Electroreduction of CO2 on Single-Site Copper-Nitrogen-Doped Carbon Material: Selective Formation of Ethanol and Reversible Restructuration of the Metal Sites

Article abstract of DOI:10.1002/anie.201907994

It is generally believed that CO₂ electroreduction to multi-carbon products such as ethanol or ethylene may be catalyzed with significant yield only on metallic copper surfaces, implying large ensembles of copper atoms. Here, we report on an inexpensive Cu-N-C material prepared via a simple pyrolytic route that exclusively feature single copper atoms with a CuN₄ coordination environment, atomically dispersed in a nitrogen-doped conductive carbon matrix. This material achieves aqueous CO₂ electroreduction to ethanol at a Faradaic yield of 55 % under optimized conditions (electrolyte: 0.1 m CsHCO₃, potential: ?1.2 V vs. RHE and gas-phase recycling set up), as well as CO electroreduction to C₂-products (ethanol and ethylene) with a Faradaic yield of 80 %. During electrolysis the isolated sites transiently convert into metallic copper nanoparticles, as shown by operando XAS analysis, which are likely to be the catalytically active species. Remarkably, this process is reversible and the initial material is recovered intact after electrolysis.

Full text of DOI:10.1002/anie.201907994

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Electroreduction of CO2 on Single-Site Copper-Nitrogen-Doped
Carbon Material: Selective Formation of Ethanol and Reversible
Restructuration of the Metal Sites
Authors: Marc Fontecave, Dilan Karapinar, Ngoc Huan Tran, Nastaran
Ranjbar Sahraie, Jingkun Li, Nadia Touati, Sandrine Zanna,
Dario Taverna, Luiz Henrique Galvão Tizei, Andrea Zitolo,
Frederic Jaouen, Victor Mougel, and David Wakerley
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.201907994
Angew. Chem. 10.1002/ange.201907994
Link to VoR: http://dx.doi.org/10.1002/anie.201907994
http://dx.doi.org/10.1002/ange.201907994

10.1002/anie.201907994
Angewandte Chemie International Edition
RESEARCH ARTICLE
Electroreduction of CO2 on Single-Site Copper-Nitrogen-Doped
Carbon Material: Selective Formation of Ethanol and Reversible
Restructuration of the Metal Sites
Dilan Karapinar,[a] Ngoc Tran Huan, [a] Nastaran Ranjbar Sahraie,[b] Jingkun Li,[b] David Wakerley[a],
Nadia Touati,[c] Sandrine Zanna,[c] Dario Taverna,[d] Luiz Henrique Galvão Tizei,[e] Andrea Zitolo,[f]
Frédéric Jaouen,[b] Victor Mougel,[a],[g]* Marc Fontecave[a]*
Abstract: It is generally believed that CO2 electroreduction to multi-
carbon products such as ethanol or ethylene may be catalyzed with
significant yield only on metallic copper surfaces, implying large
ensembles of copper atoms. Here, we report on an inexpensive Cu-
N-C material prepared via a simple pyrolytic route that exclusively
feature single copper atoms with a CuN4 coordination environment,
atomically dispersed in a nitrogen-doped conductive carbon matrix.
This material achieves aqueous CO2 electroreduction to ethanol at a
Faradaic yield of 55% under optimized conditions (electrolyte: 0.1 M
CsHCO3 , potential: -1.2V vs. RHE and gas-phase recycling set up),
as well as CO electroreduction to C2-products (ethanol and ethylene)
with a Faradaic yield of 80%. During electrolysis the isolated sites
transiently convert into metallic copper nanoparticles, as shown by
operando XAS analysis, which are likely to be the catalytically active
species. Remarkably, this process is reversible and the initial material
is recovered intact after electrolysis.
chemical feedstocks. This may be achieved in electrolyzers via
the electrochemical CO2 reduction reaction (CO2RR), providing
the opportunity to store renewable but intermittent electric energy
(sun, wind) as chemical fuels or other high-added-value products.
Copper-based materials were early identified as catalysts that
efficiently promote CO2RR in aqueous conditions to C2-products
such as ethylene and ethanol [1]. Recent years have witnessed
numerous studies on such electrode materials, reaching Faradaic
yields (FY) for multi-carbon products as high as 60% [2]. Alcohols
are desirable CO2RR products, in particular ethanol, owing to their
high energy-densities and ease of separation from the unreacted
CO2 gas stream. However, catalysts promoting alcohol production
with high selectivity are still extremely rare.[2e-k, 3] Strategies to
increase the selectivity of metallic-copper surfaces towards multi-
carbon products are emerging through tailoring the surface
properties of copper electrodes, either using a second metal [2k, 4]
,
by controlled oxidation [2g, 5] or nanostructuration [6]. Alternatively,
single-site catalysts have shown high CO2RR activity and
selectivity, however only afford single-carbon compounds, such
as CO, methane or methanol as major products.[7]
Introduction
The conversion of CO2 to multi-carbon products is currently one
of the most promising approaches to valorize CO2 as fuels or
Herein, we disclose the synthesis, characterisation and CO2RR
activity of a copper-nitrogen co-doped carbon material (Cu-N-C)
presenting well-defined isolated CuN4 sites covalently integrated
into an amorphous carbon matrix (Fig. 1a). This material shows
high selectivity for ethanol production in aqueous conditions, with
a FY up to 55%. We demonstrate that ethanol is the sole liquid-
phase product of CO2RR and that CO is a key reaction
intermediate in this process, as evidenced by the ability of the
material to catalyse the electroreduction of CO to ethanol with a
FY over 65%. In addition, while ex-situ characterization of the
material after electrolysis shows a structure identical to the
starting material with only isolated CuN4 sites, operando X-ray
absorption (XAS) results clearly evidenced a restructuring of the
material under electrocatalytic conditions, via partial and transient
conversion of CuN4 sites into Cu nanoparticles. Indeed, this is
reversible since the latter convert back to CuN4 sites after
electrolysis.
[a]
D. Karapinar, Dr. N. T. Huan, Dr D. Wakerley, Prof. Dr. V. Mougel,
Prof. Dr. M. Fontecave
Laboratoire de Chimie des Processus Biologiques, CNRS UMR
8229
Collège de France, Sorbonne Université, PSL Research University,
11 Place Marcelin Berthelot, 75231 Paris Cedex 05, France.
Correspondence to: mougel@inorg.chem.ethz.ch;
marc.fontecave@college-de-france.fr
[b]
[c]
[d]
Dr. N. Ranjbar Sahraie, Dr. J. Li, Dr. F. Jaouen
Institut Charles Gerhardt Montpellier (UMR 5253), CNRS, Université
de Montpellier
ENSCM, 2 place Eugène Bataillon, 34095 Montpellier, France.
Dr. N Touati, Dr. S. Zanna
Chimie ParisTech, PSL Research University, CNRS, Institut de
Recherche de Chimie Paris (IRCP)
11 rue Pierre et Marie Curie, 75005 Paris, France.
Dr. D. Taverna
Institut de Minéralogie et de Physique des Milieux Condensés, UMR
7590 CNRS, Sorbonne Universités, UPMC Univ Paris 06
4 place Jussieu, 75005 Paris, France.
Results and Discussion
[e]
[f]
Dr. L. H. Galvão Tizei
Laboratoire de Physique des Solides
Université Paris Sud, Bât 510, 91405 Orsay, France.
Dr. A. Zitolo
Synthesis and characterization of the Cu-N-C single-site
material
Synchrotron SOLEIL
L'Orme des Merisiers Saint-Aubin - BP 48, 91192 Gif-sur-Yvette,
France.
Current address: Department of Chemistry and Applied Biosciences,
ETH Zürich, Vladimir-Prelog-Weg 1-5, CH-8093 Zürich, Switzerland.
Pyrolytic synthesis has been identified as an effective route to
produce atomically-dispersed metal ions coordinated to nitrogen
atoms within carbon matrices [8]. By analogy to the synthetic
routes developed for Fe-N-C materials used in O2
electroreduction and CO2RR,[7e, 8a] we successfully synthesized a
[g]
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.201907994
Angewandte Chemie International Edition
RESEARCH ARTICLE
single-site Cu-N-C material via a two-step strategy: (i) dry-phase
mixing of a Zn based zeolitic imidazolate framework (ZIF-8) with
Cu(II) chloride and phenanthroline ligand via low-energy ball-
milling; (ii) pyrolysis of the resulting catalyst precursor powder at
1050 °C under Ar flow. We prepared the material Cu0.5NC,
comprising 0.5%wt Cu at step (i), which resulted in 1.4%wt Cu after
pyrolysis as determined by elemental analysis (the increase in
Cu %wt is due to ZIF-8 forming volatile products). A very small
amount (< 0.02%wt) of Zn, derived from ZIF-8, was also present
as determined by Inductively Coupled Plasma – Atomic Emission
Spectroscopy (ICP-AES). X-ray photoelectron spectroscopy
(XPS) of Cu0.5NC showed Cu 2p3/2 peaks at 932.5 and 935.0 eV
and a Cu-Auger line with a maximum at 569.0 eV (Figure S1),
indicating +II is the predominant Cu oxidation state. The analysis
of the N 1s region revealed the presence of pyridinic (398.6 eV),
pyrrolic (401.0 eV), graphitic (402.6 eV) and porphyrin-like N-Cu
signals (399.7 eV), analogous to that which is typically observed
with Metal-N-C materials (Figure S1) [7e, 7f]. The presence of Zn
was also confirmed by XPS analysis (Figure S1). At 10 K Cu0.5NC
displayed a continuous-wave electron paramagnetic resonance
(EPR) spectrum with a broad signal centered at g = 2.07, typical
for a Cu2+, d9 S=1/2 state (Figure 1b). Hyperfine coupling was not
observed, probably due to the exchange interaction within the
graphitic sheets. The threshold energy of the Cu K-edge X-ray
absorption near-edge structure (XANES) spectrum of Cu0.5NC is
8983.8 eV, similar to that of a Cu(II)-phthalocyanine molecular
analogue (8984.6 eV). This further supports a predominant +II
oxidation state of the Cu centers (Figure S2-S4). The Cu K-edge
extended X-ray absorption fine structure (EXAFS) analysis of
Cu0.5NC is shown in Figure 1c and Figure S5. Fitting this spectrum
with Cu-N and Cu-C scattering paths from the nearest neighbors
reveals the presence of four N atoms at 1.95(1) Å from the Cu
center, in agreement with values determined by EXAFS on
copper-porphyrin [9]. The CuN4 sites are integrated in the graphitic
matrix, as revealed by the need for two shells of Cu-C scattering
path at 3.10(2) and 3.36(2) Å (See Table S1) to properly fit the
EXAFS spectrum. In addition, the absence of intense peaks at ca.
Figure 1. Structural and morphological characterization of Cu0.5NC. (a)
Schematic representation; (b) EPR spectrum recorded at 10 K of Cu0.5NC
diluted in an amorphous silica matrix; (c) Cu K-edge EXAFS analysis in the
Fourier-transformed space; (d) HAADF-STEM image.
CO2 electrocatalytic reduction
Cu0.5NC was then evaluated for electrocatalytic CO2 reduction
after deposition on a carbon-based gas-diffusion layer (GDL) with
a Nafion® perfluorinated resin binder. The latter served to stabilize
the deposit but had no effect on the activity. The deposit did not
significantly alter the morphology of the GDL layer, as shown by
scanning electron microscopy (Figure S10). The electrode was
initially investigated by controlled-potential electrolysis (CPE) at
potentials ranging from –0.5 to –1.0 V (all potentials are indicated
vs. RHE) under standard ‘static headspace’ conditions, i.e. in a
closed cell, under a CO2 atmosphere, containing CO2-saturated
0.1 M aqueous solutions of various alkali bicarbonates (Li, Na, K,
Cs). Linear sweep voltammograms (LSVs), CO production rates
and current variation during electrolysis are reported in Figure
S11-S16. As shown in Figure 2a and 2b, CO and H2 were
identified as the major reaction products, with the FY for CO,
FY(CO), reaching a maximum at different potentials depending on
the electrolyte (47 % at –0.8 V for LiHCO3, 58 % at –0.7 V for
NaHCO3, 64 % at –0.6 V for KHCO3 and 74 % at –0.6 V for
CsHCO3). The maximum FY(CO) increased with increasing the
size of the cation in the order Li+<Na+<K+<Cs+. While this had
been previously observed with metallic electrodes (Au, Ag, Cu,
Hg),[10] it had not yet been reported for Metal-N4-based or other
carbon-supported catalysts. FY(H2) increased at the expense of
FY(CO) upon scanning towards more cathodic potentials. This
can be explained by considering the mass transport of the
reactants. Indeed, the CO2 concentration at the electrode surface
decreases drastically with increasing current densities due to CO2
consumption and stronger gas evolution from the electrode
(which sparges the solution from CO2) and as such protons
become the dominant substrate due to their higher concentration
in the bulk and higher diffusion coefficient. In this respect, the
CsHCO3 electrolyte is clearly unique since H2 production did not
significantly increase below –0.7 V. Such an effect has been
attributed to the stronger cation hydration of Cs+, reducing the
cation-specific adsorption on the cathode and limiting H2
2.3
Å in the Fourier-transformed spectrum (Figure S6),
characteristic for Cu-Cu scattering paths, suggests the absence
of copper clusters. The absence of crystalline copper in our
material was further confirmed by (i), powder X-ray diffraction of
the material (Figure S7), revealing only two broad reflection peaks
typical for nano-sized graphitic platelets in amorphous carbon
[7e]
materials
and (ii), transmission electron microscopy (TEM)
images showing amorphous carbon structures and no copper
clusters (Figure S8). To gain deeper insight into the copper
dispersion, atomic-resolution high-angle annular dark-field
scanning transmission electron microscopy (HAADF-STEM)
images of Cu0.5NC were captured. These images (Figure 1d and
Figure S9) show the presence of multiple bright dots of atomic
size, corresponding to isolated Cu atoms, as confirmed by
electron energy loss spectroscopy at these sites (Figure S9b).
This article is protected by copyright. All rights reserved.

10.1002/anie.201907994
Angewandte Chemie International Edition
RESEARCH ARTICLE
production.[11] Additionally we observed a total FY for all gas-
phase products well below 100% at electrode potentials below –
0.9 V, indicating the formation of liquid phase products, in
particular trace amounts of ethanol (Figure S12b). The very
limited amount of CO2 in CO2 -saturated aqueous solution under
A maximum FY(EtOH) of 43 % was found for a flow rate of
2.5 mL/min with a high and stable average current density of 16.2
mA/cm2 (Figure 2c, Figure S20 and Table S2). At this optimized
flow rate, the FY(EtOH) increased at all the potentials between –
1.0 V and –1.3 V with respect to the catalytic test carried out at 10
mL/min flow rate, but the optimal potential remained at –1.2 V
(Figure 2d). It should be noted that the FY for CO and H2 did not
vary during the electrolysis (Figure S18). As a reference material,
the parent copper-free nitrogen-doped carbon material prepared
identically, except for the absence of copper chloride in the
precursor mixture, did not show any ethanol production (Figure
S21). This demonstrated that the Zn atoms are not active in
ethanol production. Whether such a low amount of Zn present on
the material could nevertheless combine with Cu to provide the
observed selectivity seems unlikely. Indeed, recent ZnCu alloys
were reported to catalyze the electroreduction of CO2, however
exclusively into a mixture of CO and H2 with no ethanol
production[13] and CuxZn oxide catalysts were shown to produce
ethanol, in addition to CO and H2, however with an important co-
production of ethylene.[2e]
static CO2 atmosphere together with
a degassing of the
electrolyte due to the formation of gaseous products at such
negative potentials however greatly restrict the formation of more
complex multi-carbon products . This led us to modify the
operating conditions in a second step of the investigation. Indeed,
to prevent CO2 depletion during electrolysis and reach higher FY
for such multi-carbon products, we thus carried out the catalytic
tests under ‘flow’ conditions, providing a constant CO2 supply to
the electrolyte (10 mL/min), in 0.1 M CsHCO3 as the electrolyte.
To confirm that CO2 constituted the carbon source for ethanol, we
carried out the catalytic test using 13C labeled CO2 (–1.2 V,
1
2.5 mL/min CO2). The H and 13C NMR of the liquid phase after
1h electrolysis revealed the quantitative 13C labeling of ethanol
(Figure S22-S23), validating that ethanol originates from CO2
reduction. Finally, under optimized conditions (–1.2 V, 2.5 mL/min
CO2, Figure 2e) the larger cations again favored ethanol
production with FY(EtOH) increasing in the order Li+ (2%) < Na+
(5%) < K+(16%) < Cs+ (43%).
As for many CO2 reduction catalysts, the Tafel plot for Cu0.5NC is
complex and the Tafel slope is strongly potential-dependent. We
hence expressed the Tafel slopes resulting from the specific
partial current densities of the different products. A Tafel slope of
175 mV/decade could be measured between –0.4 and –0.6 V
applied potential for CO, as derived from “static headspace”
experiments and a Tafel slope of 311 mV/decade for EtOH was
estimated from “flow” experiments between –1.1 and –1.2 V
applied potential (Figure S24). These different values suggest
that the two reactions have different rate determining steps. In
addition, the Tafel slope measured for ethanol is significantly
higher than the slope of 120 mV/decade typically observed when
the rate-determining step is a reductive coupling of adsorbed CO
intermediates on metallic electrodes.[6d] Relatively large
overpotential at 1 mA/cm2 of 1 V was determined by Tafel-slope
analysis of the partial current for ethanol. This value is in the range
of the overpotentials observed with Cu foil in similar reaction
conditions[2a] but our catalyst shows a much narrower product
distribution.
Figure 2. Faradaic yields of CO2RR on Cu0.5NC. Faradaic yields of (a) CO
and (b) H2 under ‘static headspace’ conditions after 10 minutes of CPE, in 0.1
M CO2-saturated aqueous solutions of various alkali bicarbonates. FYs for
CO2RR under ‘flow’ conditions after 1 h of CPE: (c) at –1.2 V vs. RHE in a 0.1 M
CsHCO3 aqueous solution under various flow rates of CO2. (d) at 2.5 mL/min
CO2 flow-rate in 0.1 M CsHCO3 aqueous solution, at various applied potentials
during CPE. (e) at –1.2 V vs. RHE and 2.5 mL/min flow rate of CO2 in 0.1 M
aqueous solutions using various electrolyte cations. FYs of CO reduction
products after 1 h CPE at –1.2 V vs. RHE under 2.5 mL/min flow of CO in 0.1 M
CsHCO3 are reported on the right-hand side of Figure 2 (e).
CPE was carried out during one hour at potentials ranging from –1.0 V
to –1.3 V, analyzing the gaseous products by on-line gas
chromatography and the liquid phase products by quantitative 1H
NMR and ionic chromatography. Remarkably, ethanol was the sole
product found in liquid phase at these potentials, while H2 and CO
were the only detected gas-phase products (Figure S17-S18).[12] The
highest FY(EtOH) was obtained at –1.2 V (Figure S19). We further
optimized the system by investigating the effect of CO2 flow-rate on
the ethanol production, while keeping the potential fixed at –1.2 V.
Assuming that CO is the most probable first reaction intermediate
in the ethanol formation mechanism, we reasoned that the lower
selectivity of Cu0.5NC for ethanol production at high CO2 flow-
rates could be explained by a decrease of local CO concentration
at the electrode surface caused by the increased sparging. We
hence carried out CPE using Cu0.5NC in 0.1 M CsHCO3 aqueous
solution at –1.2 V for 1 hour with a continuous 2.5 mL/min flow of
CO (Figure S25). A stable current density of about 12 mA/cm2
This article is protected by copyright. All rights reserved.

10.1002/anie.201907994
Angewandte Chemie International Edition
RESEARCH ARTICLE
was observed and a FY(EtOH) of 66% was obtained , confirming
that higher FY(EtOH) could be obtained if the CO concentration
was increased (Figure 2e). Minor amounts of ethylene and
methane were also observed.
in Figure S33 and Table S3, respectively, and compared to the
crystalline Cu results. The estimated size of the Cu nanoparticles,
by means of Cu-Cu first shell parameters, is 0.47±0.04 nm. While
the coordination numbers obtained by the EXAFS analysis are
probably underestimated, because a very small fraction of Cu is
coordinated with light atoms, the experimental EXAFS spectrum
can be correctly reproduced with Cu-Cu contributions only. The
correction to be made to the CNs is small and affecting the particle
size determination only marginally (e.g. for a system containing
20% of non-metallic Cu, which is a non-negligible fraction, the
particle size estimation would be 0.53±0.05 nm).
This high FY for CO reduction to ethanol encouraged us to
explore the recycling of the gas phase during CO2RR, as
significant amounts of CO were formed and could be further
converted to ethanol. We hence carried out CPE at –1.2 V for 1
hour using a closed CO2 volume of 300 mL cycled through the
electrolyzer at a flow of 2.5 mL/min (Figure S26). In this setup the
FY(EtOH) increased to 55%, with a concomitant decrease of FY
for CO, in agreement with the favored reduction of CO to ethanol.
In these conditions, methane was also identified as a minor
product (Figure S27).
Intrigued by the exclusive presence of CuN4 sites in the ex-situ
post electrolysis samples, we exposed the material used for the
operando XAS measurements to air and, using XAS
characterization, observed the disappearance of the metallic
copper phase and the restoration of the original spectra (Figure 3
c-d). The same behavior occurred when the potential of the
electrode was switched back to + 1 V vs. RHE, after electrolysis
at –1.2 vs. RHE. This full structural reversibility operates upon
multiple reduction/oxidation cycles. ICP measurement of Cu in the
electrolyte upon cycling confirmed the absence of dissolved Cu,
consistent with a quantitative binding of Cu in CuN4 sites upon
application of positive potentials. This is likely the consequence of
the small size of the particles and the strong Cu(II)-chelating
capacity of the N4 sites of the material.
Ex-situ and operando characterization of the material
Ex-situ characterization of Cu0.5NC after electrolysis did not show
any evidence for the presence of metallic copper nanoparticles:
(i) no Cu nanoparticles could be observed in TEM images (Figure
S28); (ii) no Cu0 signal was detected in the XPS spectrum (Figure
S29) and the XANES spectrum (Figure S30) either; (iii) no
characteristic Cu-Cu feature from metallic copper or copper
oxides could be observed in the Fourier Transform (FT) of the
EXAFS spectrum which revealed that the material only contained
CuN4 sites before and after electrolysis (Figure S31). However,
the ex-situ observation of isolated CuN4 moieties may not be
sufficient to conclude that these are the catalytic sites per se. In
particular, recent studies of molecular Cu-phtalocyanine catalysts
showed that very small metallic Cu nanoparticles formed in-situ
during CO2 electrolysis, however in a reversible way [14]. Whether
such a behavior also occurs in the case of the CuN4 sites of
Cu0.5NC was studied by XAS in operando electrolysis conditions
(Figure S32). Analysis of the XANES spectra at the copper K-
edge of the material under electrocatalytic conditions revealed a
shift in the edge position from 8983.93eV to 8979.2 eV for applied
potentials below –0.6 V vs. RHE, indicative of a change of the Cu
oxidation state from +II to 0 (the metallic copper K-edge energy
position is indeed 8979 eV, see Figures 3, S3-S4). Consistently,
Cu K-edge FT-EXAFS spectra of Cu0.5NC recorded under
operating conditions demonstrated the appearance of a Cu-Cu
coordination, showing that metallic copper nanoparticles
extensively form under electrocatalytic conditions, even when
moderate potentials (from –0.6 V vs. RHE) were applied.
In agreement with XANES data, the FT-EXAFS spectrum thus
clearly indicated that Cu was essentially in the form of
nanoparticles after prolonged electrolysis at –1.2 V vs. RHE
(green spectrum Figure 3d), even though a very small peak at
around 1.5 Å was indicative of Cu atoms coordinated with light
atoms (C, N, O). It is thus very likely that these nanoparticles are
the catalytically active species. Nitrogen-doped carbon-supported
Cu nanoparticles have been previously reported as highly active
for the CO2RR to ethanol.[15]
Figure 3. Operando XAS characterization of Cu0.5NC at the Cu-K edge. (a)
K-edge XANES spectra of Cu0.5NC under no potential applied (blue
line) ,Cu0.5NC during electrolysis at –0.6 V vs. RHE ( pink line) , at –0.7 V vs.
RHE ( green line) , at –1.2 V vs. RHE ( red line) and metallic copper (black line),
(b) Fourier transform of the experimental EXAFS spectra of Cu0.5NC no potential
applied (blue line) ,Cu0.5NC during electrolysis at –0.6 V vs. RHE ( pink line) , at
–0.7 V vs. RHE ( green line) , at –1.2 V vs. RHE ( red line) and metallic copper
(black line), (c) Comparison between the K-edge XANES experimental
spectrum of Cu0.5NC under no potential applied (blue line), Cu0.5NC during
electrolysis at –1.2 V vs. RHE ( red line), after electrolysis under no potential
applied ( green line) and Cu0.5NC after electrolysis at–1.2 V vs. RHE then
sample exposed to air (orange line) (d) Fourier transform of the experimental
EXAFS spectra of Cu0.5NC under no potential applied (blue line), Cu0.5NC during
electrolysis at –1.2 V vs. RHE ( red line), after electrolysis under no potential
applied ( green line) and Cu0.5NC after electrolysis at –1.2 V vs. RHE then
sample exposed to air for 10 hours (orange line).
The mean size of these particles can be estimated by using
EXAFS coordination numbers (CNs) and mean interatomic
distances applying an equation valid for homogeneous spherical
nanoclusters.[16] The details of the fitting procedure are provided
in the experimental section. The best-fit analysis and the
structural parameters obtained for Cu0.5NC at –1.2 V are reported
This article is protected by copyright. All rights reserved.

10.1002/anie.201907994
Angewandte Chemie International Edition
RESEARCH ARTICLE
127, 5268-5271; d) 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. R. Cuenya, Nature Communications 2016, 7,
12123; e) D. Ren, B. S.-H. Ang, B. S. Yeo, ACS Catalysis 2016, 6,
8239-8247; f) Y. Song, R. Peng, D. K. Hensley, P. V. Bonnesen, L.
Liang, Z. Wu, H. M. Meyer, M. Chi, C. Ma, B. G. Sumpter, A. J.
Rondinone, ChemistrySelect 2016, 1, 6055-6061; g) M. Rahaman,
A. Dutta, A. Zanetti, P. Broekmann, ACS Catalysis 2017, 7, 7946-
7956; h) K. Zhao, Y. Liu, X. Quan, S. Chen, H. Yu, ACS Applied
Materials & Interfaces 2017, 9, 5302-5311; i) A. D. Handoko, K. W.
Chan, B. S. Yeo, ACS Energy Lett. 2017, 2, 2103-2109; j) R. A.
Geioushy, M. M. Khaled, K. Alhooshani, A. S. Hakeem, A. Rinaldi,
Electrochim. Acta 2017, 245, 456-462; k) E. L. Clark, C. Hahn, T. F.
Jaramillo, A. T. Bell, Journal of the American Chemical Society 2017,
139, 15848-15857.
Conclusion
The novel material reported herein demonstrated the amenability
of isolated copper CuN4 sites as precursors for highly active
catalysts for CO2 electroreduction to ethanol. It is unique in
allowing formation of ethanol as the only liquid product, reaching
Faradaic yields as high as 55%. Optimization of the
electrocatalytic conditions illustrated the importance of the
electrolyte cation and the CO2 flow-rate on the catalytic selectivity
for EtOH formation. As shown in Table S4, these Faradic yields
for ethanol are among the highest reported ones using Cu-based
electrocatalysts. Operando XAS experiments established the
conversion of the CuN4 sites into very small metallic Cu
nanoparticles during electrolysis. The latter are likely to be the
catalytically active species, displaying a unique reactivity. Further
studies of this material are required to understand the high
selectivity observed. Interestingly, in the present case, the
restructuring behavior is reversible which makes the material also
a very stable catalyst. This study thus illustrates that extreme care
must be taken before concluding on the nature of the active sites
in such single-atom catalysts, even at very low metal loading. By
demonstrating that the generation of multi-carbon products with
high selectivity is possible using such site-isolated inexpensive
material, this work paves the way towards new approaches to
CO2 electroreduction using single-site material precursors for
improved selectivity and efficiency.
[3]
[4]
Y. Song, W. Chen, C. Zhao, S. Li, W. Wei, Y. Sun, Angew. Chem.
Int. Ed. 2017, 56, 10840-10844.
Y. Wang, D. Wang, C. J. Dares, S. L. Marquard, M. V. Sheridan, T.
J. Meyer, Proceedings of the National Academy of Sciences 2018,
115, 278.
[5]
a) K. W. Frese, J. Electrochem. Soc. 1991, 138, 3338-3344; b) M.
Le, M. Ren, Z. Zhang, P. T. Sprunger, R. L. Kurtz, J. C. Flake,
Journal of The Electrochemical Society 2011, 158, E45-E49; c) C.
W. Li, J. Ciston, M. W. Kanan, Nature 2014, 508, 504; d) A. Dutta,
M. Rahaman, N. C. Luedi, M. Mohos, P. Broekmann, ACS Catalysis
2016, 6, 3804-3814.
Acknowledgements
V.M. acknowledges financial support from CNRS-Cellule Energie
and Fondation of College de France for the acquisition of the flow
electrochemical system. D.K. acknowledges financial support
from Fondation de l’Orangerie. D. W. W. was supported by an
Idex PSL grant (ANR-10-IDEX-001-02 PSL★), the Marie Curie
PRESTIGE Fellowship programme and the Fondation du Collège
de France. F. J. acknowledges funding from the European
Union’s Horizon 2020 research and innovation program under
grant agreement No 732840, A-LEAF. We acknowledge
Synchrotron SOLEIL (Gif-sur Yvette, France) for provision of
synchrotron radiation facilities at beamline SAMBA (proposal
number 99190060). The authors are also grateful to the French
national EPR facilities network RENARD (IR CNRS 3443). We
thank Dr Gwenaelle Rousse for powder X-ray diffraction analyses
and Dr Laurent Binet for interpretation of EPR data.
[6]
a) W. Tang, A. A. Peterson, A. S. Varela, Z. P. Jovanov, L. Bech, W.
J. Durand, S. Dahl, J. K. Norskov, I. Chorkendorff, Phys. Chem.
Chem. Phys. 2012, 14, 76-81; b,9 D. Ren, N. T. Wong, A. D.
Handoko, Y. Huang, B. S. Yeo, The Journal of Physical Chemistry
Letters 2016, 7, 20-24; c) C. Reller, R. Krause, E. Volkova, B.
Schmid, S. Neubauer, A. Rucki, M. Schuster, G. Schmid, Advanced
Energy Materials 2017, 7, 1602114; d) D. Kim, C. S. Kley, Y. Li, P.
Yang, Proc. Nat. Acad. Sci. 2017, 114, 10560-10565.
[7]
a) A. S. Varela, N. Ranjbar Sahraie, J. Steinberg, W. Ju, H.-S. Oh,
P. Strasser, Angewandte Chemie International Edition 2015, 54,
10758-10762; b) P. Kumar, H. P. Mungse, O. P. Khatri, S. L. Jain,
RSC Advances 2015, 5, 54929-54935; c) Z. Weng, J. Jiang, Y. Wu,
Z. Wu, X. Guo, K. L. Materna, W. Liu, V. S. Batista, G. W. Brudvig,
H. Wang, Journal of the American Chemical Society 2016, 138,
8076-8079; d) S. Kusama, T. Saito, H. Hashiba, A. Sakai, S.
Yotsuhashi, ACS Catalysis 2017, 7, 8382-8385; e) T. N. Huan, N.
Ranjbar, G. Rousse, M. Sougrati, A. Zitolo, V. Mougel, F. Jaouen,
M. Fontecave, ACS Catalysis 2017, 7, 1520-1525; f) W. Ju, A.
Bagger, G.-P. Hao, A. S. Varela, I. Sinev, V. Bon, B. Roldan Cuenya,
S. Kaskel, J. Rossmeisl, P. Strasser, Nature Communications 2017,
8, 944; g) M. Gong, Z. Cao, W. Liu, E. M. Nichols, P. T. Smith, J. S.
Derrick, Y.-S. Liu, J. Liu, X. Wen, C. J. Chang, ACS Central Science
2017, 3, 1032-1040; h) B. Reuillard, K. H. Ly, T. E. Rosser, M. F.
Kuehnel, I. Zebger, E. Reisner, Journal of the American Chemical
Keywords: CO2 electroreduction • copper • single-site catalyst •
ethanol • operando analysis
[1]
Y. Hori, A. Murata, R. Takahashi, Journal of the Chemical Society,
Faraday Transactions 1: Physical Chemistry in Condensed Phases
1989, 85, 2309-2326.
[2]
a) K. P. Kuhl, E. R. Cave, D. N. Abram, T. F. Jaramillo, Energy &
Environmental Science 2012, 5, 7050-7059; b) C. W. Li, M. W.
Kanan, Journal of the American Chemical Society 2012, 134, 7231-
7234; c) F. S. Roberts, K. P. Kuhl, A. Nilsson, Angew. Chem. 2015,
This article is protected by copyright. All rights reserved.

10.1002/anie.201907994
Angewandte Chemie International Edition
RESEARCH ARTICLE
Society 2017, 139, 14425-14435; i) A. S. Varela, W. Ju, P. Strasser,
Advanced Energy Materials 2018, 8, 1703614.
analyzed and shown to contain significant amounts of O2, CO and
methane. Poisoning of some surface sites or undesired side-
reaction with these impurities potentially explains the presence of
this trace acetate by-product in the liquid phase.
[8]
a) A. Zitolo, V. Goellner, V. Armel, M.-T. Sougrati, T. Mineva, L.
Stievano, E. Fonda, F. Jaouen, Nat. Mater. 2015, 14, 937-942; b) H.
T. Chung, D. A. Cullen, D. Higgins, B. T. Sneed, E. F. Holby, K. L.
More, P. Zelenay, Science 2017, 357, 479.
[13]
a) P. Moreno-García, N. Schlegel, A. Zanetti, A. Cedeño López, M.
d. J. Gálvez-Vázquez, A. Dutta, M. Rahaman, P. Broekmann, ACS
Applied Materials
& Interfaces 2018, 10, 31355-31365; b) S.
[9]
F. Carrera, E. S. Marcos, P. J. Merkling, J. Chaboy, A. Muñoz-Páez,
Lamaison, D. Wakerley, D. Montero, G. Rousse, D. Taverna, D.
Giaume, D. Mercier, J. Blanchard, H. N. Tran, M. Fontecave, V.
Mougel, ChemSusChem 2019, 12, 511-517.
Inorg. Chem. 2004, 43, 6674-6683.
[10]
a) M. Akira, H. Yoshio, Bull. Chem. Soc. Jpn. 1991, 64, 123-127; b)
M. R. Singh, Y. Kwon, Y. Lum, J. W. Ager, A. T. Bell, Journal of the
American Chemical Society 2016, 138, 13006-13012; c) J. Resasco,
L. D. Chen, E. Clark, C. Tsai, C. Hahn, T. F. Jaramillo, K. Chan, A.
T. Bell, Journal of the American Chemical Society 2017, 139, 11277-
11287.
[14]
[15]
[16]
Z. Weng, Y. Wu, M. Wang, J. Jiang, K. Yang, S. Huo, X.-F. Wang,
Q. Ma, G. W. Brudvig, V. S. Batista, Y. Liang, Z. Feng, H. Wang,
Nature Communications 2018, 9, 415.
Y. Song, R. Peng, D. K. Hensley, P. V. Bonnesen, L. Liang, Z. Wu,
H. M. Meyer III, M. Chi, C. Ma, B. G. Sumpter, A. J. Rondinone,
ChemistrySelect 2016, 1, 6055-6061.
[11]
[12]
M. R. Thorson, K. I. Siil, P. J. A. Kenis, Journal of The
Electrochemical Society 2013, 160, F69-F74.
M. Borowski, Journal De Physique Iv - J PHYS IV 1997, 7.
Trace amount of acetate – below 3.5% FE – were observed when
using 13CO2 (Figure S23-S24). The 13CO2 used in this study was
This article is protected by copyright. All rights reserved.

10.1002/anie.201907994
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents (Please choose one layout)
RESEARCH ARTICLE
[a]
Dilan Karapinar,[a] Ngoc Tran Huan,
Nastaran Ranjbar Sahraie,[b] Jingkun Li,[b]
David Wakerley[a], Nadia Touati,[c]
Sandrine Zanna,[c] Dario Taverna,[d] Luiz
Henrique Galvão Tizei,[e] Andrea Zitolo,[f]
Frédéric Jaouen,[b] Victor Mougel,[a],[g]*
Marc Fontecave[a]*
Page No. – Page No.
Electroreduction of CO2 on Single-Site
Copper-Nitrogen-Doped
Material: Selective Formation of
Ethanol and Reversible
Restructuration of the Metal Sites
Carbon
N-doped carbon material containing isolated Cu atoms prove highly selective for
CO2 electroreduction to ethanol. Operando spectroscopic characterization of the
catalyst establishes that the active species during electrolysis are transient small
Cu nanoparticles. The restructuration of metal sites is reversible.
This article is protected by copyright. All rights reserved.