques. [10,12,13,27-29]
Based on TEM and STEM imaging, the high yields of C2+ products for CuNWs covered
with an oxide sheath can be mainly attributed to their increased surface roughness.[5,25,30,31] As
previously reported, roughening the surface of Cu catalysts would lead to the generation of a
greater density of defective sites, which can enhance the adsorption of C1 intermediates and thus
improve the catalyst’s selectivity toward C2+ products.[30] In addition to surface roughness,
several other factors, including the presence of twin defects and 1-D morphology of the
nanowires, may also be involved.[6,32] Based on simulations from previous reports, the twin
boundaries and tensile surface strain, together with stacking faults arising from a stress release
mechanism, could enhance the adsorption of CO on Cu surface and increase CO coverage
density, which in turn facilitated the C-C coupling and formation of C2+ products.[6] As such, the
high yields of C2+ species may also benefit from the penta-twinned structure of the CuNWs.
Moreover, the small diameter of the nanowires and their 1-D morphology also enlarged the
specific surface area and improved the mass transport in electrocatalysis, contributing to a higher
8
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
mass activity.[33,34] As a comparison, when H2O2-oxidized Cu nanoparticles were used as the
catalyst, a FE of 41.8% toward ethylene was obtained at -1.05 VRHE, together with a geometric
partial current density of 6.8 mA cm-2 (Figure S8). The inferior selectivity and lower current
density of the nanoparticles further demonstrate the importance of both the 1-D morphology and
twinned structure of the nanowires in improving mass transport and facilitating C-C coupling.[34]
Comparing the shape evolution of two types of CuNWs during CO2RR, we proposed that
the higher C2+ selectivity of A-CuNWs relative to that of H-CuNWs could be ascribed to their
even greater surface roughness. As confirmed by the TEM and STEM images in Figure 4b and
4f, deeper cavities were observed on A-CuNWs, and they were reported to help trap the C1 and
even C2 intermediates and thus facilitate C-C coupling.[35,36] Besides, the greater extent of
roughness also revealed the potential existence of a higher abundance of high-index facets and
defective sites on the surface of A-CuNWs, which had been proven to promote the formation of
C2+ products.[7,25] In addition to selectivity, the greater surface roughness also contributed to a
larger specific surface area and thus higher mass activity of A-CuNWs compared to H-CuNWs.
As shown in Figure S5 and Table S2, double-layer capacitances of different types of nanowires
were measured, correlating their activities with surface areas. A capacitance of 140 μF·cm-2 was
obtained for A-CuNWs, which was 40% greater than that of H-CuNWs, consistent with the
higher mass activity of A-CuNWs. When using pristine CuNWs as a reference, both A-CuNWs
and H-CuNWs exhibited larger capacitances (with a factor of 3.0 and 2.2, respectively),
revealing their increase in surface roughness arising from the presence and subsequent reduction
of surface oxides. We also tried to measure the CO2RR performance of pristine CuNWs as soon
as they were prepared, with an effort to avoid significant oxidation like in the case of A-CuNWs.
Compared with the surface-oxidized nanowires, pristine CuNWs showed a relatively lower C2+
selectivity (66.0%) at -1.07 VRHE owing to the absence of extensive surface oxidation and thus
the lower surface roughness (Figure S9).
The stability of the CuNW-based catalysts was then measured at -1.02 VRHE, corresponding
to the highest ethylene selectivity. For A-CuNWs, the FE of ethylene gradually decreased as the
reaction time was extended, along with an increase in the yields of H2 and CH4 (Figure 5a),
which could be mainly attributed to the severe fragmentation of A-CuNWs during the
measurement. It has been suggested by many reports that the selectivity of Cu catalysts is highly
related to the particle size.[37,38] When the particle size of Cu decreased, below 5 nm in particular,
9
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
a dramatic increase in activity and selectivity toward HER was observed, which can be ascribed
to the increased proportion of atoms located at edges and corners. Those atoms have relatively
low coordination numbers and can serve as strong binding sites for H and CO, limiting their
diffusion on surface and further reaction to generate C2+ products. Under this circumstance, HER
was greatly facilitated due to the improved adsorption of protons on Cu surface while the
production of C2+ species was suppressed.[38] Along with the decrease in selectivity toward
ethylene, an increase in the total current density by about 40% was also observed after 22 h
(Figure 5b), confirming the fragmentation of A-CuNWs and further enlargement in specific
surface area.
When H-CuNWs served as the catalyst, the selectivity toward ethylene showed nearly no
decay after 22 h, indicating its superb stability (Figure 5c). In the meantime, the current density
of H-CuNWs also stayed relatively constant during the measurement, with an increase less than
5 mA·cm-2 (or 20%) (Figure 5d). To understand the enhanced stability, TEM imaging was
conducted on both types of CuNWs after 5 h of electrolysis and the results are shown in Figure
S10. In the case of A-CuNWs, broken nanowires and Cu nanoparticles, either separated from or
attached to the nanowires, were observed (Figure S10a–c). As for H-CuNWs, very few of them
were fragmented (Figure S10d–f). Regarding the difference in morphology, we proposed that the
different oxidation pathways and the resultant structures of oxide layers were responsible for the
difference in stability. The non-uniform oxidation and the resultant deep cavities led to easier
disintegration of nanowires, as mentioned previously, resulting in the poor stability of A-CuNWs.
In contrast, for H-CuNWs covered by a thicker and smoother CuxO layer, the metallic Cu core
was well protected during the CO2RR. Though the surface oxides were reduced during catalysis,
no large voids were developed on surface and the continuity of the nanowires was well preserved.
The preservation of 1-D morphology and suppression of fragmentation by introducing a thick,
smooth oxide sheath contributed to an improved stability for H-CuNWs.[39] However, it should
be pointed out that although higher durability was achieved for H-CuNWs, they suffered slightly
in terms of selectivity and activity when compared to A-CuNWs. Besides, the increase in current
density as a function of time also revealed that the fragmentation could be effectively retarded
for H-CuNWs, but not completely eliminated. At the moment, it remains a major challenge to
design catalysts with both good stability and high selectivity and activity. Addition of supports to
10
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
stabilize catalysts or covering the surface of a catalyst with proper ligands to prevent surface
reconstruction might be promising directions in further research.[40]
Conclusions
In summary, we have demonstrated two effective CO2RR catalysts based on CuNWs with
their surface being partially oxidized through different pathways. The A-CuNWs obtained via
oxidation by the O2 from air possessed a rough surface and a thin oxide layer with non-uniform
thickness, while the H-CuNWs obtained by oxidation with aqueous H2O2 were covered with a
relatively thicker and smoother oxide sheath. When applied to electrochemical CO2RR, both of
them exhibited high selectivity and activity toward C2+ species, ethylene in particular. The
greater extent of surface roughness, the presence of more defective sites and deeper cavities after
reduction of non-uniform oxide sheath contributed to a higher C2+ selectivity for A-CuNWs
relative to H-CuNWs. However, these features were also detrimental to the stability of A-
CuNWs due to their increased susceptibility to disintegration during electrolysis. Protected by a
thicker and smoother CuxO layer, the fragmentation of nanowires was effectively suppressed for
H-CuNWs and nearly no decay in ethylene selectivity was observed after 22 h, demonstrating
the improved catalytic stability. Taken together, this work demonstrates that controlling surface
oxidation offers an effective strategy for improving both the C2+ selectivity and stability of Cu-
based catalysts. We believe the same strategy should be extendible to other types of Cu
nanostructures for the rational development of effective catalysts toward CO2RR.
Acknowledgements
This work was supported in part by a grant from the NSF (CHE 1804970), start-up funds from
the Georgia Institute of Technology, and two grants from the Research Grant Council of Hong
Kong SAR (26206115 and 16309418). TEM and STEM imaging were performed at the
Materials Characterization Facilities of the GT’s Institute of Electronics and Nanotechnology
(IEN), a member of the National Nanotechnology Coordinated Infrastructure (NNCI) and
supported by the National Science Foundation (ECCS-1542174). Part of the STEM imaging and
EELS mapping in this research were completed at the Center for Nanophase Materials Sciences,
which is a DOE Office of Science User Facility. S. Zhu thanks the financial support from
11
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
Research Grant Council Postdoctoral Fellowship Scheme (PDFS2021-6S08). We thank Prof.
Mingjie Zhang and Mr. Rui Feng from HKUST for providing the NMR facility.
Conflict of Interest
The authors declare no conflict of interest.
Keywords
Copper nanowire • electrochemical CO2 reduction • surface oxidation • C2+ selectivity • stability
References
[1] D. Gao, R. M. Arán-Ais, H. S. Jeon, B. Roldan Cuenya, Nat. Catal. 2019, 2, 198–210.
[2] 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, I. Chorkendorff, Chem. Rev. 2019,
119, 7610–7672.
[3] R. M. Arán-Ais, D. Gao, B. Roldan Cuenya, Acc. Chem. Res. 2018, 51, 2906–2917.
[4] Y. Hori, I. Takahashi, O. Koga, N. Hoshi, J. Mol. Catal. A Chem. 2003, 199, 39–47.
[5] C. Reller, R. Krause, E. Volkova, B. Schmid, S. Neubauer, A. Rucki, M. Schuster, G. Schmid,
Adv. Energy Mater. 2017, 7, 1602114.
[6] C. Choi, T. Cheng, M. Flores Espinosa, H. Fei, X. Duan, W. A. Goddard, Y. Huang, Adv.
Mater. 2019, 31, 1805405.
[7] H. S. Jeon, S. Kunze, F. Scholten, B. Roldan Cuenya, ACS Catal. 2018, 8, 531–535.
[8] D. Kim, J. Resasco, Y. Yu, A. M. Asiri, P. Yang, Nat. Commun. 2014, 5, 1–8.
[9] Z. Q. Liang, T. T. Zhuang, A. Seifitokaldani, J. Li, C. W. Huang, C. S. Tan, Y. Li, P. De
Luna, C. T. Dinh, Y. Hu, Q. Xiao, P. L. Hsieh, Y. Wang, F. Li, R. Quintero-Bermudez, Y. Zhou,
P. Chen, Y. Pang, S. C. Lo, L. J. Chen, H. Tan, Z. Xu, S. Zhao, D. Sinton, E. H. Sargent, Nat.
Commun. 2018, 9, 3828.
[10] Y. Zhou, F. Che, M. Liu, C. Zou, Z. Liang, P. De Luna, H. Yuan, J. Li, Z. Wang, H. Xie, H.
Li, P. Chen, E. Bladt, R. Quintero-Bermudez, T. K. Sham, S. Bals, J. Hofkens, D. Sinton, G.
Chen, E. H. Sargent, Nat. Chem. 2018, 10, 974–980.
[11] H. Jung, S. Y. Lee, C. W. Lee, M. K. Cho, D. H. Won, C. Kim, H. S. Oh, B. K. Min, Y. J.
Hwang, J. Am. Chem. Soc. 2019, 141, 4624–4633.
12
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
[12] J. Kim, W. Choi, J. W. Park, C. Kim, M. Kim, H. Song, J. Am. Chem. Soc. 2019, 141,
6986–6994.
[13] D. Gao, I. Zegkinoglou, N. J. Divins, F. Scholten, I. Sinev, P. Grosse, B. Roldan Cuenya,
ACS Nano 2017, 11, 4825–4831.
[14] M. Luo, M. Zhou, R. R. Da Silva, J. Tao, L. Figueroa-Cosme, K. D. Gilroy, H. C. Peng, Z.
He, Y. Xia, ChemNanoMat 2017, 3, 190–195.
[15] J. C. Yang, B. Kolasa, J. M. Gibson, M. Yeadon, Appl. Phys. Lett. 1998, 73, 2841–2843.
[16] J. A. Eastman, P. H. Fuoss, L. E. Rehn, P. M. Baldo, G. W. Zhou, D. D. Fong, L. J.
Thompson, Appl. Phys. Lett. 2005, 87, 1–4.
[17] Z. Lyu, M. Xie, E. Aldama, M. Zhao, J. Qiu, S. Zhou, Y. Xia, ACS Appl. Nano Mater. 2019,
2, 1533–1540.
[18] H. Han, H. Choi, S. Mhin, Y.-R. Hong, K. M. Kim, J. Kwon, G. Ali, K. Y. Chung, M. Je, H.
N. Umh, D. H. Lim, K. Davey, S. Z. Qiao, U. Paik, T. Song, Energy Environ. Sci. 2019, 12,
2443–2454.
[19] D. DeNardis, D. Rosales-Yeomans, L. Borucki, A. Philipossian, Thin Solid Films 2006, 513,
311–318.
[20] M. Yin, C. K. Wu, Y. Lou, C. Burda, J. T. Koberstein, Y. Zhu, S. O’Brien, J. Am. Chem.
Soc. 2005, 127, 9506–9511.
[21] Z. Lyu, M. Xie, K. D. Gilroy, Z. D. Hood, M. Zhao, S. Zhou, J. Liu, Y. Xia, Chem. Mater.
2018, 30, 6469–6477.
[22] H. Guo, N. Lin, Y. Chen, Z. Wang, Q. Xie, T. Zheng, N. Gao, S. Li, J. Kang, D. Cai, D. L.
Peng, Sci. Rep. 2013, 3, 2323.
[23] L. L. Hung, C. K. Tsung, W. Huang, P. Yang, Adv. Mater. 2010, 22, 1910–1914.
[24] D. Raciti, M. Mao, C. Wang, Nanotechnology 2018, 29, 044001.
[25] A. D. Handoko, C. W. Ong, Y. Huang, Z. G. Lee, L. Lin, G. B. Panetti, B. S. Yeo, J. Phys.
Chem. C 2016, 120, 20058–20067.
[26] J. Huang, N. Hörmann, E. Oveisi, A. Loiudice, G. L. De Gregorio, O. Andreussi, N. Marzari,
R. Buonsanti, Nat. Commun. 2018, 9, 1–9.
[27] H. Mistry, A. S. Varela, C. S. Bonifacio, I. Zegkinoglou, I. Sinev, Y. W. Choi, K. Kisslinger,
E. A. Stach, J. C. Yang, P. Strasser, B. Roldan Cuenya, Nat. Commun. 2016, 7, 12123.
13
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
[28] F. Cavalca, R. Ferragut, S. Aghion, A. Eilert, O. Diaz-Morales, C. Liu, A. L. Koh, T. W.
Hansen, L. G. M. Pettersson, A. Nilsson, J. Phys. Chem. C 2017, 121, 25003–25009.
[29] Y. Zhao, X. Chang, A. S. Malkani, X. Yang, L. Thompson, F. Jiao, B. Xu, J. Am. Chem.
Soc. 2020, 142, 9735–9743.
[30] W. Tang, A. A. Peterson, A. S. Varela, Z. P. Jovanov, L. Bech, W. J. Durand, S. Dahl, J. K.
Nørskov, I. Chorkendorff, Phys. Chem. Chem. Phys. 2012, 14, 76–81.
[31] X. Feng, K. Jiang, S. Fan, M. W. Kanan, ACS Cent. Sci. 2016, 2, 169–174.
[32] H. Zhang, Y. Zhang, Y. Li, S. Ahn, G. T. R. Palmore, J. Fu, A. A. Peterson, S. Sun,
Nanoscale 2019, 11, 12075–12079.
[33] Z. Chen, M. Waje, W. Li, Y. Yan, Angew. Chem. Int. Ed. 2007, 46, 4060–4063.
[34] D. Huo, M. J. Kim, Z. Lyu, Y. Shi, B. J. Wiley, Y. Xia, Chem. Rev. 2019, 119, 8972–9073.
[35] T. T. Zhuang, Y. J. Pang, Z. Q. Liang, Z. Y. Wang, Y. Li, C. S. Tan, J. Li, C. T. Dinh, P. De
Luna, P. L. Hsieh, T. Burdyny, H. H. Li, M. X. Liu, Y. H. Wang, F. W. Li, A. Proppe, A.
Johnston, D. H. Nam, Z. Y. Wu, Y. R. Zheng, A. H. Ip, H. R. Tan, L. J. Chen, S. H. Yu, S. O.
Kelley, D. Sinton, E. H. Sargent, Nat. Catal. 2018, 1, 946–951.
[36] K. D. Yang, W. R. Ko, J. H. Lee, S. J. Kim, H. Lee, M. H. Lee, K. T. Nam, Angew. Chem.
Int. Ed. 2017, 56, 796–800.
[37] A. Loiudice, P. Lobaccaro, E. A. Kamali, T. Thao, B. H. Huang, J. W. Ager, R. Buonsanti,
Angew. Chem. Int. Ed. 2016, 55, 5789–5792.
[38] R. Reske, H. Mistry, F. Behafarid, B. Roldan Cuenya, P. Strasser, J. Am. Chem. Soc. 2014,
136, 6978–6986.
[39] C. W. Li, M. W. Kanan, J. Am. Chem. Soc. 2012, 134, 7231–7234.
[40] J. Fu, W. Zhu, Y. Chen, Z. Yin, Y. Li, J. Liu, H. Zhang, J. J. Zhu, S. Sun, Angew. Chem. Int.
Ed. 2019, 58, 14100–14103.
14
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
Figure 1. (a) TEM and (b) DF-STEM images of CuNWs covered by a thin, non-uniform sheath
of CuxO formed through oxidation by the O2 from air (denoted A-CuNWs). (c) EELS mapping,
(d) XRD pattern, and (e) XPS spectrum recorded from the same oxidized sample.
15
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
Figure 2. (a) TEM and (b) DF-STEM images of CuNWs covered by a smooth, relatively thick
(ca. 6 nm) sheath of CuxO formed through oxidation by aqueous H2O2 (denoted H-CuNWs). (c)
EELS mapping, (d) XRD pattern, and (e) XPS spectrum recorded from the same oxidized sample.
16
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
Figure 3. (a, b) Faradaic efficiencies of the catalysts based on A-CuNWs and H-CuNWs,
respectively. (c) Total and (d) partial current densities toward ethylene for the same types of
catalysts in 0.1 M KHCO3. The current densities were normalized to the geometric area.
17
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
Figure 4. (a) TEM, (b, c) BF-STEM images, and (d) XPS spectrum of A-CuNWs after 1 h of
electrolysis. (e) TEM, (f, g) BF-STEM images, and (h) XPS spectrum of H-CuNWs after 1 h of
electrolysis. A potential of -1.02 VRHE was applied for all measurements. The images in (c) and
(g) were taken from the regions marked by boxes in (b) and (f), respectively. The Cu
nanoparticles formed during electrolysis are marked by red circles in (a).
18
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
Figure 5. Stability tests for the catalysts based on (a, b) A-CuNWs and (c, d) H-CuNWs,
respectively, at -1.02 VRHE for 22 h: (a, c) Faradaic efficiencies of gaseous products and (b, d)
current densities normalized to the geometric area.
19
This article is protected by copyright. All rights reserved.

10.1002/anie.202011956
Angewandte Chemie International Edition
Table of Contents
Controlling the Surface Oxidation of Cu Nanowires Improves Their Catalytic Selectivity
and Stability toward C2+ Products in CO2 Reduction
Zhiheng Lyu,+ Shangqian Zhu,+ Minghao Xie, Yu Zhang, Zitao Chen, Ruhui Chen, Mengkun
Tian, Miaofang Chi, Minhua Shao,* Younan Xia*
In the electrochemical reduction of CO2, faradaic efficiencies as high as 57.7 and 52.0% are
achieved for ethylene when Cu nanowires oxidized by the O2 from air (A-CuNWs) and aqueous
H2O2 (H-CuNWs) serve as catalysts, respectively. The increased surface roughness greatly
enhances the C2+ selectivity of the catalysts, and the formation of a relatively thick, smooth oxide
sheath can improve the catalytic stability.
20
This article is protected by copyright. All rights reserved.