On the Oxidation State of Cu2O upon Electrochemical CO2 Reduction: An XPS Study

Article abstract of DOI:10.1002/cphc.201900468

The encouraging selectivity of copper oxides for the electroreduction of CO₂ into ethylene and alcohols has led to a vivid debate on the possible relation between their operando (sub-)surface oxidation state (i. e. fully reduced or partially oxidized) and this distinct reactivity. The high roughness of the Cu oxides used in previous studies on this matter adds complexity to this controversy and motivated us to prepare quasi-planar Cu₂O thin films that displayed a CO₂ reduction selectivity similar to that of oxide-derived copper catalysts reported in previous studies. Most importantly, when the post-mortem thin films were transferred for characterization in an air-free environment, X-ray photoelectron spectroscopy measurements confirmed their complete reduction in the course of the CO₂ reduction reaction. Thus, our results indicate that the selectivity of the Cu oxides featured in previous studies stems from their enhanced roughness, highlighting the importance of controlled sample transfer upon post-mortem characterization with ex situ techniques.

Full text of DOI:10.1002/cphc.201900468

Accepted Article
Title: On the Oxidation State of Cu2O upon Electrochemical CO2
Reduction: an XPS Study
Authors: Anastasia A. Permyakova, Juan Herranz, Mario El Kazzi,
Justus S. Diercks, Mauro Povia, Léa Rose Mangani, Michael
Horisberger, Alexandra Pătru, and Thomas J. Schmidt
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: ChemPhysChem 10.1002/cphc.201900468
Link to VoR: http://dx.doi.org/10.1002/cphc.201900468
A Journal of
www.chemphyschem.org

ChemPhysChem
10.1002/cphc.201900468
On the Oxidation State of Cu O
2
upon Electrochemical CO Reduction:
2
an XPS Study
Dr. Anastasia A. Permyakova,[a],‡ Dr. Juan Herranz,[a],‡,* Dr. Mario El Kazzi, Justus S.
Diercks, [a] Dr. Mauro Povia, [a] Léa Rose Mangani, [a] Michael Horisberger, [b] Dr. Alexandra
Pătru, [a] and Prof. Dr. Thomas J. Schmidt[a,c]
[a]
[a]Electrochemistry Laboratory, Paul Scherrer Institut
Forschungsstrasse 111, 5232 Villigen PSI, Switzerland
[b]Laboratory for Scientific Developments and Novel Materials, Paul Scherrer Institut
Forschungsstrasse 111, 5232 Villigen PSI, Switzerland
[c]Laboratory of Physical Chemistry, ETH Zurich
HCI G 215, Vladimir-Prelog-Weg 1-5/10, 8093 Zurich, Switzerland
‡
*
Corresponding author: juan.herranz@psi.ch; these authors contributed equally to this work
This article is protected by copyright. All rights reserved.
1

ChemPhysChem
10.1002/cphc.201900468
Abstract
The encouraging selectivity of copper oxides for the electroreduction of CO
2
into ethylene and
alcohols has led to a vivid debate on the possible relation between their operando (sub-)surface
oxidation state (i.e., fully reduced or partially oxidized) and this distinct reactivity. The high
roughness of the Cu-oxides used in previous studies on this matter adds complexity to this
controversy, and motivated us to prepare quasi-planar Cu
2
O thin films that displayed a
CO -reduction selectivity similar to that of oxide-derived copper catalysts reported in previous
2
studies. Most importantly, when the post-mortem thin films were transferred for characterization
in an air-free environment, X-ray photoelectron spectroscopy measurements confirmed their
complete reduction in the course of the CO -reduction reaction. Thus, our results indicate that the
2
selectivity of the Cu-oxides featured in previous studies stems from their enhanced roughness,
while highlighting the importance of controlled sample transfer upon post-mortem characterization
with ex situ techniques.
Introduction
The electrochemical reduction of carbon dioxide to added-value chemicals is gaining attention as
[
1]
a means to valorize CO
interest, the large overpotentials and poor selectivities displayed by most CO
electrocatalysts represent a major obstacle for the development of an implementable technology,
2
while decreasing its emissions and environmental impact. Despite this
2
-reduction
which would additionally require a rigorous assessment of the CO
2
-reduction products with the
greatest marketability potential.[2] In this regard, some studies have suggested that
CO
2
-electroreduction to formate or H
2
/CO mixtures (i.e., syngas) may be cost-competitive when
performed using Sn- or Au-based catalysts implemented in electrode/cell configurations allowing
This article is protected by copyright. All rights reserved.
2

ChemPhysChem
10.1002/cphc.201900468
to draw currents > 100 mA∙cm-2.[3−5] On the other hand, the production of ethylene or methanol
using the same setups would lead to estimated production costs ≈ 2-fold above current market
values[3,4] – a price gap that could be significantly narrowed by substituting the copper catalysts
considered in the above studies[3,4] with copper oxides, which display enhanced selectivities and
[6]
stabilities for CO -to-ethylene/-methanol reduction.
2
These distinct properties have motivated a significant research effort to understand the factors that
induce the CO
2
-reduction selectivity of these oxides, which customarily consist of nanostructured,
O) composition.[6−10] Specifically, a vivid
high surface area catalysts with a cuprous oxide (Cu
2
debate is currently ongoing regarding the oxidation state of this initial Cu
CO -reduction reaction. On the one hand, the latter takes place at potentials << 0 volts vs. the
reversible hydrogen electrode (VRHE) that are well below the Cu
2
O in the course of the
2
0
2
O/Cu equilibrium (at ≈ 0.5 VRHE),
[
11]
and from which the operando reduction of the initial oxide would be expected – an extent that
has been predicted by several computational studies[12,13] and experimentally demonstrated using
in situ X-ray diffraction[14] and Raman spectroscopy,[10,15] as well as post-mortem secondary ion
mass spectrometry (SIMS).[16] On the other hand, and in clear conflict with those works, a greater
number of recent studies implementing in situ or post-mortem X-ray absorption, photoelectron or
electron energy loss spectroscopies (XAS, XPS, EELS) have concluded that the initial oxide does
not get completely reduced in the course of the CO -reduction process. Moreover, those same
2
works additionally assigned the enhanced selectivity for ethylene/alcohols production displayed
by these catalysts to the presence of remnant oxide within their (sub-)surface.[9,17-26]
The validation of these conflicting lines of thought is complicated by the high surface roughness
0
intrinsic to these Cu
2
O catalysts, since their eventual reduction to Cu would translate into an
abundance of [100](-stepped) analogous domains[27-29] and/or in a high surface pH[28-31] that would
This article is protected by copyright. All rights reserved.
3

ChemPhysChem
10.1002/cphc.201900468
justify the CO
2
-to-ethylene selectivity displayed by these materials. Additionally, their
correspondingly large content of surface grain boundaries prone to re-oxidation upon contact with
trace amounts of oxygen[13,16] (e.g., during sample transfer for post-mortem characterization) could
cause the apparent presence of remaining (sub-)surface oxides asserted in the above studies.[9,17−26]
Thus, to suppress the ambiguous effects of the surface roughness on the oxide’s selectivity and
post-mortem oxidation state, in this work we prepared quasi-planar, low-roughness Cu
2
O thin
films (TFs) and used them as CO -reduction electrocatalysts. As we will show in the following,
2
these TFs displayed a product selectivity similar to that of oxide-derive copper catalysts,[31] and
post-mortem characterization through XPS (implementing an air-free transfer) confirmed their
0
complete reduction to Cu . Thus, our results indicate that the unique ethylene selectivity of Cu-
oxides is a result of their enhanced roughness, and suggest that the often-reported presence of
(sub-)surface oxides in these materials may stem from their re-oxidation upon air contact during
transfer for (or in the course of) post-mortem characterization.
Experimental Section
Thin film sample preparation.− Copper oxide thin films were fabricated by reactive DC magnetron
sputter deposition according to the procedure described in detail in Ref. 32. The target was a
metallic Cu disk (99.99 wt. % in purity − Umicore AG) with a diameter of 75 mm and 6 mm in
thickness. The substrate for electrochemical investigations was a silicon wafer (Si(100), type n, P-
doped − PB-Technik AG). A ≈ 4 nm thick Ti interlayer was first deposited by DC magnetron
sputtering, followed by a ≈ 20 nm Ag interlayer used to improve the adhesion of the final Cu
2
O
thin film (TF) on the substrate. Discharge was conducted at 200 W and a back-pressure of 5∙10-7
This article is protected by copyright. All rights reserved.
4

ChemPhysChem
10.1002/cphc.201900468
mbar attained by dosing a 50:50 mixture of O
2
in Ar. Samples with two different film thicknesses
(≈ 100 nm vs. ≈ 1 μm) were obtained by controlling the deposition time. The thickness of the
resulting films was determined using a stylus profilometer (Veeco Dektak-8).
Materials.− Potassium chloride (≥ 99.999 %, Merck), caesium carbonate and potassium
bicarbonate (≥ 99.999 vs. ≥ 99.95 %, respectively − Sigma Aldrich) were used without further
purification. All electrolytes were prepared by dilution of the above salts in ultrapure water
(
18.2 MΩ∙cm, Elga Purelab Ultra). Copper foil (≥ 99.999 % metals basis, 1.00 mm in thickness)
was purchased from Goodfellow Cambridge Ltd. Fine polishing MicroPolish Alumina suspension
0.05-0.3 μm) and polishing cloth (MasterTex, PSA, 8 IN) were purchased from Buehler. Carbon
(
dioxide, nitrogen, argon and hydrogen gases (all 6.0 quality) were purchased from Carbagas.
Preparation of the polycrystalline copper electrode.− For the preparation of the polycrystalline
2
copper benchmark electrode (Cupoly), Cu-foil was cut into 2.5 × 2.5 cm pieces, polished with
alumina suspension down to 0.05 μm on a polishing cloth, and subsequently cleaned by sonication
in 2-propanol (99.9 %, Chromasolv Plus for HPLC, Sigma Aldrich) and ultrapure water (18.2
MΩ∙cm, Elga Purelab Ultra) for 30 min. Finally, the copper foil was electropolished in
concentrated phosphoric acid (≥ 85 wt. % in H O, 99.999+ % metals basis, Sigma Aldrich) for 5
2
min at a potential of 2.0 V vs. a copper foil counter electrode, followed by rinsing in ultrapure
water and drying under nitrogen.
CO
in 0.1 M CsHCO
CO for 1 h at 1 bar until a pH of 6.8 was attained. CO
pressure in a gas-tight custom-made parallel plate electrochemical cell made of polyether ether
2
-Electroreduction measurements.− All electrochemical CO
electrolyte solution, which was obtained by bubbling 0.05 M Cs
electrolysis was performed at ambient
2
-reduction tests were performed
3
2
CO with pure
3
2
2
This article is protected by copyright. All rights reserved.
5

ChemPhysChem
10.1002/cphc.201900468
ketone (PEEK) and fitted with o-rings.[33,34] The cell configuration is designed to ensure a uniform
potential distribution across the surface of the working electrode by placing the counter electrode
(
platinum foil) parallel to the working electrode.[33,34] The geometric surface area of both
2
electrodes was 1.00 cm . To separate anodic and cathodic compartments and prevent the oxidation
®
of CO
2
-reduction products, an anion exchange membrane (Selemion AMV, AGC Inc.) was
placed in between. Both anodic and cathodic compartments contained small electrolyte volumes
2 vs. 1.8 mL, respectively) to ensure high concentrations of liquid products and therefore decrease
(
the detection limits. An Ag/AgCl electrode (LF-1, Harvard Apparatus GmbH) was used as the
reference. Data were converted to a reversible hydrogen electrode (RHE) potential (E) scale using
the Equation:
E (V vs. RHE) = E (V vs. Ag/AgCl) + 0.197 V + pH ∙ 0.0591 V
where pH refers to the pH of the CO
Before conducting CO electrolysis, the electrolyte was purged with CO
ensure complete CO saturation. During electrolysis, CO was constantly bubbled through the
2
-saturated electrolyte solution (i.e., pH = 6.8 – see above).
2
2
for at least 15 min to
2
2
electrolyte at a flow rate of 5 sccm controlled by a mass flow controller (Bronkhorst). For each
measurement the solution resistance was determined using electrochemical impedance
spectroscopy at the open circuit potential applying a 10 mV perturbation (1 MHz to 10 Hz). Before
the cyclic voltammetry (CV) and chronoamperometry (CA) measurements, the potentials were
partially corrected for the corresponding IR-drop by setting the Biologic SP-300 potentiostat to
8
5 % software compensation.
Product analysis.− Gas products analysis was carried out with a gas chromatograph (GC − 8610C,
SRI Instruments) equipped with a packed HaySep D column and a packed MolSieve 5A column.
This article is protected by copyright. All rights reserved.
6

ChemPhysChem
10.1002/cphc.201900468
A flame ionization detector (FID) equipped with a methanizer was used to detect hydrocarbons
with argon as the carrier gas. A thermal conductivity detector (TCD) was used to detect hydrogen
with argon as the carrier gas. Calibration of the GC was carried out using four calibration gas
mixtures in argon prepared by Carbagas. Online gaseous product GC-analysis was carried out after
CO passed through the cell directly into the gas sampling loop. For all experiments, gas analysis
2
was performed after 30 min of continuous electrolysis to obtain the faradaic efficiencies of the
various products.
Charge-neutral liquid products (e.g., ethanol) were analyzed by collecting aliquots of the
electrolyte after electrolysis from the cathode chamber, using the GC’s split/splitless injector and
Innopeg capillary column. On the other hand, formate was analyzed on diluted electrolyte samples
using ionic chromatography (IC, Metrohm 882 Compact IC plus and 863 Compact Autosampler)
with a detection limit of < 1 ppb. Liquid products’ calibration of the GC and IC devices was carried
out using calibration liquid mixtures with products concentrations ranging from 0.1 to 10 or 100
to 10,000 ppm for IC vs. GC, respectively.
Air-free transfer of post-CA samples to ultra-high vacuum (UHV).− To avoid contact with air, a
custom-made, gas-tight chamber allowing sample transfer without air intrusion[35] and the
electrochemical cell described above were placed inside a glovebox (SICCO V 1982-08) that was
constantly purged with N
2
and equipped with an O
2
-sensor (O2-BTA by Vernier Software &
Technology). Once the O
2
-content inside the glovebox reached a value ≤ 0.2 vol. %, the
electrochemical test was performed and, with the glovebox still closed, the cell was disassembled
and the sample was recovered and transferred inside the gas-tight transfer chamber. The latter was
then closed, taken out of the glovebox and directly transported and quickly mounted (in < 2 min)
in the transfer port of the UHV-setup used for Ar-sputtering and X-ray photoelectron spectroscopy
This article is protected by copyright. All rights reserved.
7

ChemPhysChem
10.1002/cphc.201900468
−
4
(
cf. below), where the transfer chamber was opened only once a back-pressure ≤ 10 mbar was
reached.
−9
Ultra-high vacuum characterization.− XPS measurements were performed at ≈ 2∙10 mbar in a
VG ESCALAB 220iXL spectrometer (Thermo Fisher Scientific) using focused monochromatized
2
Al K
α
radiation (1486.6 eV) with a beam size of ≈ 500 μm . The spectrometer was pre-calibrated
by performing measurement on a clean silver surface, whereby the Ag 3 d5/2 peak was aligned to
a binding energy of 368.25 eV with a full width at half-maximum of 0.78 eV at a pass energy of
3
0 eV. All survey spectra were recorded with a dwell time of 50 ms, using a pass energy of 50 eV
and in 0.5 eV binding energy steps; for spectra acquired in a narrow energy span, these last
parameters accounted to 30 and 0.05 eV, respectively. All recorded spectra were energy-calibrated
by aligning the Cu 2p3/2 peak to a binding energy of 932.5 eV.[36]
Argon sputtering was performed in a separated chamber directly connected in UHV to the XPS
−
9
station and with a back pressure of ≈ 2∙10 . A leak valve (Kurt J. Lesker Company) was used to
-6
dose Ar (6.0 quality – Carbagas AG) until a pressure of ≈ 9.10 mbar was reached. Sputtering for
the stipulated time was then completed using an IQE 11/35 ion source (SPECS GmbH) operated
at a voltage of 3 kV.
X-ray diffraction (XRD).− Grazing incidence XRD measurements were performed using a
Smartlab Rigaku diffractometer set in parallel beam mode. The diffractometer’s Cu K
ɑ
X-ray
source was operated at 200 mA and 45 kV and placed at 0.3 ° with respect to the sample’s surface.
A graphite Göbel mirror was used to make the beam parallel and remove the Cu-K radiation. A 2
β
2
x 0.3 mm spot size was projected on the sample using a 2 mm slit placed after the Göbel mirror.
This article is protected by copyright. All rights reserved.
8

ChemPhysChem
10.1002/cphc.201900468
The data were acquired between 2θ values of 20 and 90 °, moving the SC70 detector at a rate of
−
1
1
° min .
Scanning electron microscopy (SEM).− SEM measurements were performed in a Carl Zeiss
Ultra55 scanning electron microscope, using the in-lens detector at an acceleration voltage of 3 kV
and a working distance of ≈ 4.5 mm.
Results and Discussion
Initial properties and CO
2
-electroreduction performance of Cu
2
O thin films.− Cuprous oxide TFs
≈
100 nm in thickness were deposited on Si(100) wafers with a ≈ 20 nm thick Ag adhesion
interlayer using magnetron sputtering (cf. Experimental Section for details). As revealed by the
SEM image in Figure 1A, the resulting samples displayed a low roughness surface topography
convenient to our purposes. To quantify this asseveration, pieces of the silicon wafer covered with
the Cu
2
O TF were cut and mounted in a three-electrode electrochemical cell[33] filled with CO
2
-
saturated 0.05 M Cs
2
CO , and were submitted to cyclic voltammetry (CV) measurements at
3
−
2
various scan rates, from which a double layer capacitance of ≈ 16 μF∙cmgeom (at 0.4 VRHE) was
estimated (cf. Figs. 2A and 2C). Assuming the conversion capacitance of ≈ 30 μF∙cmsurface−2
customarily used to estimate the roughness factor (RF) of Cu-based catalysts in this potential
window[29,31] would lead to a below-unit RF of only ≈ 0.5 cmsurface2∙cmgeom−2. On the other hand,
this seemingly low roughness may be affected by imprecisions in the above normalization
capacitance,[29,31] which stems from measurements on reduced, Cu samples with an oxidation state
0
different from that of the initially oxidized Cu
2
O TF used herein. On the basis of observations for
reduced vs. oxidized Pt, Ir and Ru surfaces,[37,38] surface oxidation reportedly leads to a ≈ 2-fold
This article is protected by copyright. All rights reserved.
9

ChemPhysChem
10.1002/cphc.201900468
decrease of the intrinsic double layer capacitance; this would in terms translate in a lower
conversion capacitance (i.e., ≈ 15 μF∙cmsurface−2) leading to an RF closer to the unit value expected
for the quasi-flat surface observed by SEM.
Figure 1. Scanning electron micrographs of the as-deposited, ≈ 100 nm thick Cu
same sample after 30 minutes of chronoamperometric CO -reduction at – 1.0 VRHE in CO
Cs CO (B).
2
O thin film (A) and of the
2
2
-saturated 0.05 M
2
3
Having assessed the TF’s initial topography and roughness, grazing angle XRD measurements
confirmed the exclusive presence of crystalline Cu O domains in its bulk (cf. Fig. 3A). This
2
characterization was complemented by XPS measurements, whereby the survey scan in Figure S1
confirmed the exclusive presence of copper, oxygen and adventitious carbon on the TF surface.
Most importantly, the Cu L
3
M
4,5
M4,5 Auger spectrum in Figure 3B consisted of a broad peak
located at a binding energy of ≈ 570 eV (corresponding to a kinetic energy of ≈ 916.5 eV).
Interestingly, this energy position is precisely in-between the values reported for Cu O and CuO
2
This article is protected by copyright. All rights reserved.
1
0

ChemPhysChem
10.1002/cphc.201900468
(
at kinetic energies of ≈ 916 vs. ≈ 917 eV, respectively),[39] and therefore suggests the simultaneous
presence of both oxides on the film’s surface. This was confirmed by the satellite peaks observed
at binding energies of ≈ 935, ≈ 945, ≈ 963 and ≈ 975 eV in the corresponding core level Cu 2p
XPS displayed in Figure S2, which are assignable to CuO.[40]
Figure 2. Cyclic voltammograms recorded at various scan rates on the ≈ 100 nm thick Cu
2
O thin film prior
to and after 30 minutes of chronoamperometry (CA) at – 1.0 VRHE in CO -saturated 0.05 M Cs
2
2
CO (A vs.
3
B). Corresponding plots of the currents at 0.4 VRHE as a function of the scan rate (C), whereby the slopes of
the linear fits lead to pre- vs. post-CA capacitance values of ≈ 16 vs. ≈ 480 μF∙cmgeom-2, respectively.
This article is protected by copyright. All rights reserved.
1
1

ChemPhysChem
10.1002/cphc.201900468
Figure 3. Grazing angle X-ray diffractograms (A) of the ≈ 100 nm thick Cu
2
O thin film as-prepared and
after 30 minutes of chronoamperometry (CA) at – 1.0 VRHE in CO
2
-saturated 0.05 M Cs
2
CO (‘Before CA’
3
0
vs. ‘After CA’, respectively), along with reference diffraction peaks of Cu
numbers 1000063 and 4105040). Note that the intensities of the Cu
77 ° have been multiplied 5- or 10-fold to allow for a clearer assignment. Corresponding Cu L
Auger spectra (B), whereby post-CA samples were transferred to ultra-high vacuum for XPS analysis in
contact with air (‘After CA, air’) or by disassembly of the electrochemical cell inside a N -filled glovebox
and transport in a gas-tight transfer chamber (‘After CA, N ’). Comparison among the Cu L
2
O and Cu standards (COD card
2
O diffraction peaks at ≈ 30, ≈ 53 and
≈
3
M
4,5
M
4,5
2
2
3
4,5
M M4,5 Auger
spectra of the post-CA sample transferred in the absence of air (‘After CA, N ’) and of Ar-sputtered
2
polycrystalline copper foil (Cupoly), along with the difference of both spectra (C).
This article is protected by copyright. All rights reserved.
1
2

ChemPhysChem
10.1002/cphc.201900468
Following this initial characterization, the CO
Cu
CV measurements and filled with CO
2
-electroreduction performance of this ≈ 100 nm
O/Si(100) TF-wafers was determined in the same electrochemical cell[33] used for the above
-saturated 0.05 M Cs CO (if connected to a gas
2
2
2
3
chromatography for products’ analysis − see above Experimental Section for details). Figure 4A
displays the Faradic efficiencies (FEs) obtained in chronoamperometric (CA) tests at potentials
ranging between − 0.8 and − 1.1 VRHE (see Figs. S3 and S4 for the corresponding, overall and
product-specific currents, respectively). Upon comparison with previous results using the same
electrolyte and cell configuration,[31] the TF’s limited efficiency for ethylene production (maxing
at ≈ 20 %) and relatively large H
2
-yields (≈ 40 to ≈ 60 %) are comparable to those observed for a
thermally-produced copper oxide (cf. the OD-Cu sample in Ref. 31) − a consistency that also
extends to the minor selectivities for ethanol, CH and CO, but that does not apply to the TF’s
4
−
increase in HCOO -yield with potential, more similar to the trend reported for Cupoly in the same
study.[31] These difference among samples are highlighted in Figure 4B, which summarizes the
FEs at − 1 VRHE displayed by the Cu
latter, polycrystalline sample is in very good agreement with the results in Ref. 31 (i.e., using the
same electrolyte and cell configuration). In the case of the Cu O TF, though, the observed ethylene
and methane yields are comparatively lower than on Cupoly, and appear to be offset by a greater
2
O TF and Cupoly; chiefly, the product distribution for the
2
selectivity for H
2
-evolution.
This article is protected by copyright. All rights reserved.
1
3

ChemPhysChem
10.1002/cphc.201900468
Figure 4. Potential-dependent Faradaic efficiencies (FEs) of the ≈ 100 nm Cu
0 minute long chronoamperometric tests in CO -saturated 0.05 M Cs CO
efficiencies at – 1.0 VRHE for the ≈ 100 nm cuprous oxide thin film (‘Cu O TF’) and a polycrystalline copper
‘Cupoly’) sample (B).
2
O thin film in independent,
3
2
2
3
(A). Comparison of the Faradic
2
(
Post-CA roughness and bulk composition of the Cu
2
O thin films.− To shed light on the reasons for
O TF sample after CA at − 1 VRHE was
this product selectivity, the surface topography of a Cu
2
inspected with SEM; the corresponding images displayed in Figure 1B unveil the appearance of
numerous cracks that were not present on a Cupoly sample submitted to the same electrochemical
protocol (cf. Fig. S5). Interestingly, this surface morphology change at CO
2
-reduction potentials
is qualitatively consistent with previous observations for Cu-catalysts,[7,8,10,41,42] and has recently
been suggested to be triggered by surface-adsorbed CO produced as a CO -reduction
2
intermediate.[43] Moreover, the specific appearance of cracks in the TF is likely related to the
0
reduction of the initial Cu
2
O to Cu in the course of CO
2
-electroreduction (cf. discussion below),
This article is protected by copyright. All rights reserved.
1
4

ChemPhysChem
10.1002/cphc.201900468
which would imply a ≈ 15 % contraction of the TF’s lattice (on the basis of the face centered cubic
0
[44,45]
lattice constants of 4.27 vs. 3.62 Å reported for Cu
2
O vs. Cu , respectively)
that would trigger
the film’s in-plane fracture. Most importantly, post-CA CV measurements equivalent to those
performed on the as-prepared TF yielded a double layer capacitance value of ≈ 480 μF∙cmgeom−2
cf. Figs. 2B and 2C) that translates into a roughness factor of ≈ 16 cmsurface2∙cmgeom when
−2
(
−
2
0
considering the conversion capacitance of ≈ 30 μF∙cmsurface quoted above for Cu (i.e., relying on
the reduction of the TF’s surface – vide infra). Interestingly, this post-CA RF value is comparable
to those reported in Ref. 31 for a series of thermally- or electrochemically-prepared oxide-derived
copper catalysts, but remains ≈ 2- to ≈ 6-fold below the (post-CA) factors determined for plasma-
activated Cu materials.[18,26] Moreover, this relatively large roughness may explain the TF’s
enhanced H -evolution activity, concomitant with the selectivity for this by-product displayed by
2
0
rough Cu -samples with an abundance of low-coordinated surface sites (i.e., edges, corners and
kinks).[46-49] Finally, this operando enhancement of the film’s RF also undermined our original
purpose to decouple the effect of this variable and the (sub)surface’s oxidation state on the
catalyst’s selectivity; nevertheless, as it will be shown below, the subsequent characterization of
the post-CA TF did serve to unambiguously assess the (sub)surface’s state following CO
2
-
reduction.
In this regards, our first effort to verify the possible reduction of the cuprous oxide TF upon CO
2
-
electroreduction consisted of characterizing the post-CA TF again through grazing angle XRD.
The resulting diffractogram, plotted in Figure 3A, discloses that the initial Cu O did indeed get
2
0
reduced to Cu in the course of the CO
2
-reduction process; however, the low-intensity, broad peak
O in this post-mortem sample, either due to
the incomplete electrochemical reduction of the initial oxide and/or due to its post-CA oxidation
at a 2θ-value of ≈ 37 ° indicates the presence of Cu
2
This article is protected by copyright. All rights reserved.
1
5

ChemPhysChem
10.1002/cphc.201900468
upon contact with air in the course of the XRD measurement (lasting 70 min – see Experimental
Section for details).
Post-CA TF oxidation state assessed through (depth profiling) XPS.− Trying to discriminate these
effects with the aid of XPS, an equivalent Cu
was rapidly disassembled from the electrochemical cell and transferred in ≈ 2 min to ultra-high
vacuum (UHV) for analysis. The corresponding Cu L 4,5 Auger and 2p spectra displayed in
Figures 3B and S1 confirm the presence of copper oxide on this post-CA TF; specifically, the
2
O TF sample (i.e., having undergone CA at – 1 VRHE)
3
M M
4,5
Cu L
3
M
4,5
M
4,5 Auger in Figure 3B (more sensitive than core level Cu 2p spectra for discriminating
among Cu
2
O and Cu0)[39,40] featured the general shape observed for the initial Cu
2
O TF, along with
0
two shoulders at binding energies of ≈ 568 and ≈ 565 eV assignable to Cu . Additionally, the Cu
2
p spectrum featured in Figure S2 still displayed small contributions of the CuO-satellites
discussed above.
While these results may appear to suggest the incomplete reduction of the initial Cu O TF in the
2
course of the CO
the sample during cell disassembly and/or upon fast transfer to UHV, as previously proven in Ref.
6; thus, an equivalent TF was prepared keeping the electrochemical cell inside an N -filled
2
-reduction reaction, they could still be affected by the air-induced oxidation of
3
2
glovebox, whereby the post-CA sample was mounted inside a portable, gas-tight transfer chamber
that has previously been validated for the XPS-study of air-sensitive battery materials (see
Experimental Section for details).[35] The survey scan acquired on this air-free transferred sample
and displayed in Figure S1 only features peaks assignable to Cu, O, and adventitious carbon, along
with a minor Cs contribution arising from the supporting electrolyte used in the electrochemical
CO
2
-reduction measurement (vide infra). Most importantly, the Cu 2p XPS and L
3
M
4,5
M4,5 Auger
spectra following CA and this air-free transfer (cf. Figs. S2 and 3B, respectively) do not display
This article is protected by copyright. All rights reserved.
1
6

ChemPhysChem
10.1002/cphc.201900468
0
any signs of the oxides discussed above, and instead match the spectrum of Cu reported in
previous literature.[21,36,39] To further verify this assignment, Figure 3C displays a comparison
between this post-CA, air-free transfer spectrum and the one recorded on a sputter-cleaned Cupoly
reference sample; chiefly, the good overlay among spectra (only displaying minor background
differences that cannot be assigned to remnant oxide) unambiguously confirms the complete
reduction of the sample’s surface in the course of CO -electroreduction.
2
Motivated by these results, we proceeded to Ar-sputter this post-CA TF as to study the possible
presence of Cu-oxides in its sub-surface, which has recently been linked to the distinct product
selectivity of oxide (derived) CO
-reduction electrocatalysts.[17-26] The corresponding XPS spectra
2
appear plotted in Figure 5; sputtering for only 1 min led to the complete disappearance of the Cs 3d
and O 1s peaks initially present on the as-transferred surface and related to deposits of the
electrolyte salt,[36] while no other changes were observed after an additional 4 min of sputtering.
When the cumulative sputtering time reached 15 min, though, the recorded spectrum displayed an
additional couple of peaks at ≈ 370 eV that can be assigned to the TF’s Ag adhesion interlayer,[50]
indicating sputtering along the complete TF thickness and proving the negligible presence of
oxygen within the sample (i.e., below the XPS’s detection limit of ≈ 1 at. %). This asseveration is
fully consistent with the corresponding Cu L
3
M M4,5 Auger spectra (see inset in Fig. 5), which
4,5
are identical for all sputtering durations and the as-transferred (i.e., non-sputtered) sample and
0
therefore confirm the TF’s complete reduction to Cu along the film’s bulk thickness during CO
2
-
electroreduction.
Chiefly, these observations were excellently reproduced on a thicker, ≈ 1 μm Cu
displayed a CO -reduction product selectivity similar to that of the ≈ 100 nm sample (cf. Fig. S6)
and matching properties upon UHV characterization (see XPS spectra in Fig. S7). This lack of
2
O TF, which
2
This article is protected by copyright. All rights reserved.
1
7

ChemPhysChem
10.1002/cphc.201900468
sensitivity of the products’ distribution to the thickness of the oxide layer is further proof of the
negligible role of this oxide phase on the CO
2
-reduction mechanism,[13] and is consistent with
previous online electrochemical[51] and O secondary ion mass spectrometry studies that have
18
[15]
suggested that the latter reaction is only triggered once the oxide is reduced to Cu0.
Figure 5. Depth-profiling X-ray photoelectron spectra of the ≈ 100 nm Cu
minutes of chronoamperometry at – 1.0 VRHE in CO -saturated 0.05 M Cs CO
the electrochemical cell in a N -filled glovebox and transport to ultra-high vacuum in a gas-tight transfer
chamber (cf. ‘After CA, N ’ sample in Figures 3B and 3C). XP spectra were recorded just after transfer (‘0
min’) or following Ar-sputtering (at ≈ 9.10 mbar and 3 kV) for 1, 5 or 15 min. The inset displays the
corresponding Cu L 4,5 Auger spectra. Note the presence of Cs and O signals exclusively on the
2
O thin film subjected to 30
2
2
3
, following disassembly of
2
2
−
6
M M
3 4,5
surface of the as-transferred, non-sputtered sample (‘0 min’), as well as the detection of the TF’s Ag
adhesion layer after 15 min of sputtering.
This article is protected by copyright. All rights reserved.
1
8

ChemPhysChem
10.1002/cphc.201900468
On the other hand, as highlighted in Table 1, our observations are in stark disagreement with
numerous studies postulating the presence of residual Cu-oxide within the catalyst’s (sub-)surface
in the course of CO
-electroreduction.[9,17-26] While we cannot pinpoint the specific cause(s) that
2
may have led to this inconsistency, the post-mortem samples in Refs. 9 and 17 were reportedly
exposed to air (and thus oxidized) prior to their characterization. Contrarily, previous post-mortem
XPS[21] and EELS[22,23] measurements seemingly implementing minimized air contact upon
sample transfer still reported a partial oxidation of the catalysts following CO -electroreduction.
2
Assuming a comparable level of exposure to residual O
whereby Cu
2
in those studies[21−23] and this work,
0
2
O TFs are univocally proven to get reduced to Cu , the partial oxidation of the post-
mortem samples reported in Refs. 21−23 may be tentatively assigned to their enhanced
reactivity[13,16] with residual O
2
. Such a behavior could in terms be caused by the extremely large
surface roughness and corresponding abundance of highly reactive, low-coordinated surface sites
in those materials (featuring RFs ≈ 2- to ≈ 6-fold larger than that of the post-CA TF).[16,18,26]
Nevertheless, this relation among surface oxidation and roughness factor does not appear to be
univocal; specifically, as showcased in Table 1, samples with a roughness factor as high as
44 cmsurface2∙cmgeom have been reported to be absent of sub-surface oxides following CO
-
−2
≈
2
electroreduction (see Ref. 24), whereas the latter oxide was detected in the subsurface of catalysts
with similar or even lower RFs.21,26 Thus, further efforts should be devoted to clarify the effect of
2
5
24
other parameters (like the catalyst support or the addition of anionic species to the electrolyte )
on the reported presence of these (sub-)surface Cu-oxides, ideally minimizing possible artifacts
introduced upon sample transfer by substituting post-mortem characterization techniques with in
situ / operando ones.
This article is protected by copyright. All rights reserved.
1
9

ChemPhysChem
10.1002/cphc.201900468
Table 1. Summary of qualitative literature results (including those derived from this study) regarding the
detection of Cu-oxides in the catalyst’s (sub)surface in the course of or just after electrochemical CO
2
-
reduction, as derived from in situ or post-mortem characterization techniques and specifying whether
sample transfer involved contact with air for the latter kind. Note that for all fields the dash symbol (‘−’)
refers to information that was not specified in the cited manuscript, whereas ‘NA’ stands for ‘not applicable’
(
i.e., as in the case of air contact upon in situ characterization), and that for selected electrochemical (EC)
protocols and transfer conditions the relevant technique is reported between parenthesis.
Roughness factor
[ cmsurface ∙cmgeom
Before EC After EC
Air-contact
upon
transfer*
Characterization
technique
(Sub)surf.
Cu-oxide†
2
−2
Catalyst
EC protocol
]
Ref.
1
0 min to 2 h @
Post-mortem,
depth-profiling
XPS and AES
In.situ Raman
spec.
Cu
2
O NPs on C-
−
1.9 V vs.
2.4
.7
−
−
−
Yes
NA
NA
Yes
No
No
9
based GDE
Ag/AgCl
1
.7 μm thick
200 s @ − 0.99
V vs. RHE
Potential scan
from 0.65 to
4
10
14
2
Cu O film on Cu
Polycrystalline
Cu w/ surface
oxide layer
−
In situ GI-XRD
– 0.3 V vs. RHE
1
0 s @ − 1.0
In situ Raman
spec., post-
mortem XRD
Various Cu
nanomaterials
2
O/Cu
NA (Raman)
Yes (XRD)
−2
mA∙cm
Raman)
−
−
No
15
(
18O-enriched
1
–
0 min to 5 h @
1.0 V vs. RHE
Post-mortem
Cu
nanocubes
Cu O/Cu
2
O/Cu
22
−
21
−
Yes
Yes#
No
Yes
Yes
16
17
18
SIMS
2
5 h @ – 1.6 V
Post-mortem
XPS and XAS
Post-mortem
EDS, in situ
XAS
nanoparticles
Various plasma-
treated Cu-
electrodes
Thermally
oxidized Cu(111)
Plasma-treated Cu
nanocubes
vs. RHE
1
h @ – 0.91 V
vs. RHE (EDS)
≤ 90
−
−
Yes (EDS)
NA
NA§
−
−
In situ XPS
Yes
Yes
19
21
1 h @ – 1.0 V
Post-mortem
XPS and EDS
No (XPS)
Yes (EDS)
≤
36
vs. RHE
Electrochemically E-cycling down
oxidized,
polycrystalline Cu
Electrochemically
oxidized Cu
nanoparticles
Various plasma-
treated Cu-
to – 0.73 V vs.
RHE
55 min @
– 1.15 V vs.
RHE
−
−
−
EELS
No
No
Yes
Yes
22
23
24
25
26
−
STEM-EELS
1
h @ – 1.0 V
Post-mortem
XPS and EDS
No (XPS)
Yes (EDS)
≤ 44
−
−
No
vs. RHE
electrodes
Post-mortem
XPS, in situ
XAS
Electrodeposited
Cu-nanocubes
1 h @ – 1.1 V
−
No (XPS)
Yes/NoΩ
Yes
vs. RHE
Various plasma-
treated Cu-
1
h @ – 1.0 V
Post-mortem
No (XPS)
Yes (EDS)
−
≤ 44Φ
vs. RHE
XPS,
electrodes
This article is protected by copyright. All rights reserved.
2
0

ChemPhysChem
10.1002/cphc.201900468
3
0 min @
Post-mortem,
depth-profiling
XPS
This
work
Cu
2
O thin films
– 1.0 V vs.
≈ 1
≈ 16
No
No
RHE
*
Specifies whether the sample experienced contact with air upon transfer from the electrochemical test to
†
post-mortem characterization (i.e., it does not apply to in situ measurements). Refers to the qualitative
observation of copper oxide in the catalyst’s (sub)surface upon post-mortem or in situ characterization.
#Unclear whether the XAS-measurements were truly in situ or performed just after CO
-reduction (i.e.,
2
§
post-mortem); in the latter case the samples were reportedly dried with an air-gun. In situ near ambient
pressure XPS measurements at a CO
2
(and in some cases H O) pressure of 0.7 Torr, but without potential
2
Ω
control. Positive detection of (sub)surface oxide reported for nanocubes supported on Cu-foil, whereas no
Φ
oxide was detected when the support was a carbon paper. Sample-dependent values reported after 1 h of
-reduction at −1.0 V vs RHE, whereby the roughness of a sample with ≈ 25 cmsurf2∙cmgeom after 1 h
−2
CO
2
reportedly decreased to ≈ 12 cmsurf2∙cmgeom after 30 h of reaction.
−2
Conclusions
In summary, when implementing proper means of sample transfer, our XPS results monitoring the
Cu Auger spectra of post-mortem, low roughness TF electrodes provide clear proof of the complete
reduction of the initial Cu
2
O surface and bulk in the course of CO -reduction. As such, our
2
conclusions agree with those of recent computational[13] and experimental[15,16,31] studies by the
Ager and Bell groups, and indirectly suggest that the product selectivity of oxide-derived Cu-
catalysts is linked to their great roughness and corresponding abundance of ethylene-selective
surface domains. Ultimately, our work additionally highlights the importance of careful sample
transfer for drawing meaningful conclusions upon post-mortem surface characterization using ex
[2]
situ techniques, as well as the need for a continued development of operando characterization
tools that can circumvent these limitations.[52]
This article is protected by copyright. All rights reserved.
2
1

ChemPhysChem
10.1002/cphc.201900468
Acknowledgments
This work was supported by the Competence Center Energy & Mobility Switzerland (CCEM-CH),
Swiss Electric Research, Innosuisse and the Swiss Competence Center for Energy Research Heat
&
Electricity Storage (SCCER-HaE). The authors are thankful to Mr. Christian Marmy and Mr.
Adrian Heinritz for technical support. AAP greatly acknowledges the Danish Agency for Science,
Technology and Innovation for financial support through the CITRIS visiting scholarship at
University of California, Berkeley, and expresses her gratitude to Prof. Alexis T. Bell, Dr. Peter
Lobaccaro and Dr. Ezra L. Clark for sharing their knowledge on CO reduction and the parallel
2
plate electrochemical cell. JH greatly acknowledges the Swiss National Science Foundation for
financial support through Ambizione Energy grant PZENP2_173632. TJS thanks Innosuisse and
the Swiss Competence Center for Energy Research Heat & Electricity Storage.
References
[
[
1] D. T. Whipple, P. J. A. Kenis, J. Phys. Chem. Lett. 2010, 1, 3451–3458.
2] J. Herranz, J. Durst, E. Fabbri, A. Patru, X. Cheng, A. A. Permyakova, T. J. Schmidt, Nano
Energy 2016, 29, 4–28.
[
[
3] S. Verma, B. Kim, H.-R. Jhong, S. Ma, P. J. A. Kenis, ChemSusChem 2016, 9, 1972–1979.
4] J. Durst, A. Rudnev, A. Dutta, Y. Fu, J. Herranz, V. Kaliginedi, A. Kuzume, A. A.
Permyakova, Y. Paratcha, P. Broekmann, T. J. Schmidt. Chimia 2015, 69, 769−776.
[
5] A. Patru, T. Binninger, B. Pribyl, T. J. Schmidt, J. Electrochem. Soc. 2019, 166, F34−F43.
This article is protected by copyright. All rights reserved.
2
2

ChemPhysChem
10.1002/cphc.201900468
[
[
6] K. W. Frese, J. Electrochem. Soc. 1991, 138, 3338−3344.
7] M. Le, M. Ren, Z. Zhang, P. T. Sprunger, R. L. Kurtz, J. C. Flake, J. Electrochem. Soc. 2011,
1
58, E45−E49.
[
[
8] C. W. Li, M. W. Kanan, J. Am. Chem. Soc. 2012, 134, 7231−7234.
9] D. Kim, S. Lee, J. D. Ocon, B. Jeong, J. K. Lee, J. Lee, Phys. Chem. Chem. Phys. 2015, 17,
8
24−830.
[
[
10] D. Ren, Y. Deng, A. D. Handoko, C. S. Chen, S. Malkhandi, B. S. Yeo, ACS Catal. 2015, 5,
814-2821.
2
11] N. De Zoubov, C. Vanleugenhaghe, M. Pourbaix in Atlas of Electrochemical Equilibria in
Aqueous Solutions (Ed.: M. Pourbaix), Pergamon Press, Oxford, 1966, pp. 384−392.
[
[
[
12] X. Nie, G. L. Griffin, M. J. Janik, A. Asthagiri, Catal. Commun. 2014, 52, 88-91.
13] A. J. Garza, A. T. Bell, M. Head-Gordon, J. Phys. Chem. Lett. 2018, 9, 601–606.
14] S. B. Scott, T. V. Hogg, A. T. Landers, T. Maagaard, E. Bertheussen, J. C. Lin, R. C. Davis,
J. W. Beeman, D. Higgins, W. S. Drisdell, C. Hahn, A. Mehta, B. Seger, T. F. Jaramillo,
ACS Energy Lett. 2019, 4, 803−804.
[
[
15] L. Mandal, K. R. Yang, M. R. Motapothula, D. Ren, P. Lobaccaro, A. Patra, M. Sherburne,
V. S. Batista, B. S. Yeo, J. W. Ager, J. Martin, T. Venkatesan, ACS Appl. Mater. Interfaces
2
018, 10, 8574−8584.
16] Y. Lum, J. W. Ager, Angew. Chem. Int. Ed. 2018, 57, 551−554.
This article is protected by copyright. All rights reserved.
2
3

ChemPhysChem
10.1002/cphc.201900468
[
[
17] S. Lee, D. Kim, J. Lee, Angew. Chem. Int. Ed. 2015, 127, 14914−14918.
18] H. Mistry, A. S. Varela, C. S. Bonifacio, I. Zegkinoglou, I. Sinev, Y.-W. Choi, K. Kisslinger,
E. A. Stach, J. C. Yang, P. Strasser, B. Roldan Cuenya, Nat. Commun. 2016, 7, 12123.
[
19] M. Favaro, H. Xiao, T. Cheng, W. A. Goddard III, J. Yano, E. J. Crumlin, Proc. Nat. Acc.
Sci. 2017, 114, 6706−6711.
[
[
20] H. Xiao, W. A. Goddard III, T. Cheng, Y. Liu, Proc. Nat. Acc. Sci. 2017, 114, 6685−6688.
21] D. Gao, I. Zegkinoglou, N. J. Divins, F. Scholten, I. Sinev, P. Grosse, B. Roldan Cuenya,
ACS Nano 2017, 11, 4825−4831.
[
[
22] A. Eilert, P. Cavalca, F. S. Roberts, J. Osterwalder, C. Liu, M. Favaro, E. J. Crumlin, H.
Ogasawara, D. Friebel, L. G. M. Pettersson, A. Nilsson, J. Phys. Chem. Lett. 2017, 8,
2
85−290.
23] F. Cavalca, R. Ferragut, S. Aghion, A. Eilert, O. Diaz-Morales, C. Liu, A. L. Koh, T. W.
Hansen, L. G. M. Pettersson, A. Nilsson, J. Phys. Chem. C 2017, 121, 25003−25009.
[
[
24] D. Gao, F. Scholten, B. Roldan Cuenya, ACS Catal. 2017, 7, 5512−5120.
25] P. Grosse, D. Gao, F. Scholten, I. Sinev, H. Mistry, B. Roldan Cuenya, Angew. Chem. Int.
Ed. 2018, 57, 6192−6197.
[
[
26] D. Gao, I. T. McCrum, S. Deo, Y.-W. Choi, F. Scholten, W. Wan, J. G. Chen, M. J. Janik,
B. Roldan Cuenya, ACS Catal. 2018, 8, 10012−10020.
27] Y. Hori, I. Takahashi, O. Koga, N. Hoshi, J. Phys. Chem. B 2002, 106, 15−17.
This article is protected by copyright. All rights reserved.
2
4

ChemPhysChem
10.1002/cphc.201900468
[
28] K. J. P. Schouten, Z. Qin, E. Pérez Gallent, M. T. M. Koper, J. Am. Chem. Soc. 2012, 134,
864−9867.
9
[
[
29] F. S. Roberts, K. P. Kuhl, A. Nilsson, Angew. Chem. Int. Ed. 2015, 54, 5179−5182.
30] R. Kas, R. Kortlever, H. Yilmaz, M. T. M. Koper, G. Mul, ChemElectroChem 2015, 2, 354
–
358.
[
[
[
[
[
31] Y. Ref, B. Yue, P. Lobaccaro, A. T. Bell, J. W. Ager, J. Phys. Chem. C 2017, 121,
4191−14203.
1
32] A. Rabis, D. Kramer, E. Fabbri, M. Worsdale, R. Kötz, T. J. Schmidt, J. Phys. Chem. C
014, 118, 11292−11302.
2
33] K. P. Kuhl, E. R. Cave, D. N. Abram, T. F. Jaramillo, Energy Environ. Sci. 2012, 5,
050−7059.
7
34] P. Lobaccaro, M. R. Singh, E. L. Clark, Y. Kwon, A. T. Bell, J. W. Ager, Phys. Chem. Chem.
Phys. 2016, 18, 26777−26785.
35] G. Ferraresi, C. Villevieille, I. Czekaj, M. Horisberger, P. Novak, M. El Kazzi, ACS Appl.
Mater. Interfaces 2018, 10, 8712−8720.
[
[
36] W. Kautek, J. G. Gordon II, J. Electrochem. Soc. 1990, 137, 2672−2677.
37] W. G. Pell, A. Zolfaghari, B. E. ConwayJ. Electroanal. Chem. 2002, 532, 13−23.
This article is protected by copyright. All rights reserved.
2
5

ChemPhysChem
10.1002/cphc.201900468
[
38] N. Danilovic, R. Subbaraman, K.-C. Chang, S. H. Chang, Y. J. Kang, J. Snyder, A. P.
Paulikas, D. Strmcnik, Y.-T. Kim, D. Myers, V. R. Stamenkovic, N. M. Markovic, J. Phys.
Chem. Lett. 2014, 5, 2474−2478.
[
[
39] D. Tahir, S. Tougaard, J. Phys.: Condens. Matter 2012, 24, 175002.
40] M. C. Biesinger, L. W. M. Lau, A. R. Gerson, R. St. C. Smart, Appl. Surf. Sci. 2010, 257,
8
87−898.
[
41] Y.-G. Kim, J. H. Baricuatro, A. Javier, J. M. Gregoir, M. P. Soriaga, Langmuir 2014, 30,
5053−15056.
1
[
[
42] M. Rahaman, A. Dutta, A. Zanetti, P. Broekmann, ACS Catal. 2017, 7, 7946−7956.
43] C. M. Gunathunge, X. Li, J. Li, V. J. Ovalle, M. M. Waegele, J. Phys. Chem. C 2017, 121,
1
2337−12344.
[
[
[
44] M. E. Straumanis, L. S. Yu, Acta Cryst. 1969, A25, 676−682.
45] M. Heinemann, B. Eifert, C. Heiliger Phys. Rev. B 2013, 87, 115111.
46] D. Reyter, M. Odziemkowski, D. Bélanger, L. Roué, J. Electrochem. Soc. 2007, 154,
K36−K44.
[
[
47] R. Reske, H. Mistry, F. Behafarid, B. Roldan-Cuenya, P. Strasser, J. Am. Chem. Soc. 2014,
1
36, 6978−6986.
48] 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.
This article is protected by copyright. All rights reserved.
2
6

ChemPhysChem
10.1002/cphc.201900468
[
[
[
[
49] J. Huang, N. Hörmann, E. Oveisi, A. Loiudice, G. L. de Gregorio, O. Andreussi, N. Marzari,
R. Buonsanti, Nat. Commun. 2018, 9, 3117.
50] P. Prieto, V. Nistor, K. Nouneh, M. Oyama, M. Abd-Lefdil, R. Díaz, Appl. Surf. Sci. 2012,
2
58, 8807–8813.
51] R. Kas, R. Kortlever, A. Milbrat, M. T. M. Koper, G. Mul, J. Baltrusaitis, Phys. Chem. Chem.
Phys. 2014, 16, 12194−12201.
52] M. Povia, J. Herranz, T. Binninger, M. Nachtegaal, A. Diaz, J. Kohlbecher, D. Abbott, B.-J.
Kim, T. J. Schmidt, ACS Catal. 2018, 8, 7000−7015.
Table of Contents
2
XPS of post-mortem Cu O thin film electrodes characterized following an air-free transfer to ultra-
0
high vacuum unveils their complete transformation into Cu in the course of CO
2
-electroreduction.
Keywords
Catalysis, co-electrolysis, electrochemistry, energy conversion, photoelectron spectroscopy.
This article is protected by copyright. All rights reserved.
2
7