Effects of Electrolyte Anions on the Reduction of Carbon Dioxide to Ethylene and Ethanol on Copper (100) and (111) Surfaces

Article abstract of DOI:10.1002/cssc.201801078

The CO₂ electroreduction reaction has been investigated on Cu(100) and Cu(111) surfaces in 0.1 m aqueous solutions of KClO₄, KCl, KBr, and KI electrolyte. The formation of ethylene and ethanol on these surfaces generally increased as the electrolyte anion was changed from ClO₄^?→Cl^?→Br^?→I^?. For example, on Cu(100) at ?1.23 V versus RHE, as the electrolyte anion changed from ClO₄^? to I^?, the faradaic efficiency (FE) of ethylene formation increased from 31 to 50 %, FE_ethanol increased from 7 to 16 %, and the associated current densities increased five- and sevenfold, respectively. A remarkable total FE of up to 74 % for C₂ and C₃ products was obtained in the presence of KI. Despite surface roughening in the presence of the electrolytes, the Cu(100) electrode still enhanced the formation of C₂ compounds better than Cu(111). The favorable reduction of CO₂ to C₂ products in KI electrolyte was correlated with a higher *CO population on the surface, as shown using linear sweep voltammetry. In situ Raman spectroscopy indicated that the coordination environment of *CO was altered by the used electrolyte anion. Thus, apart from affecting the morphology of the electrode and local pH value, we propose that the anion plays a critical role in enhancing the formation of C₂ products by tuning the coordination environment of adsorbed *CO, which gives rise to more efficient C?C coupling.

Full text of DOI:10.1002/cssc.201801078

Accepted Article
Title: Effects of Electrolyte Anions on the Reduction of Carbon Dioxide
to Ethylene and Ethanol on Copper (100) and (111) Surfaces
Authors: Yun Huang, Cheng Wai Ong, and Boon Siang Yeo
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: ChemSusChem 10.1002/cssc.201801078
Link to VoR: http://dx.doi.org/10.1002/cssc.201801078
A Journal of
www.chemsuschem.org

10.1002/cssc.201801078
ChemSusChem
FULL PAPER
Effects of Electrolyte Anions on the Reduction of Carbon Dioxide
to Ethylene and Ethanol on Copper (100) and (111) Surfaces
Yun Huang,[a] Cheng Wai Ong,[a] and Boon Siang Yeo*[a]
Abstract: The CO2 electroreduction reaction has been investigated
on Cu(100) and Cu(111) surfaces in 0.1 M KClO4, KCl, KBr and KI
electrolytes. The formation of ethylene (C2H4) and ethanol (EtOH)
products on these surfaces generally increased as the electrolyte
copper electrodes was observed as the sizes of the cations
(from Li+ to Cs+) used increased. This effect was also found by
Kenis and Bell for the electroreduction of CO2 to CO on silver
electrodes.[6b, 7] Bell’s group further investigated the role of the
electrolyte cation by modeling the local pH and electric field near
the cathode.[6b, 6c] Due to the consumption of protons (supplied
from water dissociation) during CO2RR and hydrogen evolution,
the pH at the surface of the working electrode is expected to
rise.[3] However, larger cations, especially Cs+, are also
predicted to hydrolyze easily ([Cs(H2O)n]+ + H2O ® [Cs(H2O)n-
1(OH)] + H3O+), and their use will buffer the rise of local pH.[6b]
The resultant lowered local pH will facilitate the dissolution of
CO2 near the cathode and consequently enhance the CO2
reduction activity. It was further suggested that hydrated Cs+
ions were more concentrated in the outer Helmholtz plane, and
-
anion was changed from ClO4 ® Cl- ® Br- ® I-. For example, on
Cu(100) at -1.23 V vs. RHE, as the electrolyte anion changed from
-
ClO4 to I-, the Faradaic efficiencies of ethylene (FEethylene) improved
from 31 to 50%, FEethanol increased from 7 to 16%, and the current
densities of ethylene and ethanol showed respectively 5 and 7 folds
enhancement. A remarkable total FE up to 74% for C2 and C3
products were also obtained in the KI electrolyte. Despite their
roughening in the presence of the electrolytes, the Cu(100) electrode
still showed a greater propensity than Cu(111) for enhancing the
formation of C2 compounds. The favorable reduction of CO2 to C2
products in KI electrolyte was correlated with
a higher *CO
population on the surface, as shown using linear sweep voltammetry. could induce a greater electric field to stabilize *CO and *OCCO
In situ Raman spectroscopy indicates that the coordination
environment of *CO was altered by the electrolyte anion used. Thus,
apart from affecting the morphology of the electrode and local pH,
we propose that the anion plays a critical role in enhancing the
formation of C2 products, by tuning the coordination environment of
adsorbed *CO for more efficient C-C coupling.
adsorbates, the key intermediates for ethylene and ethanol
production.[6c]
The effect of electrolyte anion has also been investigated
6d, 6e]
for CO2RR.[6a,
Hori et al. reported higher selectivities
towards ethylene on polycrystalline Cu electrodes in non-
buffering KCl, KClO4 and K2SO4 electrolytes, as compared to
8]
buffers such as KHCO3 and K2HPO4.[6a,
Local pHs are
expected to rise higher when electrolytes with lower buffering
abilities are used. Thus, Hori attributed the higher ethylene
selectivities detected in the non-buffering electrolytes to the
higher local pHs at the electrode surface.[6a] This can be
rationalized in terms of the decoupled proton-electron transfer
step that occurs during the formation of ethylene. Subsequently,
the Ogura group reported selective ethylene formation on
copper electrodes coated with Cu halide in concentrated (3 M)
potassium halide electrolytes.[9] Of the three halides (Cl-, Br- and
I-) used, Br- most favored ethylene production. The halide was
proposed to aid the adsorption of CO2 via charge transfer and
stabilize the pertinent carboxyl (COOH) intermediates. The
absence of infrared spectroscopic signals belonging to adsorbed
CO also led them to propose that this species is not involved in
the formation of ethylene.[9a] Roldan-Cuenya et al. performed
CO2 reduction on preoxidized Cu catalysts in KHCO3 electrolytes
mixed with KCl, KBr or KI (non-buffers).[6e] Therein, the current
densities of ethylene and ethanol increased as the anion was
changed from Cl- ® Br- ® I-.[6e] The higher selectivity and activity
for C2 products observed on these electrodes (compared to a
copper foil reference) was not attributed to significant changes in
local pHs as a KHCO3-based buffer was used. The change in
activity was instead ascribed to effects from the adsorbed halide
anion, subsurface oxygen and roughened surface. Here, we
note that both Cu halides and preoxidized Cu electrodes are
already known to promote the formation of C2 products.[6e, 9a, 10]
Thus, it is difficult to clarify the role of the electrolyte anion on
the selectivity of CO2RR.
Introduction
Reducing CO2 emission from the use of fossil fuels is urgently
needed to prevent undesirable global warming.[1] Using
renewable electricity and catalysts to electroreduce CO2 to
chemicals and fuels offers a promising approach to reduce our
reliance on fossil fuels.[2] Copper-based catalysts are particularly
attractive for such a purpose, owing to their unique ability to
reduce CO2 to ethylene and ethanol.[3] However, the poor
product selectivity and large overpotentials required by these
catalysts make them inapt for industrial applications. Hence,
considerable resources have been invested to tune the structure
of the Cu catalysts, and their electrochemical environment, in
order to optimize their CO2–to-C2 selectivity.
The activity of Cu catalysts for the CO2 reduction reaction
(CO2RR) is influenced by their surface orientation,[4] grain
boundaries,[5] type of electrolytes used,[6] etc. The importance of
electrolyte cation was first reported by Hori et al.[6f] An increase
in the selectivities for CO, ethylene and ethanol production on
[a]
Y. Huang, C.W. Ong, Prof. Dr. B.S. Yeo
Department of Chemistry, National University of Singapore
3 Science Drive 3, Singapore 117543
E-mail: chmyeos@nus.edu.sg
Supporting information for this article is given via a link at the end of
the document.((Please delete this text if not appropriate))
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801078
ChemSusChem
FULL PAPER
Figure 1. Scanning electron micrographs of Cu(100) and Cu(111) electrodes (A-B) before CO2 reduction; after CO2 reduction in (C-D) KClO4, (E-F) KCl, (G-H)
KBr and (I-J) KI electrolytes at -1.23 V vs. RHE. Scale bar: 500 nm. (K-T) The corresponding cyclic voltammograms of the same electrodes (in N2-saturated 0.1 M
KOH, scan rate of 50 mV/s).
In this work, we examine the product distribution of CO2
reduction on Cu(100) and Cu(111) electrodes in non-buffering
KClO4, KCl, KBr and KI electrolytes. The formation of ethylene
and ethanol increased on both surfaces as the electrolyte anion
surfaces had up to 1.6´ larger surface roughness factors than
those before CO2 reduction (Table S1 of the Supporting
Information). The surface roughening could be due to formation
of CuBr/CuI particles on the Cu surfaces upon their contact with
electrolytes at open circuit condition, followed by their reduction
during CO2 electrolysis (Figure S1 of the Supporting
Information).[6e] Interestingly, the CVs of the Cu electrodes after
CO2RR resembled those of freshly-prepared ones (Figure 1K-T).
This indicates that the surface orientation of the Cu single
crystals were largely maintained after electrolysis, consistent
with additional results from high resolution and 2D X-ray
diffraction analyses of the electrodes (Figures S2 and S3 of the
Supporting Information).
-
was varied from ClO4 ® Cl- ® Br- ® I-. Effects of morphology
and local pH were examined. Linear sweep voltammetry and
Raman spectroscopy were also used to probe the anion’s effect
on CO adsorption on Cu. The effect of the anion on the
formation of C2 compounds is discussed.
Results and Discussion
Characterization of the electrodes before and after CO2
reduction
Distribution of CO2 reduction products on Cu(100) and
Cu(111) surfaces
Scanning electron microscopy (SEM) of freshly-prepared
Cu(100) and Cu(111) electrodes showed that their surfaces
were smooth (Figure 1). Their cyclic voltammograms (CV) were
consistent with previous studies, as evidenced by their OH-
adsorption/desorption peaks located between -0.2 and 0.3 V vs.
RHE (all potentials cited in this work are referenced to the
reversible hydrogen electrode).[11]
CO2 reduction was then performed on these surfaces for
40 min in 0.1 M KClO4, KCl, KBr and KI electrolytes at a
representative potential of -1.23 V. SEM images of these
surfaces after CO2RR revealed their roughening, especially
when electrolysis were performed in KBr and KI electrolytes
(Figure 1G-J).[6d] We could not discern any differences between
the way the two surfaces were roughened in each electrolyte.
Double-layer capacitance measurements showed that these
CO2 reduction was performed on Cu(100) and Cu(111)
electrodes at potentials between -0.98 and -1.38 V in 0.1 M
KClO4, KCl, KBr and KI electrolytes. The products were
analyzed using gas and liquid chromatography (Figures S4 and
S5 of the Supporting Information). The data is grouped based on
our products of interest, which are CO, formate, ethylene,
ethanol and CH4 (Figures 2 and 3, Figure S6 of the Supporting
Information). The FEs for other products including hydrogen, n-
propanol, etc. are listed in Tables S2-9 of the Supporting
Information.
Cu(100)
Starting at a potential of -0.98 V, CO production on
Cu(100) was favored in KI over KClO4 electrolyte (Figure 2A-B).
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801078
ChemSusChem
FULL PAPER
Specifically, when the electrolyte was switched from KClO4 to KI,
FECO increased from 11.8% to 22.8% and the corresponding jCO
doubled from -0.06 to -0.12 mA/cm2. At -1.08 V, ethylene and
ethanol production increased while CO evolution decreased
(Figure 2C-F). These observations indicate that CO is an
intermediate for the formation of these C2 products.[12]
The selectivity towards ethylene peaked at -1.23 V in all
four electrolytes, with FEethylene of 50.3% in KI and FEethylene of
30.6% in KClO4 (Figure 2C-D). In KI electrolyte, the optimum
jethylene reached -6.06 mA/cm2, which was 5 times higher than
that in KClO4 (-1.20 mA/cm2). The enhancement in ethanol
production follows the same trend as that of ethylene (Figure
2E-F). At -1.23 V, the optimized FEethanol increased from 7.1% to
16.4% and the corresponding jethanol improved from -0.28 to -1.98
mA/cm2 when switching electrolytes from KClO4 to KI.
It is noteworthy that the total FEs for C2 and C3 products
at -1.23 V reached as high as 74 % in KI electrolyte. This value
is the state-of-the-art as compared to those reported for plasma-
activated Cu (FEC2-C3 = ~65%),[6e] Cl-induced Cu2O-Cu (FEC2-C3
~60%)[13] and oxide-derived Cu catalysts (FEC2-C3 = ~70%).[14]
=
Figure 2. Faradaic efficiencies and partial current densities for (A-B) CO, (C-D) ethylene, (E-F) ethanol, (G-H) HCOO- and (I-J) CH4 produced on Cu(100) in 0.1
M KClO4, KCl, KBr and KI electrolytes at various applied potentials.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801078
ChemSusChem
FULL PAPER
Figure 3. Faradaic efficiencies and partial current densities for (A-B) ethylene and (C-D) ethanol produced on Cu(111) in 0.1 M KClO4, KCl, KBr and KI
electrolytes at various applied potentials.
17]
Selectivity towards formate is less favored in KBr/KI as
compared to KClO4 electrolytes (Figure 2G-H). At -0.98 V, the
FEformate reduced from 17.3% in KClO4 to 11.9% in KI electrolyte,
in contrast to the enhancement of CO and ethylene production
observed at the same conditions. This can be ascribed to the
competing production of formate and CO (and its further
reduced products).[15]
At potentials negative to -1.28 V, Cu(100) became
selective for CH4 production in KClO4, KCl and KBr electrolytes
(Figure 2I-J). The highest FECH4 of 49.1% was obtained in KBr
electrolyte at -1.38 V, while a highly suppressed FECH4 of 11.9%
was found in KI electrolyte at the same potential. Despite the
generally suppressed FECH4 in KI, the geometric current density
for CH4 in KI is comparable to that in the KBr system. This is
likely due to the increased surface roughness of the Cu
electrode in KI electrolytes.
the activation barrier for CO dimerization.[4a,
The potential-
dependent production of CO, formate and CH4 on Cu(111) in the
four electrolytes were similar to that observed on Cu(100)
(Figure S6 of the Supporting Information).
Factors responsible for the C2 enhancement
We now investigate the reasons behind the increased formation
of ethylene and ethanol as the electrolyte anion was changed
from ClO4 ® Cl- ® Br- ® I-. Is it due to electrode roughening,
-
rise in local pH induced from the non-buffering electrolytes, or
the anion’s role in tuning the *CO adsorption environment?
Morphological factor
A
Cu(100) electrode was used for two consecutive CO2
reductions (each for 40 min) at -1.23 V, first in KI and then in
KClO4 electrolyte. The production of C2 compounds in KI was
significantly reduced during the second electrolysis in KClO4
(Table 1, Table S10 of the Supporting Information). Specifically,
the FEethylene and FEethanol decreased respectively from 50.3 to
37.2% and 16.4 to 10.0%. Nonetheless, the production of
ethanol and ethylene at the 2nd electrolysis [i.e., KI-treated
Cu(100) in KClO4] is still higher than that shown by a fresh
Cu(100) in KClO4 electrolyte (FEethylene = 30.6%, FEethanol = 7.1%
at -1.23 V). Its total current density is also larger by 1.5´. This
enhancement (in the production of C2 compounds) can be
attributed to the more extensive roughening of its surface after
the 1st electrolysis in KI electrolyte. However, it is significantly
smaller than the enhancement observed using only KI as the
electrolyte. Therefore, we conclude that while a rough Cu
surface may contain catalytically-active defect sites for reducing
CO2 to C2 products, the morphology itself cannot be the sole
responsible factor.
Cu(111)
The trend in which C2 products are formed on Cu(111) in
different electrolytes is similar to that described on Cu(100).
Changing the electrolyte from KClO4 to KI increased the
optimum FEethylene from 13.3 to 35.7% and the corresponding
jethylene from -0.19 to -2.96 mA/cm2 (Figure 3A-B). The FE and
current densities of ethanol also improved (Figure 3C-D).
It is noteworthy that though Cu(100) and Cu(111) were
visibly roughened during CO2 electrolysis, their original surface
crystallographies were largely preserved (Figure 1, Figures S2
and S3). This could explain the higher ethylene selectivity
observed on Cu(100) as compared to on Cu(111). Such a
surface structure effect has been confirmed by density functional
theory (DFT) calculations and by performing CO2 reduction on
Cu foils, whose exposed facets are tuned using metal ion
cycling.[16] The propensity of Cu(100) electrodes to reduce CO2
to ethylene has also been attributed to the presence of a high
*CO coverage on the electrode during CO2RR, which lowered
Local pH
The electrolytes used in this work are non-buffers.[8] Previous
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801078
ChemSusChem
FULL PAPER
Table 1: Faradaic efficiencies for CO2 reduction products formed on Cu(100) during sequential CO2 reductions in 0.1 M KI and 0.1 M KClO4 electrolytes, and
Cu(100) in 0.1 M KClO4 electrolyte. All experiments were performed at -1.23 V for 40 min.
Faradaic Efficiency (%)
jtot (mA/cm2)
Other liquid
CH4
C2H4
CO
H2
EtOH
HCOO-
Total
products[a]
1st electrolysis of Cu(100) in KI
-12.06
-5.89
-3.93
9.2
50.3
37.2
30.6
0.5
19.9 16.4
1.5
4.8
6.5
7.1
6.0
5.1
104.9
101.9
103.5
2nd electrolysis of Cu(100) in KClO4
14.1 10.0
28.3
1.5
Cu(100) in KClO4
23.6
2.2
28.3
7.1
[a]: Other liquid products include acetaldehyde, propionaldehyde, methanol, ally alcohol and n-propanol. The standard deviations of the measurements have
been listed in Table S10 of the Supporting Information
Figure 4. (A) Linear sweep voltammograms of Cu(100) electrodes in 0.1 M KClO4, KCl, KBr and KI electrolytes saturated in CO (solid lines) and N2 (dotted lines).
Scan rate = 10 mV/s. (B) Correlation between the integrated charges beneath CO adsorption peaks (from a) and the partial current densities for ethylene (black)
and ethanol (red) produced on Cu(100) at -1.23 V in the four electrolytes.
studies have shown that local pH is dependent on the applied
current densities.[18] Herein, we model the local pH within the
diffusion layer using Gupta’s model.[18a, 19] As expected, the pH
rises when going from bulk electrolyte to the surface of the
Cu(100) during CO2 reduction at -1.23 V (Table S11 and Figure
S7 of the Supporting Information). Experimentally, Cu(100) in KI
electrolytes exhibited the largest current densities. Thus, its
surface pH (pH 11.0) was calculated to be the highest as
compared to those electrodes held in the other three electrolytes
(pHs between 10.3 and 10.4).
To investigate the effect of local pH on ethylene formation,
we analyzed the Faradaic efficiencies for ethylene on Cu(100) in
the four electrolytes at -1.23 V against the calculated surface pH
values (Figure S8 of the Supporting Information). Interestingly,
while KClO4, KCl and KBr electrolyte systems shared similar
surface pHs at ~10.4, the spread of their FEethylene values varied
greatly from 30.6 to 46.9%. The weak correlation of local pH and
FEethylene suggests that local pH is not a dominant factor for
enhancing the formation of C2 products during CO2RR in KClO4,
KBr and KCl electrolytes.
Information). By comparing the cathodic currents on Cu(100) in
CO- and N2-saturated electrolytes, we noted that the onsets of
H2 evolution were delayed under CO environment (by 44-87 mV
at -1.2 mA/cm2). We attribute this to the presence of adsorbed
*CO which blocks available sites for HER and also weakens Cu-
H bonding.[20] Copper’s HER activity will thus be lower as it lies
on the right-hand (weak-binding) side of the HER volcano
plot.[20a]
Reduction peaks also appeared at approx. -0.45 V during
the cathodic scans of Cu(100) in CO-saturated 0.1 M KClO4, KCl,
KBr and KI electrolytes (Figure 4A). These peaks could be
ascribed to CO adsorbed onto Cu(100) accompanied by electron
20b, 21]
transfer.[19,
In support of this, we note that the first CO
reduction product, ethylene, was observed only at -0.76 V on the
same surface in CO-saturated KI electrolyte (Table S12 of the
Supporting Information). Roberts et al. had also reported an
onset potential of -0.5 V vs. RHE for CO reduction on Cu(100) in
pH 7 electrolyte.[21b]
The integrated charges of the CO adsorption peaks (which
increased as the electrolyte anion was changed from ClO4- ® Cl-
® Br- ® I-) are associated with the population of *CO on the Cu
surface. Interestingly, when the current densities of ethylene and
ethanol obtained on a Cu(100) surface at -1.23 V was plotted
against these charges, a linear correlation was observed (Figure
4B). This indicates that enhancement in C2 products is linked to
Role of anion
CO adsorption on Cu single crystals: The adsorption of CO
on Cu in different electrolytes was probed by linear sweep
voltammetry (Figure 4A and Figure S9 of the Supporting
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801078
ChemSusChem
FULL PAPER
the increased *CO population on the electrode. This result
agrees with our previous studies of CO adsorption on CuCl-
Although CO production was strongest in KI electrolyte at -
0.98 V (Figure 2 and Figure S6 of the Supporting Information),
its *CO Raman bands did not show the highest intensities. This
could be related to the roughening of the surface, which affects
the optical properties of the surface. Additionally, no peaks
corresponding to Cu-I (~125/~140 cm-1) or Cu-Br (~135/~170
cm-1) could be discerned,[25] as they were obscured by the
background peak at ~100 cm-1. We also could not detect signals
belonging to further reduced species, which can be attributed to
the insufficient limits of detection afforded by our spectrometer.
derived Cu mesocrystals, Cu nanoparticles and polycrystalline
22]
Cu surfaces.[19,
Therein, we found that catalysts, which
exhibited good selectivity towards CO2 reduction to ethylene,
adsorb CO most readily. We further highlight that Verdaguer-
Cassadevall et al. have reported the presence of strongly
chemisorbed CO on Cu catalysts active for the reduction of CO2
to C2 compounds.[23]
We note that the LSVs of Cu(111) did not show the charge
transfer peak (Figure S9 of the Supporting Information). This is
consistent with previous findings by Hori et al.[20b] Nonetheless,
Effects of anions on C2 production: In this work, the formation
of ethylene and ethanol on Cu(100) and Cu(111) surfaces
generally increase as the electrolyte anion was changed from
-
suppression in HER was similarly observed in the order of ClO4
< Cl- < Br- < I-.
-
ClO4 ® Cl- ® Br- ® I-. This trend mirrors that of the binding
-
*CO binding environment on Cu single crystals probed by in
situ Raman spectroscopy: In situ Raman spectroscopy was
abilities of the anions on Cu in the order of ClO4 < Cl- < Br- < I-
[26]
.
The Cu(100) facet was also shown better than Cu(111) in
-
performed on copper in CO2-saturated 0.1 M KX (X = ClO4 , Cl-,
enhancing the formation of C2 products.
Br- and I-) electrolytes (Figure 5 and Figure S10 of the
Supporting Information). Mechanically polished polycrystalline
Cu was used for this purpose as we found it easier to focus the
Raman laser tightly on this surface as compared to the smooth
single crystal surface. The working potential was set at -0.98 V,
where strong CO production is expected (Figure 2 and Figure
We acknowledge that the electrolyte anions might compete
with active CO2RR intermediates for adsorption sites. However,
our LSV results suggest that *CO adsorption on Cu surface was
actually enhanced by the co-adsorbed anions. The resultant
high population of *CO could then in turn facilitate C-C coupling
of the CO intermediates to C2 products. [4a, 17a] Our hypothesis is
supported by Shaw et al., who revealed using DFT calculations,
that the *CO binding energy on Cu can be increased by up to
0.2 eV in the presence of co-adsorbed anions (F- and Cl-).[21a] It
is noteworthy that an infrared spectroscopy study by Scarano et
al. demonstrated the greater ability of CuCl surface to bind to
CO.[27] Our propositions also agree well with recent DFT
calculations by Nørskov and our group, which showed that
increasing *CO coverage on Cu surfaces lowers the energy
barrier for C-C coupling.[4a, 17a]
S6 of the Supporting Information).
Only Raman bands at 280, 365 and ~2060-2090 cm-1 were
recorded (Figure 5). These can be assigned respectively to the
Cu-CO frustrated rotation, Cu-CO stretching and CºO stretching
Additionally, the adsorbed anion could modify the
electronic structure of the local Cu sites, which could induce a
more positive charge on carbon atom of *CO adsorbate. Anion-
free copper sites, on the other hand, tend to promote π-back
donation (from Cu) into the antibonding 2π* orbital of *CO,[28]
resulting in a more negatively charged carbon atom of *CO
adsorbate. The electrostatic attraction between the oppositely
charged carbon atoms could render stability to the CO
dimerization product and in turn facilitate the C-C coupling
process. Goddard et al. have similarly attributed the enhanced
formation of C2 products on Cu2O-derived Cu surface to the
electrostatic stabilization between two carbon atoms induced by
Cu+ and Cu0 interface.[29] Their DFT calculations suggest such
stabilization could improve the thermodynamics and kinetics of
CO dimerization.
Figure 5. In situ Raman spectra of polycrystalline Cu cathode during
electrochemical reduction of CO2 in KClO4, KCl, KBr and KI electrolytes at –
0.98 V. The top Insert shows the red shifting of the CºO stretching bands in
-
the order of ClO4 < Cl- < Br- < I-, and the bottom Insert presents a schematic
illustration of CO adsorbed on the Cu surface.
Hori et al. first reported the use of non-buffers for
enhancing the formation of C2 products and the improved
performance was attributed to the rise of local pH at the working
electrode surface.[8] Roldan Cuenya et al. further demonstrated
increased CO2 reduction rate on preoxidized Cu electrodes in
halide containing KHCO3 buffers,[6e] and factors including
subsurface oxygen, anion adsorption and surface roughening
were proposed for the enhancement. In this work, we have
demonstrated that the enhancement in ethylene and ethanol
modes of adsorbed *CO.[24] Interestingly, the stretching vibration
of CºO decreased from 2087 to 2060 cm-1, when the electrolyte
was changed from KClO4 to KI. The weakening of CºO
stretching indicates that the coordination environment of CO was
altered by the presence of different anions.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801078
ChemSusChem
FULL PAPER
solution. The gas products were periodically analyzed by gas
chromatography (GC, 7890A, Agilent). Liquid products were collected
after chronoamperometry and quantified using headspace gas
chromatography (HSGC, Agilent, 7890B) and high-performance liquid
chromatography (HPLC, Agilent, 1260 Infinity). The product distribution
reported here is the average of at least three independent measurements
at each potential. All the currents were normalized to the exposed
geometric surface area (0.385 cm2) of the working electrode.
production (in terms of both Faradaic efficiencies and partial
current densities) strongly correlates with the population of *CO
on the surface. The presence of the adsorbed anion, specifically
I-, is suggested to be a dominant cause for the enhancement.
We show that the catalytic selectivity and activity of
electrochemical reduction of CO2 can be tuned by the judicious
selection of appropriate electrolyte anion.
Modeling of local pH: A steady-state 1-D modeling in MATLAB 8.5 was
performed to calculate the local pH within the diffusion layer in various
electrolytes, and to investigate its effect on C2 production.[18a, 19]
Conclusions
We have examined the electroreduction of CO2 on Cu(100) and
Cu(111) surfaces in 0.1 M KClO4, KCl, KBr and KI electrolytes.
The formation of ethylene and ethanol on these surfaces was
enhanced when the anion of the electrolyte was varied in the
Acknowledgements
-
order of ClO4 ® Cl- ® Br- ® I-. The use of KI as the electrolyte
This work is supported by a research fund (R-143-000-683-112)
from the Ministry of Education, Singapore. Y.H. and C.W.O
acknowledge Ph.D. research scholarships from the Ministry of
Education, Singapore. We thank Dr. Albertus D. Handoko
(Institute of Materials Research and Engineering) for the XRD
measurements.
yielded the highest FEethylene of 50% and FEethanol of 16%, and
with the total FE for C2-C3 products reaching 74% at -1.23 V.
Apart from affecting the surface morphologies and local pHs of
the electrodes, we propose that the electrolyte anion can
facilitate a higher population of adsorbed *CO, which thus
promotes their C-C coupling to C2 products.
Keywords: carbon dioxide reduction • copper single crystals •
anion • ethylene • ethanol
Experimental Section
[1]
[2]
P. Friedlingstein, R. M. Andrew, J. Rogelj, G. P. Peters, J. G. Canadell,
R. Knutti, G. Luderer, M. R. Raupach, M. Schaeffer, D. P. van Vuuren,
C. Le Quéré, Nat. Geosci. 2014, 7, 709-715.
Electrode preparation and characterization: Cu(100) and Cu(111)
single-crystal surfaces (99.99%, 10 mm diameter, MTI Corp.) were
prepared through mechanical polishing, electropolishing and acid rinsing
in 0.1 mM HClO4 solution.[30] The surface morphology of the electrodes
was characterized using scanning electron microscopy (JEOL JSM-6710
F, 5 kV, 10 µA). Their surface orientation were checked by cyclic
voltammetry in 0.1 M KOH,[11] as well as high resolution and 2D X-ray
diffraction. The surface roughness factors of these electrodes after CO2
reduction were estimated using their double-layer capacitances. CO
adsorption on the Cu single-crystal surfaces was examined by linear
sweep voltammetry.[4b] An Autolab PGSTAT100 potentiostat was used
for controlling the aforementioned electrochemistries. In situ Raman
spectroscopy (Horiba Jobin Yvon) was performed to probe the structure
of the *CO intermediates on mechanically polished polycrystalline Cu
surfaces. A He-Ne laser (633 nm, CVI Melles Griot) was employed as the
excitation source. A water-immersion objective (Olympus LUMFL, 60×,
numerical aperture = 1.10) protected with an optically transparent Teflon
film (0.013 mm, American Durafilm) was used during the in situ
electrochemical measurements.
D. Ren, N. W. X. Loo, L. Gong, B. S. Yeo, ACS Sustainable Chem. Eng.
2017, 5, 9191-9199.
[3]
[4]
M. Gattrell, N. Gupta, A. Co, J. Electroanal. Chem. 2006, 594, 1-19.
a) Y. Huang, A. D. Handoko, P. Hirunsit, B. S. Yeo, ACS Catal. 2017, 7,
1749-1756; b) Y. Hori, I. Takahashi, O. Koga, N. Hoshi, J. Mol. Catal.
A: Chem. 2003, 199, 39-47.
[5]
[6]
X. Feng, K. Jiang, S. Fan, M. W. Kanan, J. Am. Chem. Soc. 2015, 137,
4606-4609.
a) Y. Hori, A. Murata, R. Takahashi, J. Chem. Soc., Faraday Trans. 1
1989, 85, 2309-2326; b) M. R. Singh, Y. Kwon, Y. Lum, J. W. Ager, A.
T. Bell, J. Am. Chem. Soc. 2016, 138, 13006-13012; c) J. Resasco, L.
D. Chen, E. Clark, C. Tsai, C. Hahn, T. F. Jaramillo, K. Chan, A. T. Bell,
J. Am. Chem. Soc. 2017, 139, 11277-11287; d) A. S. Varela, W. Ju, T.
Reier, P. Strasser, ACS Catal. 2016, 6, 2136-2144; e) D. Gao, F.
Scholten, B. Roldan Cuenya, ACS Catal. 2017, 7, 5112-5120; f) A.
Murata, Y. Hori, Bull. Chem. Soc. Jpn. 1991, 64, 123-127.
M. R. Thorson, K. I. Siil, P. J. A. Kenis, J. Electrochem. Soc. 2013, 160,
F69-F74.
[7]
[8]
Y. Hori, in Modern Aspects of Electrochemistry, Vol. 42 (Eds.: C. G.
Vayenas, R. E. White, M. E. Gamboa-Aldeco), Springer New York, New
York, NY, 2008, pp. 89-189.
Electrochemical reduction of CO2: 0.1 M KClO4 (99.99%, Sigma-
Aldrich), KCl (99.99%, Sigma-Aldrich), KBr (99.95%, Meryer Chemical)
and KI (99.99%, Meryer Chemical) electrolytes were prepared with
ultrapure water (Type I, Barnstead, Thermo Scientific). We employed a
[9]
a) K. Ogura, J. CO2 Util. 2013, 1, 43-49; b) H. Yano, T. Tanaka, M.
Nakayama, K. Ogura, J. Electroanal. Chem. 2004, 565, 287-293.
two-compartment electrochemical cell,[31]
separated by a cation-
[10] D. Gao, I. Zegkinoglou, N. J. Divins, F. Scholten, I. Sinev, P. Grosse, B.
Roldan Cuenya, ACS Nano 2017, 11, 4825-4831.
exchange membrane (Selemion CMV, AGC Asahi Glass), to perform
CO2/CO chronoamperometric measurements using a Gamry Reference
600 potentiostat. 20 sccm of CO2 (99.999%, Linde Gas) or 5 sccm of CO
(99.9%, Linde Gas) were purged into the two compartments during the
40 min electrolysis. A Ag/AgCl electrode (Saturated KCl, Pine Research
Instrumentation) and a Pt wire (ALS Japan) were employed as reference
and counter electrodes respectively. The current interrupt mode was
used for compensating the iR drop. All potentials reported here were
referenced to the RHE. The pHs were measured to be ~3.8 for CO2-
saturated KClO4, KCl, KBr and KI solutions and 5.8 for CO-saturated KI
[11] a) V. D. Jović, B. Jovic M., J. Serb. Chem. Soc. 2002, 67, 531-546; b) K.
J. P. Schouten, E. P. Gallent, M. T. M. Koper, J. Electroanal. Chem.
2013, 699, 6-9.
[12] R. Kortlever, J. Shen, K. J. P. Schouten, F. Calle-Vallejo, M. T. M.
Koper, J. Phys. Chem. Lett. 2015, 6, 4073-4082.
[13] S. Lee, D. Kim, J. Lee, Angew. Chem., Int. Ed. 2015, 54, 14701-14705.
[14] Y. Lum, B. Yue, P. Lobaccaro, A. T. Bell, J. W. Ager, J. Phys. Chem. C
2017, 121, 14191-14203.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801078
ChemSusChem
FULL PAPER
[15] J. S. Yoo, R. Christensen, T. Vegge, J. K. Nørskov, F. Studt,
ChemSusChem 2016, 9, 358-363.
[16] K. Jiang, R. B. Sandberg, A. J. Akey, X. Liu, D. C. Bell, J. K. Nørskov, K.
Chan, H. Wang, Nat. Catal. 2018, 1, 111-119.
[17] a) J. H. Montoya, C. Shi, K. Chan, J. K. Nørskov, J. Phys. Chem. Lett.
2015, 6, 2032-2037; b) R. B. Sandberg, J. H. Montoya, K. Chan, J. K.
Nørskov, Surf. Sci. 2016, 654, 56-62.
[18] a) N. Gupta, M. Gattrell, B. MacDougall, J. Appl. Electrochem. 2005, 36,
161-172; b) J. Resasco, Y. Lum, E. Clark, J. Z. Zeledon, A. T. Bell,
ChemElectroChem 2018, 5, 1064-1072.
[19] 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.
[20] a) Y.-J. Zhang, V. Sethuraman, R. Michalsky, A. A. Peterson, ACS
Catal. 2014, 4, 3742-3748; b) Y. Hori, H. Wakebe, T. Tsukamoto, O.
Koga, Surf. Sci. 1995, 335, 258-263.
[21] a) S. K. Shaw, A. Berná, J. M. Feliu, R. J. Nichols, T. Jacob, D. J.
Schiffrin, Phys. Chem. Chem. Phys. 2011, 13, 5242-5251; b) F. S.
Roberts, K. P. Kuhl, A. Nilsson, ChemCatChem 2016, 8, 1119-1124.
[22] C. S. Chen, A. D. Handoko, J. H. Wan, L. Ma, D. Ren, B. S. Yeo, Catal.
Sci. Technol. 2015, 5, 161-168.
[23] A. Verdaguer-Casadevall, C. W. Li, T. P. Johansson, S. B. Scott, J. T.
McKeown, M. Kumar, I. E. L. Stephens, M. W. Kanan, I. Chorkendorff,
J. Am. Chem. Soc. 2015, 137, 9808-9811.
[24] B. D. Smith, D. E. Irish, P. Kedzierzawski, J. Augustynski, J.
Electrochem. Soc. 1997, 144, 4288-4296.
[25] J. Serrano, M. Cardona, T. M. Ritter, B. A. Weinstein, A. Rubio, C. T.
Lin, Phys. Rev. B 2002, 66, 245202.
[26] a) A. Ignaczak, J. A. N. F. Gomes, J. Electroanal. Chem. 1997, 420, 71-
78; b) I. T. McCrum, S. A. Akhade, M. J. Janik, Electrochim. Acta 2015,
173, 302-309; c) O. M. Magnussen, Chem. Rev. 2002, 102, 679-726.
[27] D. Scarano, P. Galletto, C. Lamberti, R. De Franceschi, A. Zecchina,
Surf. Sci. 1997, 387, 236-242.
[28] G. Blyholder, J. Phys. Chem. 1964, 68, 2772-2777.
[29] H. Xiao, W. A. Goddard, T. Cheng, Y. Liu, Proc. Natl. Acad. Sci. U. S. A.
2017, DOI: 10.1073/pnas.1702405114.
[30] I. Takahashi, O. Koga, N. Hoshi, Y. Hori, J. Electroanal. Chem. 2002,
533, 135-143.
[31] D. Ren, Y. Deng, A. D. Handoko, C. S. Chen, S. Malkhandi, B. S. Yeo,
ACS Catal. 2015, 5, 2814-2821.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801078
ChemSusChem
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
FULL PAPER
Switching the electrolyte anion
from ClO4- to I- improves the
production of ethylene and ethanol
on Cu(100) surfaces, with the
highest Faradaic efficiencies of
50% for ethylene and 16% for
ethanol. The high population of
*CO facilitated by the electrolyte
anion greatly promotes C-C
coupling to C2 products.
Y. Huang, C. W. Ong, B. S. Yeo*
Page No. – Page No.
Effects of Electrolyte Anions on
the Reduction of Carbon Dioxide
to Ethylene and Ethanol on
Copper (100) and (111) Surfaces
This article is protected by copyright. All rights reserved.