Breaking the Linear Scaling Relationship by Compositional and Structural Crafting of Ternary Cu–Au/Ag Nanoframes for Electrocatalytic Ethylene Production

Article abstract of DOI:10.1002/anie.202012631

Electrocatalytic conversion of carbon dioxide into high-value multicarbon (C₂₊) chemical feedstocks offers a promising avenue to liberate the chemical industry from fossil-resource dependence and eventually close the anthropogenic carbon cycle but is severely impeded by the lack of high-performance catalysts. To break the linear scaling relationship of intermediate binding and minimize the kinetic barrier of CO₂ reduction reactions, ternary Cu–Au/Ag nanoframes were fabricated to decouple the functions of CO generation and C?C coupling, whereby the former is promoted by the alloyed Ag/Au substrate and the latter is facilitated by the highly strained and positively charged Cu domains. Thus, C₂H₄ production in an H-cell and a flow cell occurred with high Faradic efficiencies of 69±5 and 77±2 %, respectively, as well as good electrocatalytic stability and material durability. In situ IR and DFT calculations unveiled two competing pathways for C₂H₄ generation, of which direct CO dimerization is energetically favored.

Full text of DOI:10.1002/anie.202012631

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Breaking Linear Scaling Relationship by Compositional and Structural Crafting of
Ternary Cu-Au/Ag Nanoframes
for Electrocatalytic Ethylene Production
Authors: Likong Xiong, Xiang Zhang, Hao Yuan, Juan Wang, Xuzhou
Yuan, Yuebin Lian, Huidong Jin, Hao Sun, Zhao Deng, Dan
Wang, Jiapeng Hu, Huiming Hu, Jinho Choi, Jiong Li, Yufeng
Chen, Jun Zhong, Jun Guo, Mark H Rümmerli, Lai Xu, and
Yang Peng
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.202012631
Link to VoR: https://doi.org/10.1002/anie.202012631

10.1002/anie.202012631
Angewandte Chemie International Edition
RESEARCH ARTICLE
Breaking Linear Scaling Relationship by Compositional and Struc
tural Crafting of Ternary Cu-Au/Ag Nanoframes for Electro-
catalytic Ethylene Production
Likun Xiong, ‡[a,b] Xiang Zhang,‡[a,b] Hao Yuan, ‡[c] Juan Wang,‡[d] Xuzhou Yuan, [a,b] Yuebin Lian,[a,b] Hui-
dong Jin, [a,b] Hao Sun, [a,b] Zhao Deng,*[a,b] Dan Wang, [a,b] Jiapeng Hu, [a,b] Huimin Hu, [a,b] Jinho Choi, [a,b]
Jiong Li, [d] Yufeng Chen,[c] Jun Zhong,*[c] Jun Guo,[e] Mark H. Rümmerli,[a,b] Lai Xu,*[c] and Yang
Peng*[a,b]
[a,b] L. K. Xiong, X. Zhang, H, Yuan, X. Z. Yuan, Y. B. Lian, H. D. Jin, H. Sun, D. Wang, J. P. Hu, H. M. Hu, Prof. Z. Deng, Prof. J. H. Choi, Prof. M. H. Rümmerli,
Prof. Y. Peng
Soochow Institute for Energy and Material Innovations (SIEMIS), College of Energy, Soochow University
Key Laboratory for Advanced Carbon Materials and Wearable Energy Technologies of Jiangsu Province, Soochow University, Suzhou, Jiangsu, P. R. China.
E-mail: zdeng@suda.edu.cn; ypeng@suda.edu.cn
[c] H. Yuan, Y. F. Cheng, Prof. J. Zhong, Prof. L. Xu
Institute of Functional Nano & Soft Materials (FUNSOM), Jiangsu Key Laboratory for Carbon-Based Functional Materials & Devices, Soochow University,
Suzhou, Jiangsu, P. R. China. E-mail: jzhong@suda.edu.cn; xulai15@suda.edu.cn
[d] Dr. J. Wang, Dr. J. Li, Shanghai Synchrotron Radiation Facility, Shanghai Advanced Research Institute, Chinese Academy of Sciences, P. R. China.
[e] Prof. J. Guo, Analysis and Testing Center, Soochow University, Suzhou, Jiangsu, P. R. China.
‡ These authors contributed equally.
art CO2 RR catalysts are still not satisfactory in either term of ac-
tivity, selectivity and stability[3]. A root cause of this lies in the in-
trinsic scaling relationship of intermediates binding associated
CO2RR, making it difficult to unify or optimize (in the case of syn-
gas production) product selectivity, especially when encountering
multiple reaction cascades with a spectrum of intermediates[4].
Thus, one of the central tasks in catalyst design for CO2RR
is to break the linear scaling relationship, which is both experi-
mentally and theoretically challenging and can only be done by
incorporating alternative energy forms or by increasing the com-
plexity of the catalyst[5]. For that many tactics have been explored
in constructing Cu-based catalysts, including multi-metallic alloy-
ing[6], multi-variant compositing[7], crystalline faceting[8],
nano/meso-structuring[9], strain engineering[10], subsurface-dop-
ing[11], adsorbate solvating[12], precatalytic oxidizing[13], as well as
surface reconstruction/tethering[14]. The common goal is to create
a material where the atoms are arranged to precisely manipulate
each gating mechanistic step of the reaction towards ultimate
unity of reduction products.
Abstract: Electrocatalytic conversion of carbon dioxide into
high-value multi-carbon (C2+) chemical feedstocks offers a prom-
ising avenue to liberate the chemical industry from fossil-resource
dependence and eventually close the anthropogenic carbon cycle,
but is severely impeded by the lack of high-performance catalysts.
Herein, aiming to break the linear scaling relationship of interme-
diates binding and minimize the kinetic barrier of CO2 reduction
reactions (CO2RR), ternary Cu-Au/Ag nanoframes are fabricated
to decouple the two pivotal functions of CO generation and C-C
coupling, where the former is greatly promoted by the alloyed
Ag/Au substrate and the latter is elegantly facilitated by the highly
strained and positively charged Cu domains. Thereby, extraordi-
nary Faradic efficiencies of 69 ± 5% and 77 ± 2% on C2H4 pro-
duction in H-cell and flow cell, respectively, are achieved with
great electrocatalytic stability and material durability. In-situ IR
and DFT calculations unveil two competing pathways for C2H4
generation, of which the direct CO dimerization is energetically
favored. By integrating the tandem effect, electronic modulation,
and defect engineering, the current study offers a paradigm in cat-
alyst design to break the linear scaling relationship towards im-
proved C2H4 production.
As an effort to approach this goal, herein ternary Cu-Ag/Au
nanoframes with highly strained and distorted Cu phase are tailor-
made to precisely steer the two pivotal steps towards C2+ produc-
tion, i.e., CO2 → *CO and 2 *CO→ *OCCO[15], since the alloyed
Ag/Au structures have been well recognized as one of the most
efficient and CO-specific catalyst[16], while Cu is highly potent in
Introduction
As the anthropogenic fossil fuel consumption and CO2 emis-
sion gradually approach an unsustainable level, technologies uti-
lizing CO2 as the feedstock for producing fuels and value-added
chemicals are becoming increasingly attractive and have incen-
tivized tremendous research efforts, yearning for closing the car-
bon cycle. Of particular interests, electrocatalytic CO2 reduction
reactions (CO2RR) enable to store the intermittent renewable
electricity into chemical bonds in grid scale and simultaneously
obtain deeply reduced hydrocarbons with high economic values[1].
At the core of this technology is the catalyst design and fabrication,
bearing the mission to minimize the energy input but maximize
the product yield and specificity[2]. Despite tremendous efforts and
some notable progress that have been made, so far state-of-the-
C-C coupling[17]
. In addition, the architecture of alloyed
nanoframes has been testified by previous studies capable of
greatly boosting the thermodynamics and kinetics of catalytic re-
actions such as ORR and MOR through specific surface enhance-
ment, d-band modulation, defect engineering, and active site ex-
posure[18]. Therefore, we expect the as-fabricated Cu-Ag/Au
nanoframes should manifest great prospect in efficient CO2RR
with high C2+ selectivity by breaking the scaling relations, which
will be closely inspected in the following sections with mechanistic
understanding aided by operando and DFT investigations.
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202012631
Angewandte Chemie International Edition
RESEARCH ARTICLE
Figure. 1 The evolution and atomic structure of Ag3Au nanoframes. a-d) From top to bottom are schematics, SEM images, and HAADF-STEM images
with EDS mapping of Ag (red) and Au (yellow) for pre-synthetic Ag nanocubes, Ag15Au NPs collected at the 5th min, Ag6Au NCs collected at the 15th min, and
Ag3Au NFs (respectively from left to right). e, f) High-resolution lattice details viewed by Cs-TEM on Ag3Au NFs displaying stepped surfaces, grain boundaries,
edge dislocations and vacancies.
nanoframes in aqueous condition, enabling not only the exposure
of clean facets, but also the subsequent surface amendments.
The detailed procedure for preparing the cubic Ag3Au
nanoframes (Ag3Au NFs) is given in the section of Methods.
Results and Discussion
Briefly, single-crystallinity Ag nanocubes were first produced via
a conventional polyol method, exhibiting uniform particle size with
an average edge of 51.0 ± 3.9 nm (Fig. 1a, Supplementary Fig.
1a and 2). Cetyltrimethylammonium chloride (CTAC) as the struc-
ture-inducing agent and ascorbic acid (AA) as the reducing agent
were then added together with chloroauric acid (HAuCl4) to yield
the desired framework structure after reacting for 24 minutes, re-
sulting in nanoframes with a final Ag/Au ratio of 3:1 and increased
size of 58.5 ± 3.1 nm (Fig. 1d, Supplementary Fig. 1d, 3 and Table
1). To unravel the growth mechanism as detailed in the Supple-
mentary note, we deliberately arrested the reaction intermediates
at the 5th and 15th min during the reaction (Fig. 1b, c), with all the
Previously reported methods for synthesizing bimetallic
Ag/Au nanoframes typically require a “tour de force” endeavor in-
volving controlled deposition of Au on Ag templates, followed by
selective etching of the latter[19]. Challenges remain to control the
structural uniformity and batch-to-batch reproducibility. In our pre-
vious study we were able to deploy a facile one-pot tactic for ob-
taining ultrathin octahedral Au3Ag nanoframes, but the employ-
ment of hydrophobic octadecylamine as both the solvent and cap-
ping agent restrains further compounding with Cu, which has
been deemed as the necessary constituent for obtaining C2+ prod-
ucts in CO2RR in high efficiency. To sidestep this limitation, herein
a modified protocol is adopted to synthesize cubic Ag/Au
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202012631
Angewandte Chemie International Edition
RESEARCH ARTICLE
obtained samples carefully characterized with scanning electron
microscopy (SEM), high-angle annular dark-field scanning trans-
mission electron microscopy (HAADF-STEM) and energy-disper-
sive X-ray spectroscopy (EDS) (Fig. 1a-d).
cally active surface area (ECSA), which is determined by meas-
uring the double-layer capacitance and applied throughout this
study when presenting current densities. Despite that Ag3Au NFs
exhibit an ECSA 3.2 times and 3.8 times larger than those of Au
and Ag nanocubes, respectively (Supplementary Fig. 8, Supple-
mentary Table 2), the normalized current density of the former is
still significantly higher. What’s more, the CO partial current den-
sity (jco) obtained at −0.7 VRHE for Ag3Au NFs is about 6 and 60
times higher than those of Au and Ag nanocubes, respectively
(Supplementary Fig. 9). These observations suggest Ag3Au NFs
possess intrinsically higher activities for electrocatalytic CO2 con-
version, which is further evidenced by the positively shifted onset
potential of CO2RR (Fig. 2c). Obviously, the alloyed atomic con-
figuration and abundant structural defects on Ag3Au NFs must
have greatly promoted CO2 activation, both thermodynamically
and kinetically, and therefore significantly lowered the required
overpotential and augmented the current density for CO2 conver-
X-ray diffraction (XRD) patterns taken on the Ag3Au NFs con-
firmed the alloyed Ag/Au phase without obvious phase segrega-
tion (Supplementary Fig. 4). Both the diffraction peaks at 38.16°
and 44.42°, respectively corresponding to (111) and (200) planes,
are situated between the corresponding peaks of pure Au and Ag,
indicative of atomic mixing of the two metals. This echoes the ob-
servation from EDS elemental mapping, showing homogeneous
and mingled Au and Ag distribution (Fig. 1d). Spherical-aberra-
tion-corrected TEM (Cs-TEM) was employed to further elucidate
the atomic structure of Ag3Au NFs, showing a full spectrum of lat-
tice defects and atomic misalignment such as stepped surfaces,
grain boundaries, point vacancies, edge/screw dislocations, and
twin facets etc. (Fig. 1e, f and Supplementary Fig. 5). The rich and
abundant lattice defects enable serving as not only the catalyti-
cally active sites, but also the nucleation sites for adducting het-
eroatoms.
sion[20]
.
Fig. 2c displays the Tafel slopes of Ag3Au NFs, Au and Ag
nanocubes, acquired at similar CO partial current densities, are
48.3, 88.5, and 109.1 mV dec−1, respectively. It is well known in
the two-electron pathway of CO2RR to CO, the first step is the
transformation of CO2 into *CO2¯ or *COOH, followed by the sub-
The capability of Ag3Au NFs in electrocatalytic CO2RR was
examined in CO2-saturated 0.1 M KHCO3 aqueous electrolyte,
and compared to pure Au and Ag nanocubes with similar particle
size but minimal crystalline defects (Supplementary Fig. 6 and 1a). sequent proton-coupled electron transfer (PCET) to yield CO and
As expected for typical Au and Ag based catalysts, CO and H2
were the only two reduction products detected. Ag3Au NFs exhib-
ited a decent CO selectivity beyond 90% faradic efficiency (FE) in
a broad potential range from −0.6 to −0.9 V vs. the reversible hy-
drogen electrode (RHE, all potentials in this study are henceforth
referenced to this format), and reached a maximum of 97% at
−0.72 VRHE (Fig. 2a and Supplementary Fig. 7). In contrast, for Au
OH-/H2O. Prior studies suggest the first step is usually rate-deter-
mining[21], but this is apparently not for the case of Ag3Au NFs.
Here, as the witnessed Tafel slope was only 48.3 mV dec−1, a pre-
equilibrium electron transfer step before the rate-determining step
(RDS) is implied. This strongly suggests the activation of CO2 on
Ag3Au NFs is greatly facilitated, likely by the structural defects,
whereas on Au and Ag nanocubes the first activation step could
and Ag nanocubes the highest FE were 70% at −0.8 VRHE and 86% be still rate-determining. Noteworthy, the Ag3Au NFs also demon-
at −1.0 VRHE, respectively. Fig. 2b shows the linear scanning volt-
ammograms of these three samples normalized to electrochemi-
strated great electrolytic stability, retaining nearly unchanged FE
of ~97% for CO production and ~80% of the initial current density
after a prolonged 10 h chronoamperometric test at −0.72 VRHE (Fig.
Figure 2. CO2RR performance of Ag3Au NFs in a CO2-saturated 0.1 M KHCO3 aqueous solution. a) Faradaic efficiency of CO on Ag3Au NFs and control
samples at different applied potentials. The white error bars represent standard deviations from at least three independent measurements. b) Linear sweep
voltammetry (LSV) curves for the Ag3Au NFs and control samples at a sweep rate of 20 mV s-1. c) The Tafel plots at similar CO partial current densities. d)
CO2RR stability of Au3Ag NFs at −0.72 VRHE. Left axis is the total current density normalized to ECSA and right axis is the CO Faradaic efficiency.
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202012631
Angewandte Chemie International Edition
RESEARCH ARTICLE
2d). Taken together, the as-synthesized Ag3Au NFs demonstrate
superb capacity in fleetly and selectively converting CO2 to CO,
endorsing their deployment as the CO generator for subsequent
C2+ production through a tandem fashion on further deposited
copper, which will be carried on below.
Cs-TEM further testifies the strained Cu layer on Cu3-Ag3Au
NFs, showing highly skewed lattice where lattice misalignment,
stacking faults, edge dislocations, and stepped staircases are
conspicuous and prevailing (Fig. 3d, Supplementary Fig. 13). In
order to discern whether or not this highly skewed lattice structure
is due to surface reconstruction under the post-synthetic condition
of Cu deposition, Ag3Au NFs treated with the same reaction con-
dition but without adding Cu2+ were also examined by Cs-TEM,
revealing less distorted lattice configuration (Supplementary Fig.
14) but similar crystallographic features to those of the original
Ag3Au NFs (Supplementary Fig. 5). It is thus evident that the
skewed lattice belongs to the deposited Cu phase, which is further
justified by analyses on the lattice fringes. As shown in Fig. 3d,
the area embraced within the dotted yellow lines expose majorly
two facets of Cu, i.e., Cu (110) and Cu (200) with d-spacings of
0.262 and 0.182 nm, respectively. Outside the bracketed zone,
lattice fringes of 0.141 and 0.237 nm are attributed to the alloyed
Ag/Au (220) and (111) planes, respectively. Further zooming into
the interface between the Cu and Ag/Au boundaries, distinct
phase segregation with few atomic interdiffusion can be visual-
ized (Supplementary Fig. 15), echoing the above EDX and XRD
results. Once again, the highly abrupt Cu lattice expressing rich
crystallographic abnormalities should be a consequence of
stress-release to alleviate the interfacial tension caused by lattice
The epitaxial deposition of Cu on Au/Ag at room temperature
poses a synthetic challenge because of the large lattice mismatch
(atomic mismatch of 11.1%) and hence the tendency of Cu self-
nucleation[22]. To achieve this goal high temperature and non-
aqueous solvents are typically required, but might simultaneously
damage the parent structure. Fortunately, in our case, the abun-
dant defects on Ag3Au NFs are capable of serving as the nuclea-
tion sites for Cu deposition under mild synthetic conditions. As il-
lustrated by the schematic diagram in Fig. 3a, the hydrazine-re-
duced Cu atoms are first deposited onto the defect sites of Ag3Au
NFs and then grow into larger domains. Thereby, three Cu-Ag/Au
ensembles were obtained by raising the Cu2+ feeding (Fig. 3b and
Supplementary Fig. 10), and named as Cu1-Ag3Au, Cu3-Ag3Au,
and Cu9-Ag3Au NFs based on the Cu atomic ratio co-verified by
ICP, EDS and X-ray photoelectron spectroscopy (XPS) (Supple-
mentary Table 3). EDS mapping images of the Cu3-Ag3Au NFs in
Fig. 3c clearly show Cu is separated from the alloyed Ag/Au
phase, confirming the epitaxial deposition of Cu on the Ag3Au
frameworks. The phase segregation is further witnessed by XRD,
showing separated diffraction peaks for the two constituents
(Supplementary Fig. 11). A closer look at the Cu (111) peak at
~43.2 reveals a slight 2 shift towards lower angles with de-
scending Cu content, indicative of higher strain induced by the
underlying Au/Ag lattice when the amount of deposited Cu is less
(Supplementary Fig. 12).
mismatch[23]
.
To probe the elemental interplay of the ternary structure
with respect to electronic structure modulation, XPS was taken on
Cux-Ag3Au NFs in comparison to Ag3Au NFs and Cu nanocubes.
Compared to the pristine Cu, all samples of Cux-Ag3Au NFs ex-
hibit a down-shift of binding energy for the Cu 2p orbitals, and the
Figure 3. micro-structural characterization of Cux-Ag3Au. a, Schematic diagram for Cu deposition on Ag3Au NFs with increasing Cu coverage. b, TEM
images of Cu1-Ag3Au NFs, Cu3-Ag3Au NFs and Cu9-Ag3Au NFs. c, EDX elemental mapping of Cu (purple), Ag (green) and Au (blue) on Cu3-Ag3Au NFs. d, A
representative Cs-TEM image of Cu3-Ag3Au NFs taken on one of the edge sites, showing the interlaced Cu and Ag3Au phases with rich lattice defects. The
area embraced by the yellow dotted lines belongs to Cu phase, where lattice fringes with d-spacings of 0.262 and 0.182 nm correspond to the Cu (110) and
(200) facets, respectively. Outside the Cu zone, lattice fringes of 0.141 and 0.237 nm closely match the lattice distances of alloyed Ag/Au (220) and (111)
planes, respectively.
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202012631
Angewandte Chemie International Edition
RESEARCH ARTICLE
shifting amount decreases with increasing Cu deposition (Supple-
mentary Fig. 16a), similar to the trend of XRD observations (Sup-
plementary Fig. 12). This strongly suggests the valence states of
Cu is modulated by the underlying Ag/Au substrate. Meanwhile,
both the Au 4f and Ag 3d orbitals shift to higher binding energies
with increasing Cu deposition, indicating the loss of valence elec-
trons to Cu (Supplementary Fig. 16b, c). This observation of
charge redistribution seems to contradict with the electronegativ-
ity argument as Au is the most electronegative metal among the
three. We noticed that previous studies involving binary Ag/Cu or
Au/Cu composites also witnessed the similar blue-shift of Ag 3d
and Au 4f binding energies with increasing Cu, but still appointed
Ag and Au as the electron acceptor in virtue of their lower chemi-
cal potentials[6a, 24]. Thus, XPS alone based on valence electron
analysis might not be able to accurately disclose the charge re-
distribution and Coulombic interaction in these intricate many-
body systems. In fact, to explain this apparent inconsistency, a
charge compensation model has often been called upon, that is,
the lost d charges from Ag/Au can be overcompensated by the
gain of mainly s-p charges from Cu[25], resulting in a net gain of
charges on the Ag and Au sites (and thus an overall charge loss
for Cu) in favor of the electronegativity and thermodynamics ar-
guments. Indeed, this explanation is well supported by our X-ray
absorption near-edge spectroscopy (XANES) study on Cu3-
Ag3Au NFs, showing more depleted Cu s-p states above the
Fermi level in comparison to the pristine Cu, as evidenced by the
enhanced intensity of pre-edge adsorption (Supplementary Fig.
17a). The XANES further shows the oxidation state of Cu in Cu3-
Ag3Au NFs is situated between Cu0 and Cu+. While the Fourier
transform extended X-ray absorption fine structure (FT-EXAFS)
reveals no Cu-O coordination on Cu3-Ag3Au NFs, indicating the
absence of copper oxides, the Cu-Cu coordination is significantly
reduced in comparison to Cu foil (Supplementary Fig. 17b). Taken
together, we are able to conclude that the electronic structure of
Cu can be effectively modulated by not only its own defective lat-
tice configuration, but also the underlying Ag/Au substrate
through d-d interaction and s-p overcompensation in conformity
to the electroneutrality and electronegativity rules.
The CO2RR performance of Cux-Ag3Au were first tested un-
der the same electrolytic condition as that of Ag3Au NFs in H-cell,
with liquid products quantified by 1H nuclear magnetic resonance
(NMR). In addition, Cu nanocubes of 50-60 nm (Supplementary
Fig. 18) with and without mixing with Ag3Au NFs were studied as
controls, since the Cu (100) facet has been well recognized in fa-
vor of driving C2+ selectivity[8]. Compared to Ag3Au NFs (Supple-
mentary Fig. 19), Cu1-Ag3Au NFs exhibit depressed activity for
CO production with a maximal FE of 85% at −1.0 VRHE (Fig 4a)
and a negatively shifted onset potential of −0.54 VRHE (Supple-
mentary Fig. 20). At more negative potentials lower than −1.2
VRHE, a small amount of ethylene become detectable with con-
comitant hydrogen yield. Both the decreased activity and selec-
tivity of Cu1-Ag3Au NFs towards CO production are likely due to
the partially shrouded active sites on Ag3Au NFs by the deposited
Cu domains.
As the Cu domains continually grow, it becomes the domi-
nant phase to dictate the overall behavior of CO2RR, yielding dra-
matically different product distribution on Cu3-Ag3Au NFs consist-
ing multiple C1+ and C2+ products (Fig. 4b). Interestingly, the FE
and partial current density of CO on Cu1-Ag3Au NFs and those of
C2H4 on Cu3-Ag3Au NFs at the corresponding potentials seem to
be well correlated (Fig. 4a, b, Supplementary Fig. 21a, b), show-
ing similar variation trends with the applied bias. Specifically, the
maximal FE and partial current density of C2H4 on Cu3-Ag3Au NFs
is 69 ± 5% and 13.06 mA cm-2 at −1.2 VRHE, coinciding with those
Figure 4. CO2RR performance of Cux-Ag3Au. a-c) The FE and selectivity of various reduction products at different bias tested in 0.1 M KHCO3 aqueous
solution for Cu1-Ag3Au NFs, Cu3-Ag3Au NFs, and Cu9-Ag3Au NFs, respectively. The white error bars represent standard deviations from at least three
independent measurements. d) Chronoamperometric i–t curve (left axis) of Cu3-Ag3Au NFs along with the corresponding FE of ethylene (right axis) at −1.2
VRHE for a period of 31 h.
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202012631
Angewandte Chemie International Edition
RESEARCH ARTICLE
1
high values of CO (81 ± 6% and 12.85 mA cm-2) on Cu1-Ag3Au
NFs at the same potential. This suggests a tandem mechanism of
C-C coupling with the overflow of CO from the Ag/Au phase to Cu
domains, since the absorbed CO is well known as a crucial inter-
mediate for ethylene production[17]. In fact, Jaramillo’s group has
reported a diffusion model based on the Poisson-Nernst-Planck
theory depicting the diffusion length of CO from Au to Cu via a
shared interface are 3 to 4 orders of magnitude lower than that
crossing the Helmholtz layer at the electrode/electrolyte inter-
face[6c]. This argument is further supported by our control studies
employing the physically mixed Ag3Au NFs and Cu nanocubes in
a mass ratio of 1:1 (Supplementary Fig. 22). At the potential of
−1.2 VRHE, the mixed catalysts showed an FE of 31% for C2H4,
which is 1.24 times greater than that of the Cu nanocubes alone
(Supplementary Fig. 23), indicating that Ag3Au NFs do contribute
to producing C2H4 via a tandem fashion. Nonetheless, the eth-
ylene selectivity of this mixed catalyst is still far inferior to that of
Cu3-Ag3Au NFs, signifying the importance of atomic proximity be-
tween Cu and the Ag/Au substrate in promoting the concentration
of absorbed CO on the former. When the Cu phase becomes ex-
cessive, as is the case of Cu9-Ag3Au NFs, a large proportion of
the Ag3Au NFs is covered by Cu. This weakens the CO spillover
effect and eventually leads to both reduced CO and C2H4 produc-
tion (Fig 4c). Collectively, our observations here strongly suggest
that a delicate balance and close proximity between the Cu and
Ag/Au phases is pivotal for exploiting the tandem reactions to-
wards efficient C-C coupling.
Using H NMR to examine the liquid products produced by
Cu3-Ag3Au NFs, we found some minor oxygenated reduction
products at relatively higher potentials between −0.6 and −1.2
VRHE (Supplementary Fig. 24), which is in line with previous re-
ports but out of the scope of current study[6b, 6c]. At more negative
bias, both the FE and partial current density of CH4 gradually in-
crease, achieving an appreciable FE of 40% at −1.4 VRHE on Cu3-
Ag3Au NFs, but at the expense of ethylene FE drop to 43%. This
enhanced methane production at more negative potentials was
also observed by others and attributed to a drastically different
CO2RR pathway involving absorbed *H (rather than *CO) as the
key intermediate, followed by CO adduction[6a]. Lastly, it is impres-
sive to see Cu3-Ag3Au NFs exhibit a fairly stable FE of C2H4 pro-
duction with only about 13% decrease in current density over the
31 h continuous electrolytic process (Fig. 4d), which is atop the
state-of-the-art CO2 RR catalysts today (Supplementary Table 4).
What’s more, to test the adaptability of Cu3-Ag3Au NFs in
flow-cell configuration, CO2 RR was carried out on gas-diffusion
electrodes in 1 M KOH (Supplementary Fig. 25). As show in Fig.
5a, under such a strong alkaline condition the potentials to yield
deeply reduced hydrocarbons are upshifted significantly. While
the overall trend of C2H4 production with varying bias is similar to
that observed in H cell (Fig. 4b), the highest C2H4 FE of 77 ± 2%
is now observed at −0.65 VRHE. The polarization curve displayed
in Fig. 5b shows the electrolytic current density can surge to 400
mA cm-2 until the potentiostat limit is reached. Further chronoam-
perometric i-t test at a fixed potential of −0.65 VRHE (Fig. 5b inset)
Figure 5. CO2RR performance of Cu3-Ag3Au in a flow cell and operando XAFS and IR characterizations. a) Potential-dependent FE of various reduction
products on Cu3-Ag3Au NFs in 1M KOH aqueous solution. The white error bars represent standard deviations from at least three independent measurements.
b) Linear sweep voltammetry curve of Cu3-Ag3Au NFs in 1M KOH at a scan rate of 10 mV s–1; Insert is the chronoamperometric i–t curve (left axis) of Cu3-
Ag3Au NFs along with the corresponding FE of ethylene (right axis) at −0.65 VRHE for a testing period of 6 h. c) Operando copper K-edge XAFS spectra of Cu3-
Ag3Au NFs at 0 min (green), 60 min (blue), and 80 min (cyan) in 0.1 M KHCO3; Copper foil (pink) and Cu2O (orange) are included for references. d) Operando
ATR-IR spectra recorded while ramping down the potential from −0.4 to −1.2 VRHE on Cu3-Ag3Au NFs.
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202012631
Angewandte Chemie International Edition
RESEARCH ARTICLE
shows the catalyst is capable of delivering a stabilized current
density of ~232 mA cm-2 up to 6 hours with the FE maintained
above 60% throughout the testing period. Moreover, SEM and
TEM images taken on the post-electrolytic catalyst reveal no ma-
jor structural changes, despite a few fragmented nanoframes, af-
ter the extensive electrolysis (Supplementary Fig. 26). Impres-
sively, CS-TEM can still reveal atomic lattice defects and strains
on the frameworks as illustrated in Supplementary Fig. 27. There-
fore, in our flow-cell demonstration Cu3-Ag3Au NFs manifests
great perspective in catalyzing CO2 RR under heavy-duty condi-
tions with great C2H4 selectivity.
To gain deeper insights into the CO2RR mechanistics on
Cu3-Ag3Au NFs, operando XAFS were carried out to inspect the
Cu valence state change with time (Fig. 5c) and in-situ attenuated
total reflection surface-enhanced infrared absorption spectros-
copy (ATR-SEIRAS) were employed to monitor the evolution of
absorbed reaction intermediates from −0.4 to −1.2 VRHE in CO2-
saturated 0.1 M KHCO3 (Fig. 5d, Supplementary Fig. 28). Under
white-line peaks (Fig. 5c), and are nearly identical to those of the
original powder samples (Supplementary Fig. 29). These obser-
vations unequivocally testify the presence of structurally induced
Cu+ (0 <  < 1) in Cu3-Ag3Au NFs throughout the whole testing
period even under highly reductive environment. Although being
controversial, some previous studies pointed out the presence of
Cu+ is in favor of intermediates binding during CO2RR and helps
boost the CO dimerization to form C2H4 at modest overpoten-
tials[13]
.
Supplementary Table 5 summarizes the detected CO2RR in-
termediates in the ATR-SEIRAS test as well as the corresponding
band assignments. Four major peaks can be explicitly discerned
from the serial IR spectra between 1450 and 2150 cm-1. The band
at 2075 cm-1 emerging from -0.5 VRHE is attributed to the C≡O
[26]
stretching mode of linear-bonded CO adsorbates (COad,L
)
,
reaching its maximum intensity at -1.0 VRHE and then decreasing
again at lower potentials. This indicates the initial formation of CO
intermediates and subsequent transmutation into other hydrocar-
Figure 6. Mechanistic investigation by DFT calculations. a) The free energy changes of the first two hydrogenation steps on Ag(100)-Cu(100) (cyan),
Au(100)-Cu(100) (orange) and Ag3Au(100)-Cu(100) (blue) for yielding *CO intermediates. b) Partial density of states (PDOS) of the proposed structure with
absorbed CO. The area highlighted by the yellow background indicates coupling between the 3d orbitals of Cu and the 2p orbitals of C and O. c) The free
energy of intermediates-binding along the reaction coordinates following the two pathways towards C2H4 and the pathway towards CH4 formation at 0 V. Insets
are the corresponding intermediate configurations on the Au/Ag substrate (green background) and Cu domains (blue background). d) A general mechanistic
overview of ethylene production on Cu3-Ag3Au NFs. The Ag3Au substrate serves as the efficient CO generator for subsequent ethylene formation on segregated
Cu domains, following majorly the direct dimerization pathway, but also the pre-hydrogenation to form *CHO first.
the applied bias of −1.0 VRHE, the XANES spectra acquired at 0,
60 and 80 min show almost completely overlapped pre-edge and
bons at more negative bias. It is worthy to note that the highest
CO peak intensity at -1.0 VRHE coincides with the maximal CO
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202012631
Angewandte Chemie International Edition
RESEARCH ARTICLE
partial current density at the same potential observed for Cu3- Conclusion
Ag3Au NFs (Supplementary Fig. 21b). Meanwhile, the peak
shows a blue shift with increasing bias, suggesting a strong elec-
tronic interaction between CO and the metal substrate that causes
the Stark shift.
For multi-electron participated electrode reactions such as
CO2RR involving a variety of intermediates, breaking the linear
scaling relationship of intermediates binding and minimizing the
kinetic barrier of RDS is paramount in catalyst design for promot-
ing both electrocatalytic activity and selectivity. Following this
tenet, ternary catalysts were fabricated in the current study to sep-
arate the two functions of CO generation and C-C coupling in
CO2RR, where the former is greatly promoted by the alloyed
Ag/Au substrate and the latter is elegantly facilitated by the highly
strained and distorted Cu domains. In addition, the nanoporous
framework design of the catalyst is further beneficial for the expo-
sure of active sites, as well as the mass transportation of ions and
gases. As such, the optimized catalyst of Cu3-Ag3Au NFs pre-
sents an extraordinary FE of 69 ± 5% and 77 ± 2% on C2H4 pro-
duction in H-cell and flow cell, respectively, apart from remarkable
electrocatalytic stability and material durability. Through operando
IR spectroscopy and DFT calculations, we unveiled two compet-
ing pathways for C2H4 generation, of which the direct CO dimeri-
zation is suggested to be favored at low overpotentials. At more
negative bias, the intermediate of *CHO is involved, leading to the
yield of both C2H4 and CH4. It is further suggested that the energy
barrier of the RDS is greatly lowered by the positively charged Cu
with highly skewed lattice due to the strong lattice mismatch and
electronic interaction between the Ag/Au substrate and Cu adden-
dum. Consequently, this study offers a paradigm to break the lin-
ear scaling relationship of CO2RR towards improved C2H4 produc-
tion by integrating the tandem effect, electronic modulation, defect
engineering, and open structure design.
The band at 1725 cm-1, intensifying with descending poten-
*
tials, is in close proximity to the stretching of absorbed CHO[27]
,
which serves as not only the precursor for subsequent C-C cou-
*
pling to form OCCHO, but also a key intermediate for the for-
mation of CH4 at more negative bias. While water shows a wide
O-H band within the range of 1600-1700 cm-1, the peak located at
1563 cm-1 growing conspicuously with decreasing potential can
be assigned to the C-O vibration of absorbed *OCCOH, a strong
evidence for the direct CO-CO dimerization towards C2H4 for-
mation[28]. Taken together from these spectroscopic observations,
there should exist two competing reaction pathways for C2H4 for-
*
mation on Cu3-Ag3Au NFs. One involves the dimerization of CO
to form *OCCO, followed by surface hydrogenation or proton-cou-
*
pled electron transfer to form OCCOH. The other engages the
absorbed *H or *CO to form *CHO first, and then the subsequent
CO-CHO coupling.
First-principle calculations were further performed to illumi-
nate the tandem process and reaction energetics towards C2H4
formation (see computing details in Methods and Supplementary
information). DFT results show that the Cu-born Ag3Au (100) sub-
strate (Supplementary Fig. 30) could effectively catalyze the first
*
hydrogenation step from CO2 to COOH requiring only 0.31 eV
reaction free energy, in contrast to that of 0.83 and 0.81 eV for the
Cu-born Ag (100) and Au (100) surfaces, respectively (Figure 6a,
Supplementary Fig. 31). This suggests the alloyed Ag/Au phase
is highly in favor of CO production, endorsing its role as the CO
generator for subsequent deep reduction on Cu. Next, the adsorp-
*
tion energies of CO (ΔGco) on various sites of Cu3-Ag3Ag NFs
Acknowledgements
were examined (Supplementary Table 6). The calculated ΔGco on
the sites of pristine Ag3Au substrate nonadjacent to Cu is −0.47
eV, whereas on sites near and atop the Cu cluster the values are
−0.96 and −1.01 eV, respectively. This indicates the CO overflow
from the Ag3Au substrate to the Cu cluster is thermodynamically
driven.
This work is supported by Natural Science Foundation of China
(No. 21701118, No. 91961120, No.U1932211), Natural Science
Research Project of Jiangsu Higher Education Institutions of
China (18KJA480004), the Key Technology Initiative of Suzhou
Municipal Science and Technology Bureau (SYG201934) Six Tal-
ent Peaks Project in Jiangsu Province (No. XCL-057, XCL-062,
TD-XCL-006) and Priority Academic Program Development
(PAPD) of Jiangsu Higher Education Institutions. We thank Soo-
chow University Facility Center for use of their shared facilities.
Also we thank Thermal Fischer for helping set up the in-situ IR. In
particular, we acknowledge the help from Taiwan Light Source
(TLS) and Shanghai Synchrotron Radiation Facility (SSRF). This
project is also funded by the Collaborative Innovation Center of
Suzhou Nano Science & Technology the 111 Project, and Joint
International Research Laboratory of Carbon-Based Functional
Materials and Devices.
Once *CO is migrated to the Cu cluster, whether it is hydro-
genated or directly coupled with another CO molecule would be
the key step that determines the reaction path[15]. Our calculation
results suggest that the direct coupling process requires 0.50 eV
of reaction free energy, less than that required for the hydrogena-
tion of *CO to *CHO (0.87 eV, Figure 6c). Moreover, the calcu-
lated partial density of states (PDOS) shown in Figure 6b suggest
close coupling between Cu-3d and C/O-2p orbitals, which should
also facilitate the charge transfer for CO dimerization. Therefore,
the C2 path through direct CO-CO dimerization is more favored at
low overpotentials, while at more negative bias, the CO2 reduction
reaction could undergo both C2 pathways shown in the inset of
Figure 6c. Note that at even lower reduction potentials, the C1
path for methane formation can be taken[29], requiring an addi-
tional 0.67 eV from *CHO. These results are in good agreement
with the potential-dependent C2H4 and CH4 selectivity on Cu3-
Ag3Au NFs witnessed from experiments. Summing up, our calcu-
lations here clearly show the CO spillover from Au/Ag to Cu is
thermodynamically favored, and both the *CO*CO coupling and
*CO→*CHO hydrogenation steps are rate-determining and mech-
anistics-gating towards ethylene production (Figure 6d).
Author contributions
Y. P. conceived the idea and led the project. L. K. X. and X. Z.
designed and performed the majority of the experiments. D. W.
and M. H. R. took the Cs-TEM images and analyzed the data. H.
M. H., J. C., H. Y. and L. X. carried out the DFT calculations. Y. F.
C., J. W., J. L. and J. Z carried out XAFs characterizations. X. Z.
Y., H. D. J. and H. S. helped on the SEM characterizations. J. P.
H., J. G. and Y. B. L took the TEM images. Z. D. devoted to the
8
This article is protected by copyright. All rights reserved.

10.1002/anie.202012631
Angewandte Chemie International Edition
RESEARCH ARTICLE
Surf. Sci. 2016, 654, 56-62; d) W. Luo, X. Nie, M. J. Janik, A. Asthagiri,
ACS Catal. 2016, 6, 219-229.
electrochemical data analysis. L. K. X., Z. D. and Y. P. wrote the
paper. All the authors have discussed the results.
[9]
a) R. Reske, H. Mistry, F. Behafarid, B. Roldan Cuenya, P. Strasser, J.
Am. Chem. Soc. 2014, 136, 6978-6986; b) C. Choi, T. Cheng, M. Flores
Espinosa, H. Fei, X. Duan, W. A. Goddard, Y. Huang, Adv. Mater. 2019,
31, 1805405; c) K. Manthiram, B. J. Beberwyck, A. P. Aivisatos, J. Am.
Chem. Soc. 2014, 136, 13319-13325; d) 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.
Keywords: Cu-Au/Ag Nanoframes • electroreduction • CO2 • lin-
ear scaling relationship • ethylene
[10] a) R. Reske, M. Duca, M. Oezaslan, K. J. P. Schouten, M. T. Koper, P.
Strasser, J. Phys. Chem. Lett. 2013, 4, 2410-2413; b) E. L. Clark, C.
Hahn, T. F. Jaramillo, A. T. Bell, J. Am. Chem. Soc. 2017, 139, 15848-
15857; c) T. Adit Maark, B. R .K. Nanda, J. Phys. Chem. C. 2017, 121,
4496-4504.
[11] a) 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; b) T. T. Zhuang, Z. Q. Liang, A. Seifitokaldani, Y. Li,
P. De Luna, T. Burdyny, F. Che, F. Meng, Y. Min, R. Quintero-Bermudez,
C. T. Dinh, Y. Pang, M. Zhong, B. Zhang, J. Li, P. N. Chen, X. L. Zheng,
H. Liang, W. N. Ge, B.J. Ye, D. Sinton, S. H. Yu, E. H. Sargent, Nat.
Catal. 2018, 1, 421-428; c) 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.
Reference
[1]
a) R. P. Ye, J. Ding, W. Gong, M. D. Argyle, Q. Zhong, Y. Wang, C. K.
Russell, Z. Xu, A. G. Russell, Q. Li, M. Fan, Y. G. Yao, Nat. Commun.
2019, 10, 5698; b) P. De Luna, C. Hahn, D. Higgins, S. A. Jaffer, T. F.
Jaramillo, E. H. Sargent, Science 2019, 364, 3506.
[2]
[3]
[4]
M. G. Kibria, J. P. Edwards, C. M. Gabardo, C. T. Dinh, A. Seifitokaldani,
D. Sinton, E. H. Sargent, Adv. Mater. 2019, 31, 1807166.
D. Gao, R. M. Arán-Ais, H. S. Jeon, B. Roldan Cuenya, Nat. Catal. 2019,
2, 198-210.
a) Y. Y. Birdja, E. Pérez-Gallent, M. C. Figueiredo, A. J. Göttle, F. Calle-
Vallejo, M. T. M. Koper, Nat. Energy. 2019, 4, 732-745; b) R. Kortlever,
J. Shen, K. J. Schouten, F. Calle-Vallejo, M. T. M. Koper, J. Phys. Chem.
Lett. 2015, 6, 4073-4082.
[12] a) C. T. Dinh, T. Burdyny, M. G. Kibria, A. Seifitokaldani, C. M. Gabardo,
F. P. G. De Arquer, A. Kiani, J. P. Edwards, P. De Luna, O. S. Bushuyev,
Science 2018, 360, 783-787; b) G. Iijima, T. Inomata, H. Yamaguchi, M.
Ito, H. Masuda, ACS Catalysis 2019, 9, 6305-6319.
[5]
[6]
a) Pérez-Ramírez, N. López, Nat. Catal. 2019, 2, 971-976; b) M. B. Ross,
P. De Luna, Y. Li, C. T. Dinh, D. Kim, P. Yang, E. H. Sargent, Nat. Catal.
2019, 2, 648-658; c) Z. W. Seh, J. Kibsgaard, C. F. Dickens, I. Chorken-
dorff, J. K. Norskov, T. F. Jaramillo, Science 2017, 355; d) Y. Li, Q. Sun,
Adv. Energy. Mater. 2016, 6.
[13] a) P. De Luna, R. Quintero-Bermudez, C. T. Dinh, M. B. Ross, O. S.
Bushuyev, P. Todorović, T. Regier, S. O. Kelley, P. Yang, E. H. Sargent,
Nat. Catal. 2018, 1, 103-110; b) 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; c) Q. Zhu, X. Sun, D. Yang, J. Ma, X. Kang, L.
Zheng, J. Zhang, Z. Wu, B. Han, Nat. Commun. 2019, 10, 3851.
[14] a) D. Wakerley, S. Lamaison, F. Ozanam, N. Menguy, D. Mercier, P.
Marcus, M. Fontecave, V. Mougel, Nat. Mater. 2019, 18, 1222-1227; b)
C. M. Gunathunge, X. Li, J. Li, R. P. Hicks, V. J. Ovalle, M. M. Waegele,
J. Phys. Chem. C. 2017, 121, 12337-12344; c) M. G. Kibria, C. T. Dinh,
A. Seifitokaldani, P. De Luna, T. Burdyny, R. Quintero-Bermudez, M. B.
Ross, O. S. Bushuyev, F. P. Garcia de Arquer, P. Yang, D. Sinton, E. H.
Sargent, Adv. Mater. 2018, 30, 1804867.
a) J. Huang, M. Mensi, E. Oveisi, V. Mantella, R. Buonsanti, J. Am. Chem.
Soc. 2019, 141, 2490-2499; b) Y. C. Li, Z. Wang, T. Yuan, D. H. Nam,
M. Luo, J. Wicks, B. Chen, J. Li, F. Li, F. P. G. de Arquer, Y. Wang, C. T.
Dinh, O. Voznyy, D. Sinton, E. H. Sargent, J. Am. Chem. Soc. 2019, 141,
8584-8591; c) C. G. Morales-Guio, E. R. Cave, S. A. Nitopi, J. T. Feaster,
L. Wang, K. P. Kuhl, A. Jackson, N. C. Johnson, D. N. Abram, T.
Hatsukade, C. Hahn, T. F. Jaramillo, Nat. Catal. 2018, 1, 764-771; d) S.
Ma, M. Sadakiyo, M. Heima, R. Luo, R. T. Haasch, J. I. Gold, M. Yamau-
chi, P. J. Kenis, J. Am. Chem. Soc. 2017, 139, 47-50; e) D. Kim, C. Xie,
N. Becknell, Y. Yu, M. Karamad, K. Chan, E. J. Crumlin, J. K. Norskov,
P. Yang, J. Am. Chem. Soc. 2017, 139, 8329-8336; f) H. Zhang, X.
Chang, J. G. Chen, W. A. Goddard, B. Xu, M. J. Cheng, Q. Lu, Nat. Com-
mun. 2019, 10, 3340.
[15] K. J. Schouten, Z. Qin, E. Perez Gallent, M. T. M. Koper, J. Am. Chem.
Soc. 2012, 134, 9864-9867.
[16] S. Back, M. S. Yeom, Y. Jung, ACS Catal. 2015, 5, 5089-5096.
[17] a) Y. Lum, J. W. Ager, Nature Catalysis 2018, 2, 86-93; b) A. J. Garza,
A. T. Bell, M. Head-Gordon, ACS Catal. 018, 8, 1490-1499; c) T. K. Todo-
rova, M. W. Schreiber, M. Fontecave, ACS Catal. 2019, 10, 1754-1768.
[18] a) C. Chen, Y. Kang, Z. Huo, Z. Zhu, W. Huang, H. L. Xin, J. D. Snyder,
D. Li, J. A. Herron, M. J. S. Mavrikakis, Science 2014, 343, 1339-1343;
b) L. Xiong, Z. Sun, X. Zhang, L. Zhao, P. Huang, X. Chen, H. Jin, H.
Sun, Y. Lian, Z. Deng, M. H. Rummerli, W. Yin, D. Zhang, S. Wang, Y.
Peng, Nat. Commun. 2019, 10, 3782.
[7]
a) H. Wang, Y. K. Tzeng, Y. Ji, Y. Li, J. Li, X. Zheng, A. Yang, Y. Liu, Y.
Gong, L. Cai, Y. Li, X. Zhang, W. Chen, B. Liu, H. Lu, N. A. Melosh, Z. X.
Shen, K. Chan, T. Tan, S. Chu, Y. Cui, Nat. Nanotechnol. 2020, 15, 131-
137; b) F. Li, Y. C. Li, Z. Wang, J. Li, D.-H. Nam, Y. Lum, M. Luo, X.
Wang, A. Ozden, S.-F. Hung, B. Chen, Y. Wang, J. Wicks, Y. Xu, Y. Li,
C. M. Gabardo, C. T. Dinh, Y. Wang, T. T. Zhuang, D. Sinton, E. H. Sar-
gent, Nat. Catal. 2019, 3, 75-82; c) F. Li, A. Thevenon, A. Rosas-Her-
nandez, Z. Wang, Y. Li, C. M. Gabardo, A. Ozden, C. T. Dinh, J. Li, Y.
Wang, J. P. Edwards, Y. Xu, C. McCallum, L. Tao, Z. Q. Liang, M. Luo,
X. Wang, H. Li, C. P. O'Brien, C. S. Tan, D. H. Nam, R. Quintero-Bermu-
dez, T. T. Zhuang, Y. C. Li, Z. Han, R. D. Britt, D. Sinton, T. Agapie, J. C.
Peters, E. H. Sargent, Nature 2020, 577, 509-513; d) F. P. G. de Arquer,
C. T. Dinh, A. Ozden, J. Wicks, C. McCallum, A. R. Kirmani, D. H. Nam,
C. Gabardo, A. Seifitokaldani, X. J. S. Wang, Science 2020, 367, 661-
666.
[19] a) M. McEachran, D. Keogh, B. Pietrobon, N. Cathcart, I. Gourevich, N.
Coombs, V. Kitaev, J. Am. Chem. Soc. 2011, 133, 8066-8069; b) J. Ahn,
D. Wang, Y. Ding, J. Zhang, D. Qin, ACS Nano 2018, 12, 298-307; c) M.
A. Mahmoud, M. A. El-Sayed, J. Am. Chem. Soc. 2010, 132, 12704-
12710.
[20] R. G. Mariano, K. McKelvey, H. S. White, M. W. J. S. Kanan, Science
2017, 358, 1187-1192
[21] a) M. Liu, Y. Pang, B. Zhang, P. De Luna, O. Voznyy, J. Xu, X. Zheng,
C. T. Dinh, F. Fan, C. J. N. Cao, Nature 2016, 537, 382-386; b) Y. Chen,
C. W. Li, M. W. Kanan, J. Am. Chem. Soc. 2012, 134, 19969-19972.
[22] Y. Xia, K. D. Gilroy, H. C. Peng, X. Xia, Angew. Chem. Int. Ed. 2017, 56,
60-95.
[8]
a) Y. Wang, Z. Wang, C.-T. Dinh, J. Li, A. Ozden, M. Golam Kibria, A.
Seifitokaldani, C. S. Tan, C. M. Gabardo, M. Luo, H. Zhou, F. Li, Y. Lum,
C. McCallum, Y. Xu, M. Liu, A. Proppe, A. Johnston, P. Todorovic, T. T.
Zhuang, D. Sinton, S. O. Kelley, E. H. Sargent, Nat. Catal. 2019, 3, 98-
106; b) F. S. Roberts, K. P. Kuhl, A. Nilsson, Angew. Chem. Int. Ed. 2015,
54, 5179-5182; c) R. B. Sandberg, J. H. Montoya, K. Chan, J. K. Nørskov,
[23] M. J. Yacamán, J. Ascencio, H. Liu. Gardea-Torresdey, Sci. Technol. B.
2001, 19, 1091-1103.
9
This article is protected by copyright. All rights reserved.

10.1002/anie.202012631
Angewandte Chemie International Edition
RESEARCH ARTICLE
[24] J. H. Kim, H. Woo, S. W. Yun, H.W. Jung, S, Back, Y, Jung, Y. T. Kim,
2017, Appl. Catal. B-Environ, 2017, 213, 211-215.
[25] a) M. Kuhn, T. K. Sham, Phys. Rev. B. 1994, 49, 1647-1661; b) T. Sham,
A. Hiraya, M. Watanabe, Phys. Rev. B. 1997, 55, 7585.
[26] S. Zhu, B. Jiang, W. B. Cai, M. Shao. J. Am. Chem. Soc. 2017, 139,
15664-15667.
[27] A. Wuttig, C. Liu, Q. L, M. Yaguchi, C. H. Hendon, K. Motobayashi, M.
Osawa, Y, Surendranath, ACS Cent. Sci. 2016, 2, 522-528.
[28] E. Perez-Gallent, M. C. Figueiredo, F. Calle-Vallejo, M. T. Koper, Angew.
Chem. Int. Ed. 2017, 56, 3621-3624.
[29] T. Cheng, H. Xiao, W. A. Goddard, P. Natl. Acad. Sci. 2017, 114, 1795-
1800.
10
This article is protected by copyright. All rights reserved.

10.1002/anie.202012631
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
We prepared the Cu-Ag/Au nanoframes to break the intrinsic scaling relationship of intermediates binding associated CO2RR, and
achieving extraordinary Faradic efficiencies of 69 ± 5% and 77 ± 2% on ethylene production in H-cell and flow cell, respectively. Both
the *CO*CO coupling and *CO→*CHO hydrogenation steps are rate-determining and mechanistic-gating towards ethylene production
11
This article is protected by copyright. All rights reserved.