Copper-Based Metal–Organic Porous Materials for CO2 Electrocatalytic Reduction to Alcohols

Article abstract of DOI:10.1002/cssc.201600693

The electrocatalytic reduction of CO₂ has been investigated using four Cu-based metal–organic porous materials supported on gas diffusion electrodes, namely, (1) HKUST-1 metal–organic framework (MOF), [Cu₃(μ₆-C₉H₃O₆)₂]_n; (2) CuAdeAce MOF, [Cu₃(μ₃-C₅H₄N₅)₂]_n; (3) CuDTA mesoporous metal–organic aerogel (MOA), [Cu(μ-C₂H₂N₂S₂)]_n; and (4) CuZnDTA MOA, [Cu_0.6Zn_0.4(μ-C₂H₂N₂S₂)]_n. The electrodes show relatively high surface areas, accessibilities, and exposure of the Cu catalytic centers as well as favorable electrocatalytic CO₂ reduction performance, that is, they have a high efficiency for the production of methanol and ethanol in the liquid phase. The maximum cumulative Faradaic efficiencies for CO₂ conversion at HKUST-1-, CuAdeAce-, CuDTA-, and CuZnDTA-based electrodes are 15.9, 1.2, 6, and 9.9 %, respectively, at a current density of 10 mA cm^?2, an electrolyte-flow/area ratio of 3 mL min cm^?2, and a gas-flow/area ratio of 20 mL min cm^?2. We can correlate these observations with the structural features of the electrodes. Furthermore, HKUST-1- and CuZnDTA-based electrodes show stable electrocatalytic performance for 17 and 12 h, respectively.

Full text of DOI:10.1002/cssc.201600693

DOI: 10.1002/cssc.201600693
Full Papers
Copper-Based Metal–Organic Porous Materials for CO2
Electrocatalytic Reduction to Alcohols
[a]
[b]
[b]
[b]
[a]
Jonathan Albo,* Daniel Vallejo, Garikoitz Beobide, Oscar Castillo, Pedro CastaÇo,
[c]
and Angel Irabien
The electrocatalytic reduction of CO2 has been investigated
using four Cu-based metal–organic porous materials supported
on gas diffusion electrodes, namely, (1) HKUST-1 metal–organic
framework (MOF), [Cu (m -C H O ) ] ; (2) CuAdeAce MOF,
tion of methanol and ethanol in the liquid phase. The maxi-
mum cumulative Faradaic efficiencies for CO2 conversion at
HKUST-1-, CuAdeAce-, CuDTA-, and CuZnDTA-based electrodes
are 15.9, 1.2, 6, and 9.9%, respectively, at a current density of
3
6
9
3
6 2 n
ꢀ
2
ꢀ2
[
Cu (m -C H N ) ] ; (3) CuDTA mesoporous metal–organic aero-
10 mAcm , an electrolyte-flow/area ratio of 3 mLmincm ,
3
3
5
4
5 2 n
ꢀ2
gel (MOA), [Cu(m-C H N S )] ; and (4) CuZnDTA MOA,
and a gas-flow/area ratio of 20 mLmincm . We can correlate
these observations with the structural features of the electro-
des. Furthermore, HKUST-1- and CuZnDTA-based electrodes
show stable electrocatalytic performance for 17 and 12 h, re-
spectively.
2
2
2 2 n
[
Cu Zn (m-C H N S )] . The electrodes show relatively high
0.6 0.4 2 2 2 2 n
surface areas, accessibilities, and exposure of the Cu catalytic
centers as well as favorable electrocatalytic CO reduction per-
2
formance, that is, they have a high efficiency for the produc-
Introduction
The reduction of the CO concentration in the atmosphere has
coupled to a renewable energy source, such as wind or solar
energy, could generate carbon-neutral fuels or industrial chem-
2
become a critical challenge for sustainable development. There
[8]
are different approaches to mitigate CO emissions from the
icals that are derived conventionally from petroleum. In
2
[
1–3]
use of fossil fuels, including CO capture
and subsequent
recent years, many investigators have studied the electrocata-
2
[
4–6]
[9]
storage or its conversion into valuable chemicals.
CO2 is
lytic reduction of CO on metallic and modified electrodes in
2
[10]
considered to be a carbon source for the synthesis of valuable
chemicals, as it is abundant in the atmosphere and nontoxic
aqueous, nonaqueous, and ionic-liquid media under differ-
ent operating conditions and system configurations to pro-
duce a range of useful products for industrial chemistry (i.e.,
formic acid, methane, ethane, ethylene, propylene, methanol,
and ethanol). In particular, the challenges for the conversion of
CO2 into high-energy-density alcohols, such as methanol
compared with the other C source, CO; therefore, the benefi-
1
cial reuse of CO is an interesting approach for the future. To
2
this end, several methods have been adopted for the conver-
sion and activation of CO such as chemical, thermochemical,
2
photochemical, biochemical, electrochemical, and hydrother-
(CH OH), are great, but the potential rewards are even great-
3
[
7]
[8,11,12]
mal methods. Among them, electrocatalytic valorization ap-
pears to be a promising strategy owing to its simple procedure
and ambient operation conditions. In addition, this technology
er.
Among the different cathode metals applied, Cu uniquely
produces hydrocarbons at high reaction rates over sustained
[9,13,14]
periods of time;
therefore, it is the strongest candidate
[
a] J. Albo, P. CastaÇo
for CO2 electrocatalytic reduction. However, Cu generates
a range of reaction products, and the selectivity of each prod-
Department of Chemical Engineering
University of the Basque Country (UPV/EHU)
PO Box. 644, 48080, Bilbao (Spain)
E-mail: jonathan.albo@ehu.es
[15]
uct tends to be low. Thus, to improve the selectivity, other
catalyst structures should be considered to make the electroca-
[
b] D. Vallejo, G. Beobide, O. Castillo
Department of Inorganic Chemistry
University of the Basque Country (UPV/EHU)
PO Box. 644, 48080, Bilbao (Spain)
talytic reduction of CO at Cu-based surfaces technically and
2
economically viable.
In the last decade, metal–organic porous materials (MOPMs)
and, particularly, metal–organic frameworks (MOFs), also
known as porous coordination polymers or porous coordina-
tion networks, have shown many potential applications as new
multifunctional materials. MOFs are hybrid materials containing
three well-differentiated sites to which the catalytic function
can be allocated, namely, the metallic component, the organic
linker, and the pore space. MOFs are considered as ideal candi-
[
c] A. Irabien
Department of Chemical & Biomolecular Engineering
University of Cantabria (UC)
Avda. Los Castros, 39005, Santander (Spain)
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under http://dx.doi.org/10.1002/
cssc.201600693.
This publication is part of a Special Issue around the “1st Carbon Dioxide
Conversion Catalysis” (CDCC-1) conference. A link to the issue’s Table of
Contents will appear here once it is complete.
[
16]
dates for CO adsorption, separation, and reduction through
2
[17]
catalyzed reactions. This is because of their combined favor-
ChemSusChem 2016, 9, 1 – 11
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
These are not the final page numbers! ÞÞ

Full Papers
able properties of large surface area, high porosity, tunable
pore-size, and shape-selective character. Indeed, MOFs are par-
ticularly suitable for electrochemical reactions as a result of
these features in addition to their high electronic conductivi-
To face the challenge of synthesizing effective and stable
CO reduction electrocatalysts for the continuous production
2
of alcohols, in this work, we have evaluated four MOPMs (two
MOFs and two MOAs) containing Cu (Figure 1) as gas diffusion
electrodes (GDEs), which are named hereafter MOPM-GDEs:
(1) a benchmark MOF with formula [Cu (m -C H O ) (OH ) ]
2 3 n
[
18–26]
[19]
ties.
For example, Kumar et al. studied the electrocata-
lytic reduction of CO on Cu-based MOF (HKUST-1) films with
2
3
6
9
3
6 2
an electrolyte consisting of a DMF solution of tetrabutylammo-
(C H O =benzene-1,3,5-tricarboxylate), commonly known as
9 3 6
[32]
nium tetrafluoroborate saturated with CO . Cyclic voltammetry
MOF-199 or HKUST-1, in which the accessible metallic moiet-
2
I
[33]
revealed that the electrochemically generated Cu species were
ies are adsorption sites for CO2;
(2) a microporous
very selective for CO reduction, although the main product
copper(II)–adeninate–acetate coordination framework with for-
2
[
20]
was oxalic acid. In the same year, Hinogami et al. synthesized
a copper rubeanate MOF (CR-MOF) supported on carbon films
mula [Cu (m -adeninate) (m -OOC(CH ) )] ·xH O (CuAdeAce), in
2
3
2
2
3 2
n
2
which the Watson–Crick faces of the adenine are sites for CO
2
[34]
as electrodes. The onset potential for CO reduction at the CR-
adsorption;
(3) a Cu MOA built from successive junctions
2
MOF electrode was approximately 0.20 V higher than that ob-
with bis-bidentate dithiooxamidate (DTA) and named CuDTA;
and (4) a MOA with the same synthetic premise as (3) but with
oxides of Cu and Zn, named CuZnDTA. The coordination
frameworks of MOAs lack intrinsic pore systems; therefore, we
have processed the materials as nanofibrous aerogels to in-
crease their surface areas and the accessibility of the catalytic
centers. Each MOPM was deposited on a gas diffusion layer to
form a characteristic gas–solid–liquid three-phase interface,
which allows the mass-transfer limitations usually found in
[
23]
served on a Cu metal electrode. Hod et al. synthesized iron–
porphyrin-based MOFs for the electrocatalytic conversion of
CO ; these materials exhibited high active-site exposure
2
1
5
2
(
ꢁ10 sites per cm ) and nearly 100% Faradaic efficiency (FE)
[24]
for the production of CO+H mixtures. Kornienko et al. ob-
2
tained 76% Faradaic efficiency and high stability for 7 h using
Co–porphyrin MOFs. On the contrary, examples of metal–
[
27–31]
organic aerogels (MOAs) are relatively scarce
compared
with the more conventional MOFs. To the best of our knowl-
edge, no studies have dealt with their use as electrocatalysts
electrochemical systems to be overcome to enhance the CO
2
[4]
reduction performance. Then, we characterized the GDEs
through a set of analytical techniques and cyclic-voltammetry
analyses and finally tested the materials in the electrocatalytic
for CO reduction.
2
reduction of CO using a continuous filter-press electrochemi-
2
cal cell under different operating conditions.
Results and Discussion
Surface characterization of the GDEs
The MOPM-GDEs were characterized by SEM, attenuated total
reflectance (ATR) FTIR spectroscopy, and PXRD to understand
their structural and morphological properties in relation to
their capability for CO electrocatalytic reduction. In all cases,
2
the SEM images recorded at low magnification (5000ꢁ, see
Figure S3.1 of the Supporting Information) show that homoge-
nous films cover the entire sprayed GDE surface. At high mag-
nification (25000ꢁ), the microstructures of the HKUST-1 and
CuAdeAce GDEs reveal their polycrystalline natures (Figure 2a
and b) with strongly aggregated sub-micrometric crystals and
micrometric octahedral crystals, respectively. The images of the
CuDTA and CuZnDTA GDEs (Figure 2c and d) reveal filamen-
tous structures composed of highly crosslinked fibers with di-
ameters of 5 to 20 nm, comparable to those of the corre-
sponding as-prepared materials.
All of the ATR-FTIR spectra feature a set of peaks at n˜ =1305,
ꢀ
1
1
210, 1150, 1060, and 975 cm , which corresponds to the anti-
symmetric and symmetric stretching of the sulfonate, perfluori-
nated, and ether groups of the tetrafluoroethylene copolymer
ꢂ
used as the surfactant (Nafion ). Although the MOPMs show
less-intense peaks overlapped with those of the surfactant in
the low-energy range, the peaks arising from the coordination
framework can be distinguished at higher wavenumbers ( n˜ =
Figure 1. Structural details of the selected metal–organic porous materials
ꢀ
1
(MOPMs).
1310–1700 cm ). The detailed spectroscopy of the MOPM-
&
ChemSusChem 2016, 9, 1 – 11
www.chemsuschem.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Cyclic-voltammetry characterization
To examine the electrocatalytic behavior of the prepared GDEs,
cyclic voltammetry was performed for the MOPM-GDEs in CO2-
saturated (0.5m KHCO ) aqueous solutions; the voltammo-
3
grams after five scans with the current densities (j) normalized
to the geometric area of the MOPM-GDEs are shown in
Figure 4. The results are compared to the current–voltage re-
sponse of a Cu plate.
Figure 2. SEM images at 25000ꢁ magnification of (a) HKUST-1, (b) CuA-
deAce, (c) CuDTA, and (d) CuZnDTA; not real colors.
GDEs and the analysis of the main vibration modes are provid-
ed in the Supporting Information (see Figure S3.3 and
Table S3.1).
As the performance of a MOF is closely related to its crystal-
linity, HKUST-1 and CuAdeAce GDEs were further characterized
by PXRD (Figure 3). Both MOF-GDEs show characteristic (002)
and (004) reflections of the graphite sites at 2q=26.5 and
Figure 4. Cyclic voltammograms for the MOPM-GDEs in a CO
.5m KHCO aqueous solution.
2
-saturated
0
3
54.68, respectively. Although, the preferred orientation of
graphite gives rise to an outsized peak, the signals correspond-
ing to the MOFs are clearly distinguishable at 2q<228. As
shown in the inset graphic, all of the observed peaks fit the ex-
pected lattice-plane reflections.
Large differences between the voltammetric profiles of the
MOPM-GDEs can be seen in Figure 4, and HKUST-1 and CuZnD-
TA are the most promising candidates for the electroreduction
process. All of the voltammetric profiles show a reduction pro-
cess that starts at approximately ꢀ1 V versus Ag/AgCl and is
associated with the reduction of CO . On the other hand, for
2
the applied voltage, the CuAdeAce GDE shows almost no re-
sponse variation, which reveals a low activity for conducting
electrons. It is also important to note the synergic effect of Cu
and Zn in the reduction response, as denoted by the remarka-
bly lower current densities at lower onset potentials for
CuZnDTA compared with those of CuDTA. This result is in
[5]
agreement with those for Cu O/ZnO and Cu O GDEs. Further-
2
2
more, the oxidative peak at ꢀ0.8 V, previously assigned to the
[5]
transformation of Zn to ZnO, is assigned in this work to the
formation of oxidized subproducts from the CO reduction re-
2
action because it is present for both CuDTA and CuZnDTA
GDEs, and each voltammetry profile shown in Figure 5 corre-
sponds to the fifth cycle.
To further confirm the reduction of CO , the voltammetry
2
profiles of the most promising GDEs (HKUST-1 and CuZnDTA)
in the presence and absence of CO (in an Ar-saturated solu-
2
tion) are shown in Figure 5a and 5b. The decrease of the cur-
rent intensity for both GDEs is an indication that CO is re-
2
duced at an onset potential lower than ꢀ1 V versus Ag/AgCl.
Thus, the intrinsic oxidation–reduction of the GDEs might be
neglected, particularly for HKUST-1 GDE, for which the reduc-
tion response in the absence of CO is close to the response
2
Figure 3. PXRD patterns of HKUST-1 and CuAdeAce GDEs.
for a Cu plate in a CO -saturated KHCO aqueous solution.
2
3
ChemSusChem 2016, 9, 1 – 11
www.chemsuschem.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
voltammetry profiles (Figure 4). This electrocatalytic per-
formance is partially correlated with the surface area of the
GDEs, as summarized in Table S2.1 in the Supporting Informa-
tion: the HKUST-1 GDE exhibits the highest surface area
2
ꢀ1
(
1710 m g ) and the highest electrocatalytic performance,
whereas CuDTA and CuZnDTA GDEs exhibit much lower sur-
2
ꢀ1
face areas (270 and 260 m g , respectively) and significantly
poorer electrocatalytic performances. Nevertheless, the surface
area is not the only parameter that controls the catalytic per-
formance, as CuAdeAce GDE with an intermediate area
2
ꢀ1
(
500 m g ) shows the lowest FE values. The last results could
be anticipated from the cyclic-voltammetry profiles displayed
in Figure 4, which indicated the lowest electron conductivity
for the CuAdeAce GDE. Thus, the electrocatalytic performance
should be related to additional features of the GDEs that re-
quire additional analysis of the Cu active sites, particularly the
II
accessibility of the pentacoordinate Cu centers, which are hin-
dered sterically by the surrounding ligands. This leads to the
preliminarily conclusion that MOPMs with unsaturated coordi-
nation positions exposed in the pore system are preferred for
the enhancement of the performance of the electrocatalytic re-
duction of CO to alcohols.
2
The maximum CH OH and C H OH production rates for the
3
2
5
ꢀ
6
ꢀ2 ꢀ1
HKUST-1 GDEs (r
=5.62ꢁ10 molm
s
and rC2H5OH =
CH3OH
ꢀ2 ꢀ1
ꢀ
6
5
.28ꢁ10 molm s ) correspond to product concentrations
ꢀ
1
in the catholyte of 0.54 and 0.73 mgL
for CH OH and
3
C H OH, respectively. The formation of both alcohols, CH OH
2
5
3
[4,15,35–46]
and C H OH, has been reported previously,
whereas
2
5
Figure 5. Cyclic-voltammetry responses in medium saturated with CO
KHCO ) and Ar of (a) HKUST-1 and (b) CuZnDTA.
2
(0.5m
Cu-based GDEs are more selective towards the formation of
[4]
3
CH OH over C H OH. The maximum FEs of CH OH and
3
2
5
3
C H OH were 54.8 and 31.4% for Cu O and Cu O/ZnO GDEs, re-
2
5
2
2
Influence of current density on CO reduction performance
spectively, at applied potentials of ꢀ1.39 and ꢀ1.16 V versus
Ag/AgCl. Furthermore, the CuO GDE led to a higher selectivi-
2
[4]
The results for the continuous electrocatalytic reduc-
tion of CO in a filter-press electrochemical cell are
2
presented hereafter. The quantitative reduction per-
formances (production rate, r, and Faradaic efficiency,
FE) regarding the liquid-phase product distribution
ꢀ
2
at different current densities (j=5–40 mAcm ) are
shown in Figure 6 for the prepared MOPM-GDEs. The
electrocatalytic reduction of CO2 with the MOPM-
GDEs leads to the formation of CH OH and C H OH.
3
2
5
It should be noted that the carbon paper without
MOPMs did not produce any measurable liquid prod-
uct. The FEs were calculated for a 6-electron pathway
for CO reduction to CH OH and a 12-electron path-
2
3
way to C H OH. A constant electrolyte-flow/area ratio
2
5
(
Q /A) and gas-flow/area ratio (Q /A) of 2 and
e g
ꢀ
1
ꢀ2
2
0 mLmin cm , respectively, were applied. These
conditions were found previously to be optimal for
[
4]
the electrocatalytic reduction of CO2.
As shown in Figure 6, the product distribution and
process efficiency are correlated with the current
density applied to the system. The HKUST-1 and
CuZnDTA GDEs are the most active electrocatalysts
Figure 6. Rates (r) for CH
3
OH (*) and C
2
H
5
OH (*) formation and Faradaic efficiencies
as a function of the current density (j) ap-
(
FEs) in the electrocatalytic reduction of CO
2
for the reduction of CO , in agreement with the
2
plied with (a) HKUST-1, (b) CuAdeAce, (c) CuDTA, and (d) CuZnDTA; the lines are only
guides.
higher reduction responses observed in the cyclic-
&
ChemSusChem 2016, 9, 1 – 11
www.chemsuschem.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
[39]
ty for C H OH (FE=15.5%) in a 0.2m KHCO solution, and
2
5
3
trace amounts of CH OH and C H O were also detected. Anoth-
3
3
8
er example is the recent study conducted by Gutierrez-Guerra
[
46]
et al.
for different Cu-based GDEs, which afforded CH OH
3
and C H OH with selectivities of 80 and 10%, respectively, at
2
5
[
15]
an applied current of ꢀ30 mA. Kuhl et al. reported a total of
6 different CO2 reduction products (including CH OH and
1
3
C H OH) across a range of potentials. They hypothesized that
2
5
the chemistry involved in the CꢀC coupling reactions to form
C –C products occurs through an enol-like surface intermedi-
2
3
ate, which desorbs to convert to its diol or ketone form. Cer-
tainly, the CꢀC bond formation is one of the most critical fac-
tors to be considered in the design of electrocatalysts for the
production of alcohols, and further experimental work is
needed to fully elucidate the CO reduction steps to form alco-
2
hols using Cu-based GDEs.
As shown in Figure 6, the rates for CO reduction to CH OH
2
3
ꢀ
2
and C H OH did not improve at j>10 mAcm . At this point,
2
5
the maximum r values can be obtained for all MOPM-GDEs.
The total Faradaic efficiency (FE , cumulative efficiency for the
T
formation of CH OH and C H OH) drops drastically as the cur-
3
2
5
ꢀ
2
rent increases from j=10 to 40 mAcm . This result could be
explained by the consumption of the additional current by
side reactions; hence, the optimal current density is
ꢀ
2
1
0 mAcm for all MOPM-GDEs. Under these conditions, the
FE values are 10.9 and 7.3% for the HKUST-1 and CuZnDTA
T
GDEs, respectively. The remaining product is expected to be
mainly H , which competes with the electrocatalytic reduction
2
ꢀ
2
T
Figure 7. Total rates (r ) at j=10 mAcm for the MOPM-GDE with (a) differ-
of CO to alcohols and affects the GDE stability negatively. The
2
e 2 g
ent electrolyte flow rates (Q /A) and (b) CO gas flow rates (Q /A); the lines
latter observation is caused by the fact that H favors the
2
are only guides.
[
4]
leaching of the active material from the GDE.
ꢀ
1
–2
Moreover, the increases in Q /A from 10 to 20 mLmin cm
g
yields an increase of the CO electrocatalytic conversion rate
2
Influence of electrolyte flow rate and gas flow rate
ꢀ
6
(
Figure 7b,
e.g.,
from
rT =13.94ꢁ10
to
18.57ꢁ
ꢀ
6
ꢀ2 ꢀ1
Previous studies demonstrated that variations in Qe and Qg
could lead to significant mass-transfer differences in the cell
and, thus, in the total rate for CO transformation, r , and the
10 molm s
for HKUST-1 GDE). This observation reveals
ꢀ
1
that the overall kinetics at Q /A=10 mLmin are controlled
g
by the external transport of CO to the GDE actives sites. A fur-
2
T
2
[
4,5]
ꢀ1
ꢀ2
cumulative FE of the process.
These effects can be ob-
ther increase of Q /A above 20 mLmin cm leads to a drastic
T
g
[
4]
served even for the application of GDEs, for which mass-
decrease in r , which is then attributed to the leaching of
T
[
6,47,48]
transfer limitations are expected to be overcome partially.
active material from the GDE, in agreement with previous find-
[
4,48,50]
In an attempt to improve the CO conversion efficiency, ad-
ings.
Thus, the optimal balance point is at Q /A=
g
2
ꢀ
1
–2
ditional experiments were conducted at different Q /A and Q /
20 mLmin cm , at which enough CO gas supply is provided
e
g
2
A values. The results are presented in Figure 7. The increase in
for the reaction well before a massive detachment of active
material occurs.
ꢀ
1
–2
Q /A from 1 to 3 mLmin cm led to a significant enhance-
e
ment in the CO2 electrocatalytic conversion rate (Figure 7a,
Overall, the optimal conditions for the CO electrocatalytic
2
ꢀ1
ꢀ
6
ꢀ6
ꢀ2 ꢀ1
ꢀ2
e.g., from r =5.80ꢁ10 to 18.57ꢁ10 molm
s
for HKUST-
reduction on MOPM-GDEs is Q /A=3 mLmin cm and Qg/
e
T
ꢀ
1
ꢀ2
1
GDE). The use of low Q /A ratios is preferred as the concen-
A=20 mLmin cm . The optimal electrocatalytic performan-
ces, in terms of r and FE, for the MOPM-GDEs are shown in
Table 1. To interpret the electrocatalytic activity further, the
total formation rates were normalized to the active Cu surface
e
tration of alcohols in the liquid would be higher; therefore, for
an optimal process operation, a compromise needs to be met
between the concentration of alcohols in the product and
their formation rate. Further increases in Q /A led to a drastic
area of each GDE, r . The active Cu area (a) was measured
e
T,a
decrease in r owing to the leaching of the active material
through pulse chemisorption, as described in the Experimental
Section. The results were compared to those obtained for
a filter-press electrochemical cell equipped with a Cu plate at
an applied potential of ꢀ1.3 V versus Ag/AgCl (j=
T
[
4]
from the GDE. Furthermore, a low Q /A allows the infiltration
e
of the catholyte into the GDE structure, which increases the
diffusion time and enhances CO2 electrocatalytic per-
[
4,48,49]
ꢀ2
formance.
10.83 mAcm ).
ChemSusChem 2016, 9, 1 – 11
www.chemsuschem.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
found for CuAdeAce implies that the scarce
Table 1. r and FE for the electrocatalytic conversion of CO
2
at MOPM-GDEs. j=
II
ꢀ2
ꢀ1
ꢀ2
ꢀ1
ꢀ2
amount of accessible Cu sites are highly active
10 mAcm , Q
e
/A=3 mLmin cm , Q
g
/A=20 mLmin cm
.
centers and it provides a clue for the design of
more efficient electrocatalysts for CO2 reduction
based on {Cu (m -adeninate) [m -OOC(CH ) ]} pad-
ꢀ
6
ꢀ2 ꢀ1
GDE
E
a
Cu
r [10 molm
s
]
FE [%]
2
[
V]
[cm ]
r
CH3 OH
r
C2 H5OH
r
T
r
T,a
FECH3 OH FEC2 H5 OH FE
T
2
3
2
2
3 2 n
HKUST-1
ꢀ0.9
66.48 9.68
0.13 1.25
8.90
0.43
3.58
5.64
–
18.58
1.68 13.1
6.86
11.57
8.7
0.28 5.6
0.7
0.13 1.9
6.52 3.4
10.3
0.5
4.1
6.5
–
15.9
1.2
6
9.9
4.6
dlewheel motifs.
CuAdeAce ꢀ1.75
On the other hand, the comparison of the nor-
malized rates for CuDTA and CuZnDTA (0.13 and
CuDTA
ꢀ1.41 53.69 3.28
CuZnDTA
ꢀ1.25
ꢀ1.3
1.78 5.93
8.7
[
a]
ꢀ2 ꢀ1
Cu plate
–
–
4.6
6.52 molm s ) supports the previously inferred
II
ꢀ
2
ꢀ1
ꢀ2
crucial role that Zn centers play on the per-
[a] Data from Ref. [5] at j=10.83 mAcm and Q
e
/A=2 mLmin cm
.
formance of MOF-GDEs.
Long-term stability of MOPM-GDEs
The alcohol formation rate after 90 min on-stream varied
ꢀ
6
ꢀ6
ꢀ2 ꢀ1
from r =1.68ꢁ10 to 18.58ꢁ10 molm
s
for CuAdeAce
T
Finally, the stabilities of the MOPM-GDEs were tested over an
extended period. The evolution of the cumulative Faradaic effi-
and HKUST-1 GDEs, respectively. The latter GDE also displays
the highest FE (ꢁ16%) among the studied electrodes. These
values, together with those obtained for CuZnDTA GDE, are re-
markably greater than those for a Cu plate and show the great
opportunities brought by MOPMs for the electrocatalytic re-
ciency over 5 h on-stream, FE , for the observed optimal experi-
T
mental conditions is shown in Figure 8.
duction of CO . This enhanced performance is related to the
2
demonstrated higher activity for the reduction of protons by
0
CuO in comparison with that of Cu for the electrocatalytic re-
x
[
24,35–37,51,52]
[37]
duction of CO2.
For example, Lan et al. investigat-
ed the electrocatalytic reduction of CO at a Cu/CuO (core–
2
shell) catalyst in 1m KHCO3 with a flow reactor. This study
I
proved that transformations between Cu, Cu, and CuO oc-
curred as a function of applied potential, which at the same
time did not affect CH OH production severely. Thus, the high
3
yield of CH OH obtained at ꢀ1.72 V versus Ag/AgCl using
3
a Cu/CuO (core–shell) electrocatalyst was higher than that ob-
tained using Cu foil. Furthermore, the formation rates reached
using the HKUST-1 GDEs are in the range of those values re-
ported previously for air-oxidized Cu foil and electrochemically
ꢀ
5
ꢀ2 ꢀ1
oxidized Cu foil (rꢁ2ꢁ10 molm s ) at potentials of ꢀ1.2
Figure 8. Time-dependence of FE for the MOPM-GDEs under the following
ꢀ
2
ꢀ1
ꢀ2
[
53]
e
conditions: j=10 mAcm , Q /A=3 mLmin cm , and Qg/
to ꢀ1.5 V versus Ag/AgCl. Nevertheless, the formation rates
are still below those reported for Cu O GDEs (r =11.9ꢁ
ꢀ
1
ꢀ2
A=20 mLmin cm ; the lines are only guides.
2
T
ꢀ
5
ꢀ2 ꢀ1
[54]
10
molm
s
at ꢀ1.05 V versus Ag/AgCl) and also those
with Cu O/Zn-based GDEs (r=4.74ꢁ
molm s ). This result is related to the beneficial prop-
achieved
As indicated in Figure 8, during the first minutes of the ex-
periment, the electrolyte needs to diffuse through the internal
structure of the GDE to form a typical three-phase interface
throughout the whole GDE and, thus, enlarge the contact
2
ꢀ
5
ꢀ2 ꢀ1 [4]
10
I
erties of Cu for CO electrocatalytic reduction processes. Previ-
2
I
ous studies demonstrated that Cu presents both intermediate
[55]
hydrogen overpotentials and CO adsorption properties, which
promote CO2 reduction in aqueous solutions in comparison
area. The electrocatalytic activity (FE ) decays to a plateau
T
and remains almost stable during the rest of the run (pseudos-
tationary state), except that for the CuDTA GDE, which drops
to zero after 120 min. This deactivation observed for all
0
II
[36,51]
with Cu or Cu -based electrocatalysts.
Interestingly, if the formation rates are normalized to the
active Cu surface available, CuAdeAce shows the highest activi-
MOPM-GDEs is attributed to the decrease of their active-site
ꢀ
6
ꢀ2 ꢀ1
[48,49]
ty (r =13.1ꢁ10 molm s ). Conversely, the lowest rates
areas
or the degradation of the MOPMs owing to their lim-
T,a
were observed for the HKUST-1 GDEs (rT,a =0.28ꢁ
ited stabilities in water, in which they undergo hydrolysis,
amorphization, or phase transformations, even at room tem-
ꢀ
6
ꢀ2 ꢀ1
10
molm s ). To explain such unexpected behavior, first it
[56,57]
must be considered that the idealized crystal structure sug-
gests that the coordination framework of CuAdeAce lacks ac-
perature.
In this sense, the MOPM-GDEs used for 30 min
were analyzed further and compared with the as-prepared
ones. The ATR-FTIR spectroscopy measurements (Figure 9a
and b) of the MOF-GDEs (HKUST-1 and CuAdeAce) show a de-
pletion of the intensity of the main reference signal attributed
to the MOF-GDE; this suggests that a degradation of the mate-
rial occurs during the run.
II
cessible unsaturated Cu coordination positions, and this leads
2
ꢀ1
to a low Cu active-surface value (0.13 cm g ), which is prob-
Cu
ably attributable to the presence of crystal defects that render
II
some Cu sites unhindered and available for N O chemisorp-
2
tion. Therefore, the superior value of the normalized rate
&
ChemSusChem 2016, 9, 1 – 11
www.chemsuschem.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Figure 10. Comparison of the PXRD patterns of fresh and used (a) HKUST-
1
and (b) CuAdeAce.
FTIR spectra recorded at the end of the run (300 min, see Sup-
porting Information) reveal a series of emerging peaks that evi-
dence the formation of copper(II) hydroxycarbonate (mala-
chite), which is a plausible degradation path for all Cu-MOPMs
and not detectable at the middle of the run. Furthermore, at
the end of the run, the decay of the signals corresponding to
CuDTA is greater than that observed for CuZnDTA, and this ex-
plains the differences in the FE trend at the last part of the
T
run.
The FTIR and PXRD analyses of the fresh and used MOPM-
GDEs give a qualitative clue to the activity loss, whereas
a quantitative one is obtained from the relative FET losses.
Figure 9. ATR-FTIR spectra for fresh and used (a) HKUST-1, (b) CuAdeAce,
(c) CuDTA, and (d) CuZnDTA.
After 5 h on-stream, the relative FE drops are 40, 65, 98, and
T
51% for HKUST-1, CuAdeAce, CuDTA, and CuZnDTA GDEs, re-
spectively. These relative activity losses indicate that all MOPM-
GDEs (except CuDTA) retain an intermediate efficiency despite
the cited long-term degradation. In a practical sense, a third
reason for the deactivation should be indicated, namely the
leaching of the MOPM from the rest of the GDE structure
Accordingly, the PXRD analysis of the HKUST-1 and CuA-
deAce GDEs (Figure 10) shows a significant reduction of crys-
tallinity, which is more acute for HKUST-1. It must be pointed
out that the performance loss of HKUST-1 is smaller than that
expected from the drastic crystallinity loss; thus, the remaining
efficiency level can be attributed to the preservation of the
local structure, as suggested by the FTIR spectra. Coming back
to the FTIR analysis (Figure 9), the loss of intensity for the
MOA-GDEs (Figure 9c and 9d) is not so evident; therefore, the
[4]
(carbon support). In addition, the probable agglomeration of
particles and defects in the catalytic layer during the prepara-
tion of the MOPM-GDE would likely assist tunneling and in-
crease the unwanted H formation owing to the easy access of
2
[51]
water to catalytic sites. On the contrary, the GDEs containing
HKUST-1 and CuZnDTA are able to retain moderate formation
initial FE decay can be related to a shallow surface degrada-
T
ꢀ
6
ꢀ2 ꢀ1
tion/modification of the CuDTA and CuZnDTA nanofibers that
is not detectable in the FTIR spectra at this stage, as the bulk
of the material remains unaltered. However, in all cases, the
rates (KHUST-1, r =18.58ꢁ10 molm s ; CuZnDTA, rT =
T
ꢀ
6
ꢀ2 ꢀ1
11.57ꢁ10 molm s ) and Faradaic efficiencies (KHUST-1,
FE =15.9%; CuZnDTA, FE =9.9%) for as long as 12 or 17 h
T
T
ChemSusChem 2016, 9, 1 – 11
www.chemsuschem.org
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
and surpass the stability value reached recently for MOF-GDEs
conversion of CO to alcohols in continuous operation. The
2
[
19]
for the electrocatalytic transformation of CO2.
modularity of these systems yields many opportunities for fur-
ther performance improvements and open new directions in
electrocatalysis.
These results make the use of MOPMs valuable for the CO2
electrocatalytic conversion to alcohols in continuous operation,
although further work is required to design materials with the
same favorable properties and a higher stability for a technoe-
conomically viable CO valorization process. The outstanding
2
challenges remain in the design of catalyst systems featuring
Experimental Section
(
(
1) selectivity for CO reduction with minimum H generation,
2 2
Preparation of MOPM-GDEs
2) high conversion efficiency at low electrochemical overpo-
tentials, and (3) long-term stability. Furthermore, the detailed
mechanisms for the overall catalytic system remain unclear. We
Synthesis of the MOPMs: HKUST-1 was prepared by a previously
[58]
described solvent-free synthetic route. In a first step, stoichio-
metric amounts of benzene-1,3,5-tricarboxylic acid and copper(II)
acetate monohydrate were ground together to ensure a homoge-
neous mixture and placed in the reaction vessel. The reagent mix-
ture was oven-heated for approximately 50 h at a heating rate of
hope to elucidate the reactions steps for CO conversion to
2
value-added products on MOPM-GDEs.
ꢀ1
Conclusions
28Ch to a maximum temperature of 1208C. Polycrystalline CuA-
deAce material was prepared by the slow addition of acetic acid to
This work demonstrates the ability of Cu-containing metal–or-
ganic porous materials (MOPMs) supported in gas diffusion
electrodes (GDEs) to promote the electrocatalytic conversion
of CO2 to alcohols. We successfully prepared, characterized,
and tested four different MOPM-GDEs. Specifically, two metal–
organic frameworks, (1) [Cu (m -C H O ) ] (HKUST-1) and
II
an aqueous solution containing adenine and Cu salt in equimolar
[59]
proportions. Both MOFs were washed thoroughly with water to
remove unreacted reagents and remaining byproducts. The gener-
al procedure to prepare the MOAs (CuDTA and CuZnDTA) proceed-
ed as follows. The corresponding metal acetate (or metal salt mix-
ture) was dissolved in a mixture of N,N’-dimethylacetamide (DMA)
and DMF in a 60:40 volumetric ratio, aided by an ultrasonic tip
3
6
9
3
6 2 n
(
2) [Cu (m -C H N ) ] (CuAdeAce), and two metal–organic aero-
3 3 5 4 5 2 n
(Vibra-Cell VCX130 20 kHz and 130 W, Sonics) at 80% of its power
gels, (3) [Cu(m-C H N S )]
(CuDTA) and (4) [Cu Zn (m-
0.6 0.4
2
2
2
2
n
for 2 min. Then, dithiooxamide (H DTA) ligand, basified with trie-
2
C H N S )] (CuZnDTA). The characterization involved structural
2
2
2 2 n
thylamine, was dissolved in the same solvent mixture and added
into the metal-ion-containing solution with the system maintained
in an ultrasound bath (ULTRASONS-H, Selecta) at a temperature of
288 K. Once the metal–organic gel reached a certain consistency, it
was allowed to age at room temperature for 1 d. Thereafter, the
materials were washed first through immersion in pure DMF to
remove the unreacted species and then by successive solvent ex-
changes in DMF/ethanol mixtures and pure ethanol to replace the
solvent. In each exchange step, the contact between the solvent
and gel was 24 h to ensure an efficient exchange. To prepare the
aerogels, an E3100 critical-point dryer from Quorum Technologies
equipped with gas-inlet, vent, and purge valves and a thermal
and cyclic-voltammetry analyses, whereas the testing during
the electrocatalytic reduction of CO was performed in a contin-
2
uous setup consisting of a filter-press electrochemical cell
under ambient conditions.
The analyses of the electrolysis products showed that meth-
anol and ethanol were formed predominately as the liquid
products from CO reduction. An enhanced performance for
2
CO2 conversion was achieved through the application of
HKUST-1 and CuZnDTA GDEs at a current density (j) of
ꢀ
2
1
3
2
1
1
1
0 mAcm , an electrolyte-flow/area ratio (Q /A) of
e
ꢀ
1
ꢀ2
mLmin cm , and
a
gas-flow/area ratio (Q /A) of
bath was employed. Firstly, the gel was immersed in liquid CO
at
2
g
ꢀ
1
ꢀ2
2
93 K and 50 bar for 1 h. After this, the exchanged ethanol was re-
0 mLmin cm , at which moderate formation rates (KHUST-
ꢀ
6
ꢀ2 ꢀ1
moved through the purge valve. This process was repeated five
times. Subsequently, the sample was dried under supercritical con-
ditions at a temperature of 311 K and a pressure of 85–95 bar. Fi-
nally, under constant temperature (311 K), the chamber was vented
slowly to atmospheric pressure. Details on the characterization of
the prepared MOPMs are provided in the Supporting Information
(S2).
,
r =18.58ꢁ10 molm s ;
CuZnDTA,
rT =11.57ꢁ
T
ꢀ
6
ꢀ2 ꢀ1
0
molm s ) and Faradaic efficiencies (KHUST-1, FE =
T
5.9%; CuZnDTA, FE =9.9%) could be obtained. These results
T
denoted that MOPMs with unsaturated coordination positions
exposed in the pore system are preferred for the enhancement
of the performance of the electrocatalytic reduction of CO to
2
alcohols. Interestingly, if the formation rates were normalized
to the active Cu surface available for each MOPM, CuAdeAce
showed a superior activity. This gives a clue for the design of
Preparation of the GDEs: The MOPM-GDEs were prepared by the
[4]
2
procedure described in a previous study. The GDEs (A=10 cm )
were prepared by airbrushing a catalytic ink onto a porous carbon
paper (type TGP-H-60, Toray Inc.). The catalyst loading in the GDEs
more-efficient CO reduction electrocatalysts including paddle-
2
ꢀ
2
wheel motifs built from N-donor ligands that preserve square-
was kept at 1 mgcm , which is an effective loading for enhanced
II
[5]
planar coordination geometries around the Cu atoms and, as
CO electrocatalytic reduction performance. The catalytic ink was
2
formed by a mixture of the synthesized MOPMs (HKUST-1, CuA-
deAce, CuDTA, and CuZnDTA) as electrocatalysts, Nafion disper-
a result, produce open metal sites prone to interact strongly
with guest molecules throughout the porous network.
ꢂ
sion 5 wt% (Alfa Aesar) as binder, and isopropyl alcohol (IPA,
Sigma–Aldrich) as the vehicle with a 70:30 catalyst/Nafion mass
ratio and 3% solids (catalyst+Nafion). The mixture was sonicated
for 15 min and then airbrushed onto the carbon papers, and the
resulting MOPM-GDEs were dried and rinsed with deionized water
before use.
Finally, the stability of the HKUST-1 GDE was confirmed for
as long as 17 h of operation and can be attributed to the pres-
ervation of the local structure, even if a significant reduction in
crystallinity occurred during the experimental time. These re-
sults make the use of MOPMs valuable for the electrocatalytic
&
ChemSusChem 2016, 9, 1 – 11
www.chemsuschem.org
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Characterization of the prepared MOPM-GDEs
The electrochemical behavior of the materials was evaluated by
cyclic voltammetry with a PGSTAT 302N potentiostat (Metrohm,
Autolab B.V.) under GPES software control using a conventional
three-electrode electrochemical cell. The current–voltage curves
ꢀ
1
were obtained with a scan rate of 50 mVs at potentials ranging
from 0 to ꢀ1.6 V versus Ag/AgCl in a CO -saturated 0.5m KHCO
2
3
(
Panreac) aqueous solution as the electrolyte. Portions of the
MOPM-based GDEs were used as working electrodes, and glassy
carbon and Ag/AgCl (sat. KCl) were used as the counter and refer-
ence electrode, respectively. The current density is expressed as
the total current divided by the geometric surface area of the elec-
trodes.
Scheme 1. Schematic representation of the experimental setup.
Elemental analyses (C, H, N) were performed with a Euro EA ele-
mental analyzer (Eurovector, Milan, Italy), whereas the metal con-
tent was determined by inductively coupled plasma atomic emis-
sion spectrometry (ICP-AES) with a Horiba Yobin Yvon Activa instru-
ment (Kyoto, Japan). The IR spectra were recorded with an Shimad-
zu FTIR 8400S spectrometer (Shimadzu, Kyoto, Japan) in the 4000–
peristaltic pumps (Watson Marlow 320, Watson Marlow Pumps
Group). In this study, the filter-press electrochemical system pos-
sesses three inputs (catholyte, anolyte, and CO
two outputs (catholyte-CO and anolyte), which make possible the
formation of a gas–solid–liquid interface for the electrocatalytic re-
separately) and
2
2
[4]
duction of CO in the gas phase.
2
ꢀ1
4
00 cm wavenumber region with a PIKE MIRacle universal ATR
All of the experiments were performed under galvanostatic condi-
tions (i.e., at a constant current density) with an AutoLab PGSTAT
sampling accessory equipped with a ZnSe crystal. The PXRD pat-
terns were collected using a Phillips X’PERT powder diffractometer
3
02N potentiostat (Metrohm, Autolab B.V.). The current density
(
Panalytical, Eindhoven, The Netherlands) with CuK radiation (l=
ꢀ2
a
ranged from j=5 to 40 mAcm . The experimental time was
1
.54060 ꢃ) over the range 5<2q<708 with a step size of 0.028
9
0 min, for which pseudostable conditions are ensured according
and an acquisition time of 2 s per step at 258C. The N (77 K) phys-
[4,5]
2
to our previous analyses.
Liquid samples were taken every
isorption data of the materials (vacuum at 1508C for 12 h) was re-
corded with a Quantachrome Autosorb-iQ-MP instrument (Quan-
tachrome Instruments, Florida, United States). Field-emission scan-
ning electron microscopy (SEM) studies were performed with
a JEOL JSM-7000F microscope. Before the analysis, all of the GDEs
were coated with 5 nm of chromium.
1
5 min from the catholyte tank. To quantify the concentration of
each product in the liquid phase, the samples were analyzed in du-
plicate in a headspace gas chromatograph (GC–MS QP2010, Ultra
Shimadzu) equipped with a flame ionization detector (FID). The
compounds were separated using a DB-Wax 30 mꢁ0.25 mmꢁ
0
.25 mm column with an injection and detector temperature of 250
The Cu exposure (a) of the GDEs was determined by N O pulse
chemisorption with an AutoChem 2920 analyzer (Micromeritics,
Georgia, USA) coupled to a Omnistar (Balzers Instruments, New
and 2708C, respectively. Helium was used as the carrier gas at
2
ꢀ
1
a flow rate of 50 mLmin . The identification of the obtained prod-
ucts was further confirmed by a headspace GC–MS instrument
(N5975B) equipped with a 60 mꢁ250 mmꢁ1.40 mm DB-624 capilla-
ry column. The product concentrations were averaged from at
least three replicates. The standard deviations of all experiments
were below 19.2%.
Jersey, USA) mass spectrometer. Part of the N O can be reduced to
2
N and otherwise chemisorbed or left unreacted. In particular, we
2
calculated the Cu exposure through the cumulative disappearance
3
ꢀ1
of N O. The GDEs were first treated at 1008C in a 50 cm min
2
stream of 10 vol% H in Ar (Air Liquide, Madrid, Spain) over 2 h.
2
The performance of the electrochemical processes were evaluated
through the rate of product formation, r (i.e., the product obtained
per unit of cathode area and time), and the Faradaic efficiency, FE
3
ꢀ1
Then, the GDEs were kept at 358C in a 50 cm min stream of He,
3
and 20 pulses (0.25 cm each) of 10 vol% N O in He (Air Liquide,
2
Madrid, Spain) were applied. The signals of the N O and N were
2
2
(i.e., the selectivity of the reaction for the formation of the different
recorded in the effluent by the mass spectrometer at m/z=44 and
products), according to Equation (1):
19
2
8, respectively. Cu was assumed to have a density of 1.63ꢁ10
2
Cu atoms per m .
FEð%Þ ¼ ðznFÞ=q ꢂ 100
ð1Þ
z is the theoretical number of electrons exchanged to form the de-
sired product, n is the number of moles produced, F is the Faradaic
constant (96485 Cmol ), and q is the total charge applied in the
Electrochemical cell and experimental conditions
ꢀ1
The prepared MOPM-GDEs were evaluated for the continuous elec-
process.
trocatalytic reduction of CO in a filter-press electrochemical cell
2
(
Micro Flow Cell, ElectroCell A/S) under ambient conditions. A
Nafion 117 cation-exchange membrane was used to separate the
cathode and anode compartments. The MOPM-GDEs were em-
ployed together with a platinized titanium electrode as the anode
and a Ag/AgCl (sat. KCl) reference electrode assembled close to
the cathode. A schematic representation of the experimental plant
is shown in Scheme 1. The cathode side of the reactor was fed
Acknowledgements
The authors gratefully acknowledge the financial support from
the Spanish Ministry of Economy and Competitiveness (MINECO)
under projects CTQ2013-48280-C3-1-R and CTQ2013-48280-C3-3-
R and MAT2013-46502-C2-1-P, and the Juan de la Cierva program
(JCI-2012-12073).
ꢀ1
ꢀ2
with CO gas (99.99%) at Q /A=10 to 40 mLmin cm . A 0.5m
2
g
KHCO (Panreac) aqueous solution was used as both the catholyte
3
ꢀ
1
ꢀ2
and anolyte at Q /A=1–4 mLmin cm . The electrolytes were
e
pumped from the catholyte and anolyte tanks to the cell by two
ChemSusChem 2016, 9, 1 – 11
www.chemsuschem.org
9
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
[
30] Q. Yang, X. Tan, S. Wang, J. Zhang, L. Chen, J. P. Zhang, C. Y. Su, Micropo-
rous Mesoporous Mater. 2014, 187, 108–113.
Keywords: alcohols · copper · electrodes · metal–organic
frameworks · reduction
[31] H. Zhu, X. Yang, E. D. Cranston, S. Zhu, Adv. Mater. 2016, DOI: 10.1002/
adma.201601351.
[
[
[
[
[
[
32] S. S. Chui, S. M. Lo, J. P. Charmant, A. G. Orpen, I. D. William, Science
[
1] M. E. Boot-Handford, J. C. Abanades, E. J. Anthony, M. J. Blunt, S. Branda-
ni, N. Mac Dowell, J. R. Fernandez, M. C. Ferrari, R. Gross, J. P. Hallet, R. S.
Haszeldine, P. Heptonstall, A. Lyngfelt, Z. Makuch, E. Mangano, R. T. J.
Porter, M. Pourkashanian, G. T. Rochelle, N. Shah, J. G. Yao, P. S. Fenell,
Energy Environ. Sci. 2014, 7, 130–189.
1
999, 283, 1148–1150.
33] K. Sumida, D. L. Rogow, J. A. Mason, T. M. McDonald, E. D. Bloch, Z. R.
Herm, T. Bae, J. R. Long, Chem. Rev. 2012, 112, 724–781.
34] S. Pꢆrez-YꢄÇez, G. Beobide, O. Castillo, M. Fischer, F. Hoffmann, M.
Frçba, J. Cepeda, A. Luque, Eur. J. Inorg. Chem. 2012, 5921–5933.
35] T. Y. Chang, R. M. Liang, P. W. Wu, J. Y. Chen, Y. C. Hsieh, Mater. Lett.
[2] J. Albo, T. Yoshioka, T. Tsuru, Sep. Purif. Technol. 2014, 122, 440–448.
[3] J. Albo, A. Irabien, J. Chem. Technol. Biotechnol. 2012, 87, 1502–1507.
[4] J. Albo, A. Irabien, J. Catal. 2015, DOI: 10.1016/j.jcat.2015.11.014.
[5] J. Albo, A. Sꢄez, J. Solla-Gullꢅn, A. Irabien, Appl. Catal. B 2015, 176–177,
2
009, 63, 1001–1003.
36] M. Le, M. Ren, Z. Zhang, P. T. Sprunger, R. L. Kurtz, J. C. Flake, J. Electro-
chem. Soc. 2011, 158, E45–E49.
37] Y. Lan, S. Ma, J. Lu, P. J. A. Kenis, Int. J. Electrochem. Sci. 2014, 9, 7300–
7
09–717.
6] A. Del Castillo, M. Alvarez-Guerra, A. Irabien, AIChE J. 2014, 60, 3557–
564.
[
7308.
3
[
[
38] F. Jia, X. Yu, L. Zhang, J. Power Sources 2014, 252, 85–89.
39] D. Chi, H. Yang, Y. Du, T. Lv, G. Sui, H. Wang, J. Lu, RSC Adv. 2014, 4,
[
[
7] G. Centi, S. Perathoner, Catal. Today 2009, 148, 191–205.
8] J. Albo, M. Alvarez-Guerra, P. CastaÇo, A. Irabien, Green Chem. 2015, 17,
37329–37332.
2304–2324.
[
40] R. Kas, R. Kortlever, A. Milbrat, M. T. M. Koper, G. Mul, J. Baltrusaitis, Phys.
Chem. Chem. Phys. 2014, 16, 12194–12201.
[9] J. Qiao, Y. Liu, F. Hong, J. Zhang, Chem. Soc. Rev. 2014, 43, 631–675.
[
10] M. Alvarez-Guerra, J. Albo, E. Alvarez-Guerra, A. Irabien, Energy Environ.
[
[
41] C. W. Li, J. Ciston, M. W. Kanan, Nature 2014, 508, 504–507.
42] D. Ren, Y. Deng, A. D. Handoko, C. S. Chen, S. Malkhandi, B. S. Yeo, ACS
Catal. 2015, 5, 2814–2821.
Sci. 2015, 8, 2574–2599.
[
[
11] I. Ganesh, Renewable Sustainable Energy Rev. 2014, 31, 221–257.
12] A. Goeppert, M. Czaun, J. P. Jones, G. K. Surya Prakash, G. A. Olah, Chem.
Soc. Rev. 2014, 43, 7995–8048.
13] M. Gattrell, N. Gupta, A. Co, J. Electroanal. Chem. 2006, 594, 1–19.
14] A. A. Peterson, F. Abild-Pedersen, F. Studt, J. Rossmeisl, J. K. Nørskov,
Energy Environ. Sci. 2010, 3, 1311–1315.
[
[
[
[
43] R. Kas, R. Kortlever, H. Yilmaz, T. M. Koper, G. Mul, ChemElectroChem
2
015, 2, 354–358.
[
[
44] C. Genovese, C. Ampelli, B. C. Marepally, G. Papanikolaou, S. Perathoner,
G. Centi, Chem. Eng. Trans. 2015, 43, 2281–2286.
45] A. Javier, J. H. Baricuatro, Y. G. Kim, M. P. Soriaga, Electrocatalysis 2015, 6,
[
[
[
[
[
[
[
[
15] K. P. Kuhl, E. R. Cave, D. N. Abram, T. F. Jaramillo, Energy Environ. Sci.
493–497.
2
012, 5, 7050–7059.
16] Y. Liu, Z. U. Wang, H. C. Zhou, Greenhouse Gases Sci. Technol. 2012, 2,
39–259.
17] J. Gascon, A. Corma, F. Kapteijn, F. X. Llabrꢆs I Xamena, ACS Catal. 2014,
, 361–378.
46] N. Gutiꢆrrez-Guerra, L. Moreno-Lꢅpez, J. C. Serrano-Ruiz, J. L. Valverde,
A. de Lucas-Consuegra, Appl. Catal. B 2016, 188, 272–282.
2
[47] Q. Wang, H. Dong, H. Yu, RSC Adv. 2014, 4, 59970–59976.
[48] Q. Wang, H. Dong, H. Yu, J. Power Sources 2014, 271, 278–284.
[49] J. Wu, P. P. Sharma, B. H. Harris, X. D. Zhou, J. Power Sources 2014, 258,
4
18] A. Domꢆnech, H. Garcꢇa, M. T. Domꢆnech-Carbꢅ, F. Llabrꢆs-i-Xamena, J.
Phys. Chem. C 2007, 111, 13701–13711.
19] R. S. Kumar, S. S. Kumar, M. A. Kulandainathan, Electrochem. Commun.
189–194.
[
50] S. Lee, H. Ju, R. Machunda, S. Uhm, J. Kwang Lee, H. Jin Lee, J. Lee, J.
Mater. Chem. A 2015, 3, 3029–3034.
51] K. W. Frese, J. Electrochem. Soc. 1991, 138, 3338–3344.
52] L. M. Aeshala, R. G. Uppaluri, A. Verma, J. CO Util. 2013, 3–4, 49–55.
2
53] E. Andrews, M. Ren, F. Wang, Z. Zhang, P. Sprunger, R. Kurtz, J. Flake, J.
Electrochem. Soc. 2013, 160, H841–H846.
2012, 25, 70–73.
[
[
[
20] R. Hinogami, S. Yotsuhashi, M. Deguchi, Y. Zenitani, H. Hashiba, Y.
Yamada, ECS Electrochem. Lett. 2012, 1, H17–H19.
21] T. Maihom, S. Wannakao, B. Boekfa, J. Limtrakul, J. Phys. Chem. C 2013,
117, 17650–17658.
[
54] D. P. Summers, S. Leach, K. W. Frese, J. Electroanal. Chem. 1986, 205,
22] I. Hod, P. Deria, W. Bury, J. E. Mondloch, C. W. Kung, M. So, M. D. Samp-
219–232.
son, A. W. Peters, C. P. Kubiak, O. K. Farha, J. R. Hupp, Nat. Commun.
[
[
55] A. Li, H. Wang, J. Han, L. Liu, Front. Chem. Sci. Eng. 2012, 6, 381–388.
56] L. M. Huang, H. T. Wang, J. X. Chen, Z. B. Wang, J. Y. Sun, D. Y. Zhao, Y. S.
Yan, Microporous Mesoporous Mater. 2003, 58, 105–114.
2
015, 6, 8304.
[
[
[
[
23] I. Hod, M. D. Sampson, P. Deria, C. P. Kubiak, O. K. Farha, J. T. Hupp, ACS
Catal. 2015, 5, 6302–6309.
[
[
[
57] S. Hausdorf, J. Wagler, R. Mossig, F. O. Mertens, J. Phys. Chem. A 2008,
24] N. Kornienko, Y. Zhao, C. S. Kley, C. Zhu, D. Kim, S. Lin, C. J. Chang, O. M.
Yaghi, P. Yang, J. Am. Chem. Soc. 2015, 137, 14129–14135.
25] W. Xia, A. Mahmood, R. Zou, Q. Xu, Energy Environ. Sci. 2015, 8, 1837–
112, 7567–7576.
58] M. Lanchas, S. Arcediano, A. T. Aguayo, G. Beobide, O. Castillo, J.
Cepeda, D. Vallejo-Sꢄnchez, A. Luque, RSC Adv. 2014, 4, 60409–60412.
59] S. Pꢆrez-YꢄÇez, G. Beobide, O. Castillo, J. Cepeda, A. Luque, A. T.
Aguayo, P. Romꢄn, Inorg. Chem. 2011, 50, 5330–5332.
1866.
26] S. Lin, C. S. Diercks, Y. B. Zhang, N. Kornienko, E. M. Nichols, Y. Zhao,
A. R. Paris, D. Kim, P. Yang, O. M. Yaghi, C. J. Chang, Science 2015, 349,
1208–1213.
[
[
27] M. R. Lohe, M. Rose, S. Kaskel, Chem. Commun. 2009, 6056–6058.
28] S. Xiang, L. Li, J. Zhang, X. Tan, H. Cui, J. Shi, Y. Hu, L. Chen, C. Y. Su, S. L.
James, J. Mater. Chem. 2012, 22, 1862–1867.
[29] L. Li, S. Xiang, S. Cao, J. Zhang, G. Ouyang, L. Chen, C. Y. Su, Nat.
Received: May 24, 2016
Commun. 2013, 4, 1174.
Published online on && &&, 0000
&
ChemSusChem 2016, 9, 1 – 11
www.chemsuschem.org
10
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

FULL PAPERS
Closing the loop: Metal–organic porous
materials are effective electrocatalysts
for the continuous electrochemical con-
J. Albo,* D. Vallejo, G. Beobide,
O. Castillo, P. CastaÇo, A. Irabien
&& – &&
version of CO to alcohols, a process
2
that could promote the transition to
a low-carbon economy. The modularity
of these systems yields many opportuni-
ties for further performance improve-
ments and opens new directions in
electrocatalysis.
Copper-Based Metal–Organic Porous
Materials for CO Electrocatalytic
2
Reduction to Alcohols
ChemSusChem 2016, 9, 1 – 11
www.chemsuschem.org
11
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ