Electro-Assisted Reduction of CO2 to CO and Formaldehyde by (TOA)6[α-SiW11O39Co(-)] Polyoxometalate

Article abstract of DOI:10.1002/ejic.201500389

We report here on the multiproton-multielectron electrochemical reduction of CO₂ in homogeneous solution by using (TOA)₆[α-SiW₁₁O₃₉Co(-)] (TOA = tetraoctyl ammonium; - = vacant position in the coordination sphere of Co) as an electrocatalyst. First, the electrochemical behavior of (TOA)₆[α-SiW₁₁O₃₉Co(-)] was analyzed in detail by cyclic voltammetry in dichloromethane, studying the influence of the presence of protons and/or CO₂. These preliminary results were further used to optimize the conditions of electrolysis in terms of reduction potentials. Analysis of the electrolysis products in the gas and liquid phases show the formation of CO and HCHO without formation of H₂. Our results tend to show that the (TOA)₆[α-SiW₁₁O₃₉Co(-)] polyoxometalate is a catalyst for CO₂ electroreduction, with unique selectivity. The cobalt derivative of the silico-undecatungstate [α-SiW₁₁O₃₉Co(-)]^6- is a catalyst for the multiproton-multielectron electrochemical reduction of CO₂, with unique selectivity.

Full text of DOI:10.1002/ejic.201500389

FULL PAPER
DOI:10.1002/ejic.201500389
Electro-Assisted Reduction of CO to CO and
2
Formaldehyde by (TOA) [α-SiW O Co(_)]
6
11 39
Polyoxometalate
Marcelo Girardi,[
a,b,c,d,e]
Sébastien Blanchard,
[a]
[
b,c,d,e]
[f]
[f]
Sophie Griveau,
Philippe Simon, Marc Fontecave,
[
b,c,d,e]
[a]
Fethi Bedioui,
and Anna Proust*
Keywords: Polyoxometalates / Reduction / Carbon dioxide fixation / Electrochemistry / Cobalt / Tungsten
We report here on the multiproton-multielectron electro-
chemical reduction of CO in homogeneous solution by using
TOA) [α-SiW11 39Co(_)] (TOA = tetraoctyl ammonium; _ =
vacant position in the coordination sphere of Co) as an elec-
trocatalyst. First, the electrochemical behavior of (TOA)
α-SiW11 39Co(_)] was analyzed in detail by cyclic voltam-
2
ence of protons and/or CO . These preliminary results were
2
further used to optimize the conditions of electrolysis in terms
of reduction potentials. Analysis of the electrolysis products
in the gas and liquid phases show the formation of CO and
(
6
O
6
-
HCHO without formation of H
the (TOA) [α-SiW11 39Co(_)] polyoxometalate is a catalyst
for CO electroreduction, with unique selectivity.
2
. Our results tend to show that
[
O
6
O
metry in dichloromethane, studying the influence of the pres-
2
Introduction
ture. A variety of molecular catalysts have been tested under
[
7–11]
[12–16]
homogeneous
and supported conditions,
and in
Techniques that can be used to capture and transform
carbon dioxide are being developed in response to the
major environmental and economical issues associated with
the greenhouse effect and to the depletion of fossil fuel re-
the presence of proton sources, to activate the reaction. In
the case of iron–porphyrin electrocatalysts, detailed investi-
gations of the reduction mechanism have pointed out the
efficiency of the proton-coupled electron transfer (PCET)
that occurs when the proton source is close to the catalytic
serves. The reduction of CO to valuable fuels has thus been
2
[1,2]
addressed by using both photo-
and electrochemical ap-
[17,18]
center.
[
3–6]
proaches.
The direct electro-assisted reduction of CO2
Transition-metal-substituted polyoxometalates (TMSPs)
fulfill several of the relevant criteria for being competitive
on a bare electrode is, however, a kinetically slow process
that is characterized by large overpotentials because of the
multielectronic nature of the reactions and because of the
CO electrocatalysts: (1) polyoxometalates (POMs) are sol-
2
uble oxides of the early transition metals, and they are char-
acterized by great molecular diversity that encompasses
fundamental reorganization of the CO molecular struc-
2
(
multi)vacant species acting as all-inorganic ligands towards
[
a] Sorbonne Universités, UPMC Univ. Paris 06, Institut Parisien
[
19]
de Chimie Moléculaire, UMR CNRS 8232, Université Pierre et extra transition-metal cations,
especially abundant first-
Marie Curie,
row transition-metal cations, which, in turn, could fix CO2;
(
sensitive to the presence of protons
used as electrocatalysts for the multiproton-multielectron
reduction of nitrate to ammonia, or for hydrogen pro-
duction;
could promote the formation of hydrogen-bond networks in
the vicinity of the CO coordination center to favor PCET.
Whereas there have been a few articles dealing with the
7
5005 Paris, France
2) many POMs display reversible redox processes that are
E-mail: anna.proust@upmc.fr
[
20–24]
http://www.ipcm.fr/nouvelle-traduction-1-presentation-742
b] PSL Research University, Chimie ParisTech, Unité de
Technologies Chimiques et Biologiques pour la Santé,
and have been
[
[
[
[
[
75005 Paris, France
[
25–27]
c] INSERM, Unité de Technologies Chimiques et Biologiques
pour la Santé (no. 1022),
(3) although they are weak bases, POMs
75005 Paris, France
d] CNRS, Unité de Technologies Chimiques et Biologiques pour
la Santé UMR 8258,
2
75005 Paris, France
[
28–31]
activity of TMSPs in the photoreduction of CO2,
re-
e] Université Paris Descartes, Sorbonne Paris Cité, Unité de
Technologies Chimiques et Biologiques pour la Santé,
sults related to the electrocatalytic reduction of CO are still
expected. The decoloration of a blue toluene solution of
reduced [α-SiW11
gether with the modification of the cyclic voltammogram
of [α-SiW O Co(_)] recorded in dichloromethane under
CO2 have been reported.
2
75005 Paris, France
f] Laboratoire de Chimie des Processus Biologiques, UMR 8229
Collège de France-CNRS-Université Pierre et Marie Curie
Université Paris 06,
6+n)–
O
39Co(_)](
upon CO bubbling, to-
2
1
1 Place Marcelin Berthelot, 75231 Paris Cedex 05, France
6–
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201500389.
11 39
[
32,33]
However, no more infor-
Eur. J. Inorg. Chem. 2015, 3642–3648
3642
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
mation regarding the nature of the products or the electro- initial cathodic peak located at –1.55 V increased (almost
catalytic character of the reaction were given. We thus de- sixfold) with an anodic shift of the onset of the peak of
cided to reinvestigate this reaction in more detail to evaluate 0.1 V, whereas a second electroreductive event was observed
the potential of TMSPs in the electro-assisted reduction of at –1.90 V. On the backward scan, a modest increase of the
CO2.
Kozik et al. have reported on the coordination of CO
anodic peak at –1.34 V was visible, whereas a new anodic
process appeared at –1.08 V. The latter is associated with
the second reduction process located at –1.9 V. Similar ob-
2
[32]
6
–
to nonreduced [α-SiW O Co(_)] in nonpolar solvent.
1
1
39
Given that electrochemical studies could not be performed servations to those made by cyclic voltammetry were ob-
in toluene, other solvents were tested. Whereas preliminary tained by using the Au/Hg RDE technique. Indeed, the in-
experiments showed no interaction of the POM with CO2 tensity of the first reduction wave increased by a factor of
in acetonitrile, electrolysis of a CO -saturated dichloro- almost two with an ill-defined plateau, whereas a second
2
methane solution of (TOA) [α-SiW O Co(_)] (TOA = ill-defined cathodic wave appeared with an onset at approxi-
6
11 39
tetraoctyl ammonium) at –1.5 V vs. Hg Cl /Hg (1 m LiCl) mately –1.9 V (Figure S1).
2
2
on a Hg working electrode gave a blue solution. Analysis
of the gas above the solution revealed the formation of CO,
albeit in low faradic yield; no other reduction product could
be detected. To improve the activity of the TMSP under
study, small amounts of water and acid (TBAHSO ) were
4
added. The electro-assisted reaction led again to the forma-
tion of CO and also, as determined by the results of a color-
imetric Nash test at the end of the electrolysis, to the forma-
tion of formaldehyde, a four-electron-four-proton reduction
product of CO . Whereas the two-electron reduction prod-
2
–
ucts CO and HCOO are commonly identified, HCHO is
very seldom observed with molecular catalysts,[ one note-
34]
worthy contribution being the production of HCHO in
CO -saturated aqueous solution at glassy carbon electrodes
2
modified with electropolymerized films of metal–terpyrid-
[
35]
ine complexes.
These findings encouraged us to explore
in more detail the ability of (TOA) [α-SiW O Co(_)] to
6
11 39
act as an electrocatalyst for the four-electron reduction of
CO2.
Results
Evaluation of the Electro-Assisted Processes by Cyclic
Voltammetry
First, the electrochemical behavior of (TOA) [α-Si- Figure 1. Comparison of the cyclic voltammograms with an Hg/Au
6
–
1
electrode at a scan rate of 100 mVs in CH
2
Cl
2
+ 0.1 m TBABF
4
W O Co(_)] was assessed in dichloromethane (CH Cl )
solution by using cyclic and hydrodynamic voltammetric
methods. Figure 1 shows the cyclic voltammograms ob-
1
1
39
2
2
solution under argon either without (curve I) or with (TOA)
α-SiW11 39Co(_)] (curve II) and under CO with (TOA)
6
[
O
2
6
[α-Si-
W
11
O
39Co(_)] (curve III). * Despite 30 min of Ar bubbling, traces
tained at an amalgamated Au/Hg electrode for a 1 mm solu- of O
tion of (TOA) [α-SiW O Co(_)] in CH Cl containing
2
could not be avoided.
6
11 39
2
2
0
.1 m TBABF as supporting electrolyte, either in the pres-
Electrochemical analysis of (TOA) [α-SiW O Co(_)]
4
6 11 39
ence or absence of CO . In the absence of CO (Figure 1, was then carried out in the presence of up to 100 mm H O.
2
2
2
curve II), a quasi-reversible redox system, attributed to In all cases, under argon as well as under CO , the electro-
2
(
TOA) [α-SiW O Co(_)] was observed at –1.47 V vs. chemical responses were identical to those obtained without
6 11 39
Hg Cl /Hg (1 m LiCl) followed by a second, weak, irrevers- added water (Figure 1 curves II and III, and Figure 2,
2
2
ible reduction peak at –1.80 V [note that in the following curves I and II).
the nature of the reference electrode Hg Cl /Hg (1 m LiCl) On the other hand, addition of up to 10 equiv. of
2
2
will be omitted for simplification]. The electrochemical TBAHSO as a proton donor, together with 40 equiv. of
4
analysis using an Au/Hg rotating disk electrode (RDE) H O, in the absence of CO , induced progressive and large
2
2
showed a well-defined reduction wave located at –1.5 V (see modification of the shape of the cyclic voltammograms
the Supporting Information, Figure S1). Under CO , a (Figure 2, curve III). Under these conditions, on the for-
2
significant modification of the electrochemical behavior ward scan, the initial reduction peak at –1.55 V was posi-
of (TOA) [α-SiW O Co(_)] was observed (Figure 1, tively shifted (with an onset at ca. –1.25 V) concomitant
6
11 39
curve III). First, on the forward scan, the intensity of the with an increase of its intensity, whereas a second reduction
Eur. J. Inorg. Chem. 2015, 3642–3648
3643
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Electrolysis under Ar
First, reduction of 1 mm (TOA) [α-SiW O Co(_)] in
6
11 39
CH Cl + 0.1 m TBABF solution was carried out under
2
2
4
argon at controlled potentials (from –1.5 to –1.8 V). In all
cases, no color change of the initial green solution and a
very low electrolysis current were observed, indicating that
no reduction of the POM occurred under these conditions
(data not shown).
Electrolysis under CO2
Under a saturated CO atmosphere, a clear increase of
2
the initial intensity of the electrolysis current at –1.5 V of a
1
0 mm (TOA) [α-SiW O Co(_)] in CH Cl2 + 0.1 m
6 11 39 2
TBABF solution was observed, together with a deep-blue
4
coloration, thus attesting to the reduction of the POM
framework. The darkening of the coloration paralleled a
progressive decrease of the current intensity from ca.
1
8
000 μA to reach a plateau at ca. 100 μA after 2 h. After
0 min, a charge of 3.7 C was passed, and the analysis of
the gas in the head space indicated production of CO
(
3.5ϫ10–6 mol), corresponding to a faradic yield of 18%.
No other reduction product (H , HCOOH, HCHO, ...) was
2
detected.
Figure 2. Cyclic voltammograms of a 1 mm solution of (TOA)
α-SiW11O39Co(_)] in CH Cl + 0.1 m TBABF and 40 mm water
at 100 mV/s without TBAHSO under argon (curve I) and under
CO (curve II) and with 10 mm TBAHSO under argon (curve III)
and under CO (curve IV).
6
-
Electrolysis under CO and Effect of Protons
2
[
2
2
4
To increase the efficiency of the reaction, 10 equiv. of
acid (10 mm) and 40 equiv. of water were added to the POM
(1 mm) in the cathodic compartment. In this case, upon re-
4
2
4
2
duction under CO at –1.5 V, together with the detection of
2
CO, a positive Nash test proved the formation of the four-
peak was observed at –1.64 V. A third reduction wave, pre-
sumably associated with the reduction of protons, was also
observed starting at approximately –1.8 V. On the reverse
scan, an anodic peak located at approximately –1 V was
observed that was associated with the cathodic peak at
electron-reduced product of CO , formaldehyde. Notably,
2
neither H nor other CO reduction products were detected.
2
2
Under these conditions, the time-dependent formation of
CO is shown in Figure 3. After a short lag phase, CO was
produced continuously to reach 3.7ϫ10–5 mol (TON = 3.7)
after 66 h of electrolysis. Given that during the same time a
charge of 52 C was passed, this corresponded to a faradic
–
1.64 V. Addition of CO induced only moderate modifica-
2
tion of this cyclic voltammogram, yet a slight anodic shift
ca. 30 mV) of all electrochemical features and a reproduc-
(
ible weak decrease of the intensity of the reoxidation peak,
now located at –0.97 V (Figure 2, curve IV) were observed.
Similar observations were made by using an Au/Hg RDE.
It was observed that, in the presence of TBAHSO , the in-
4
tensity of the reduction wave of (TOA) [α-SiW O Co(_)]
6
11 39
increased, whereas subsequent addition of CO did not sig-
2
nificantly modify the shape or the position of the reduction
wave at –1.67 V (see the Supporting Information, Fig-
ure S1). This analysis determined the working potential
used in electrolysis experiments.
Electrolysis
To gain a better understanding of the redox activity of
(
TOA) [α-SiW O Co(_)], a range of electrolysis experi-
6 11 39
ments were performed with a mercury pool as working elec-
trode. The products present in both gaseous and liquid
phases were analyzed (see the Experimental Section and the
Supporting Information for detailed information about
analysis methodologies).
Figure 3. Evolution of CO production (GC analysis of the head
space) as a function of electrolysis time. Reaction conditions: elec-
trolysis at –1.5 V vs. Hg
SiW11 39Co] + 40 mm H O + 10 mm TBAHSO
solution in CH Cl saturated in CO
2
Cl
2
/Hg (1 m LiCl) of 1 mm (TOA)
6
[
O
2
4
+ 0.1 m TBABF
4
2
2
2
.
Eur. J. Inorg. Chem. 2015, 3642–3648
3644
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Table 1. Amount of formaldehyde detected after 4.5 h electrolysis at –1.5 V vs. Hg
2
Cl
2
/Hg (1 m LiCl) at an Hg electrode of a 1 mm solution
of (TOA)
TBAHSO
6
[α-SiW11
2 2 4 2 2
O39Co(_)] in CH Cl + 0.1 m TBABF containing 40 mm H O and under CO as a function of the concentration of
4
, electrolysis charge passed and the treatment of the solution at the end of the electrolysis.
Concentration of TBAHSO
4
10 mm
100 mm
Charge injected
4.8 C
6.5 C
Experimental conditions
Solution exposed to air
at the end of electrolysis CO
Solution kept under
for 12 h after
Solution exposed to air at
the end of electrolysis
Solution kept under CO
2
for 12 h after electrolysis
2
electrolysis
2.6
Amount of formaldehyde formed [μmol] 1.3
1.0
2.8
efficiency of 13% for the catalytic CO production. Keeping to air, typically 3.6ϫ10–7 mol for a total charge of 4 C and
the electrolysis conditions rigorously analogous during such a faradic yield of 3.5%, which is four times less than for a
a long period is difficult; thus, some variations in CO pro- comparable electrolysis with the same charge under CO2.
duction was found to occur over 66 h; in the worst case, Considering that the reduced POMs could dehalogenate
–
5
2
.5ϫ10 mol of CO was detected (TON 2.5).
dichloromethane in the presence of air to produce
As indicated above, the only other detectable product HCHO,[36] an electrolysis was performed under air in the
was formaldehyde (HCOH). However, its production suf- presence of protons and water: Under these conditions,
fered from a lack of reproducibility: several electrolyses 5ϫ10–6 mol of formaldehyde were detected, with a faradic
were carried out under the same conditions under a static efficiency of 20% as 10 C were injected into the cathodic
CO atmosphere over 4.5 h, after which the blue reaction compartment.
2
medium was opened to air and the amount of formaldehyde
Table 1 shows the results obtained when, at the end of
was quantified. The results are presented in Figure 4. The electrolysis, the resulting solution was divided into two
data show that the amount of HCHO varied from parts: one was left under CO for 12 h before analysis, while
2
–
7
–6
2
.1ϫ10 to 2.2ϫ10 mol, spanning an order of magni- the second was opened to air at the end of electrolysis. The
tude, with the faradic yield varying from 25% (low charge latter fraction turned reddish within a few minutes, whereas
and high formaldehyde content) to 0.8% (high charge, low fading under CO was much slower, taking a few hours to
2
formaldehyde content). The influence of the proton concen- proceed. The analysis of the two solutions showed that the
tration is also shown in Figure 4, with an experiment car- concentration of formaldehyde for the sample kept under
ried out in the presence of 100 mm TBAHSO (tenfold in- CO was much higher than that for the sample immediately
4
2
+
crease in H ): a charge of 6.5 C was injected after 4.5 h, exposed to air after electrolysis, irrespective of the concen-
and an amount of 1.0ϫ10–6 mol of HCHO was detected.
tration of the acid (and the electric charge passed) during
the electrolysis.
Finally, 23 h electrolysis of 10 mL of 1 mm POM +
4
0 mm H O + 10 mm TBAHSO in CH Cl under CO led
2 4 2 2 2
–5
to the formation, after air reoxidation, of 1.75ϫ10 mol
of HCHO after the injection of a charge of 16.1 C and a
40% faradic yield. This corresponds to a TON of 1.75, indi-
cating that the reaction is catalytic, albeit again slow.
Discussion
The shape of the cyclic voltammograms of (TOA)6
[
SiW O Co(_)] in CH Cl is modified upon addition of
11 39 2 2
CO with a sixfold increase of the reduction peak intensity.
2
[
33]
This is consistent with a previous report
authors attributed this increase to an electrocatalytic re-
by the POM. However, no reduction of the
in which the
Figure 4. Amount of formaldehyde detected after 4.5 h electrolysis
at –1.5 V vs. Hg
solution of (TOA)
under CO containing 40 mm H
of the electrolysis charge passed. Circles: 10 mm TBAHSO
00 mm TBAHSO
2
Cl
2
/Hg (1 m LiCl) at an Hg electrode of a 1 mm duction of CO
39Co(_)] in CH Cl + 0.1 m TBABF
O and TBAHSO , as a function
; square:
2
6
[α-SiW11
O
2
2
4
POM occured when the electrolysis was conducted under
argon. This suggests that the low-intensity voltammetric
cathodic peak observed under argon may not result from
the reduction of (TOA) [SiW O Co(_)]. To generate
2
2
4
4
1
4
.
6
11 39
Control experiments were also performed: under argon,
a coordinatively unsaturated cobalt center, K [α-Si-
6
in the presence of water and acid, the solution turned blue W O Co(OH )] is transferred from an aqueous solution
11
39
2
upon reduction, indicating reduction of the POM in the to a toluene solution containing TOABr as a phase-transfer
presence of protons. Neither CO nor H were detected in reagent: the pink organic solution containing the (TOA)6
2
these experiments. Surprisingly, a small amount of form- [α-SiW O Co(OH )] is then dried under vacuum, and the
11
39
2
aldehyde was detected after 4.5 h electrolysis and opening resulting oily product is redissolved in dichloromethane.
Eur. J. Inorg. Chem. 2015, 3642–3648
3645
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
However, it has been shown that during the phase transfer However, this latter process is much more intricate. Indeed,
from water to toluene, partial protonation of POMs could formaldehyde is also formed, at a very low yield, after elec-
[
37–39]
occur.
This was confirmed by preliminary pulsed EPR trolysis of a (TOA) [SiW O Co(_)] solution under argon
6 11 39
experiments that established the presence of protons in the and exposure to air before analysis of the electrolysis prod-
vicinity of the paramagnetic Co center,[40] and by colorimet- ucts. The most likely source of carbon in this case is the
ric titration with TBAOH of a freshly prepared (TOA) [Si- solvent, CH Cl . It was reported that reduced POMs can
6
2
2
W O Co(_)] redissolved in anhydrous acetonitrile, which indeed promote dehalogenation reactions of CH Cl2
1
1
39
2
accounted for the presence of approximately 0.1 equiv. H+ through various mechanisms, leading to a range of prod-
[
36]
per POM. Thus, we suggest that the weak cathodic peak ucts, among which is formaldehyde. These reactions have
observed under argon results from the reduction of approx- different rates, the fastest for formaldehyde production be-
imately 10% of the POMs that are monoprotonated ing in the presence of air. In our case, post-electrolysis treat-
{
(TOA) [HSiW O Co(_)]} during the phase-transfer pro- ment relies on aerobic reoxidation of the reduced blue solu-
5 11 39
cess, whereas the major fraction of unprotonnated POMs tion before titration of the formaldehyde, and this could
cannot be reduced under these conditions. explain the detection of formaldehyde even when the elec-
Bubbling CO , a π-acceptor ligand, or adding protons to trolysis is carried out under argon. Although care was taken
2
the media shifts the reduction potential toward more posi- to keep the cell air-tight, one can also not rule out, to some
tive values. The POM-based reduction wave thus appears extent, air contamination during the electrolysis. These de-
at –1.55 and –1.25 V in the presence of CO or protons, halogenation processes could explain the lack of reprodu-
2
respectively. The observed increase in intensity is thus not cibility of formaldehyde production during the 4.5 h elec-
the signature of an electrocatalytic process, but simply the trolysis, with a high charge associated with high dehaloge-
appearance of POM-based waves.
nation and low HCHO amount.
The reduction processes of the POM in the presence of
However, when the electrolyzed solution is reoxidized un-
both a proton source and CO appears to be dominated by der CO
rather than air, the formaldehyde content increases
the former. The modifications of the cyclic voltammograms (Table 1). This finding, together with the fact that the form-
positive shift of the reduction peak and decrease in inten- aldehyde yield increases when the charge is low (little de-
sity of the associated anodic peak related to the POM redox halogenation processes), clearly indicates that the detected
process) are indicative that the reduced protonated POM formaldehyde in the electrolysis under CO mainly origi-
degradation.
2
2
(
2
still interacts with CO2.
nates from CO rather than from CH Cl
2 2 2
During the electrolysis in the presence of CO , CO is
2
always detected. Interestingly, its production is not sensitive
to the presence of air (either during or at the end of the Conclusions
electrolysis), in contrast to that of formaldehyde (see be-
The TMSP (TOA) [SiW O Co(_)] promotes the elec-
6
11 39
low). Moreover, addition of protons does not affect the pro-
duction of CO, because preliminary experiments without
added water or protons yielded CO at approximately the
same rate. In both cases, with or without water and protons,
a lag time is observed, corresponding to the formation of
the blue solution. It is thus reasonable to suggest that the
first step of the electrolysis is the reduction of the POM
framework, followed by a slow electron transfer from the
trocatalytic reduction of CO to CO and, more interest-
2
ingly, performs the four-electron/four-proton reduction of
the former to a more valuable product, formaldehyde. This
catalyst is selective towards CO reduction, because no H2
2
production was detected during electrolysis at various tim-
escales. These results reinforce our conviction that, due to
their proton- and electron-storage properties, POMs are at-
tractive candidates to promote the multiproton/multielec-
POM to CO . Formation of CO, a two-electron reduction
2
tron reduction of CO to useful chemicals.
+
2
product of CO , in the absence of H , has previously been
2
observed in other systems and is thought to proceed ac-
cording to the following equation:[
41,42]
Experimental Section
–
2–
2
CO
2
+ 2 e Ǟ CO
3
+ CO
Chemicals: Analytical grade solvents were purchased from Sigma–
Irrespective of the step following the reduction of (TOA)6- Aldrich. Dichloromethane was distilled from calcium hydride.
[
SiW O Co(CO )], either electron transfer to CO or reac- Tetrabutylammonium hydrogensulfate (TBAHSO , 97%) was pur-
11 39 2 2
4
tion with another CO molecule, it is likely to be slow, be- chased from Sigma–Aldrich, tetraoctylammonium bromide
cause fading of the blue color under CO proceeds over a
few hours, in agreement with the fact that no increase of
the current of the voltammetric peak could be observed.
Nevertheless, the system appears to be quite robust, because
it remains active after more than 66 h electrolysis, yielding
a TON of 3.7. (TOA) [SiW O Co(_)] is thus able to pro-
2
(
TOABr, 97%) and sodium tetrafluoroborate (NaBF
supplied by Alpha-Aesar and cobalt(II) chloride hexahydrate
CoCl ·6H O, 98%) by Labo Sciences. Tetrabutylammonium tetra-
fluoroborate (TBABF ) was prepared by mixing 200 mL of 1.3 m
NaBF with 200 mL of 1 m TBAHSO in water under vigorous
stirring. A white precipitate was formed, and additional water (ca.
00 mL) was added. The precipitate was filtered through a frit and
4
, 97%) were
2
(
2
2
4
4
4
6
11 39
4
mote the electrocatalytic reduction of CO to CO.
2
washed until the wastewater reached pH ≈ 5. The white powder
In the presence of protons and water, the formation of was redissolved in dichloromethane and dried with MgSO . The
4
CO is concomitant with the production of formaldehyde. organic phase was filtered and finally concentrated under vacuum
Eur. J. Inorg. Chem. 2015, 3642–3648
3646
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
to obtain a white powder that was stored under argon. Azoviolet
was used as colorimetric indicator for the titration of protons in
distilled acetonitrile. The pink solution became purple after the
the mercury pool and diminished by supplementary gas bubbling
if needed. Solutions were constantly stirred throughout the bulk
electrolysis experiments.
8 2
equivalence was reached. K [α-SiW11O39]·13H O was prepared ac-
Analysis of Electrolysis Products:[11] It must be noted that all analy-
sis measurements were corrected for the solution volumes at the
end of the bulk electrolysis experiments. Indeed, with time, di-
chloromethane evaporated slowly and was absorbed by the natural
rubber stoppers under static conditions. H measurements were
2
performed by gas chromatography with a Shimadzu GC-2014
equipped with a Quadrex column and a Thermal Conductivity De-
cording to a reported procedure.[
43]
Synthesis of K
[α-SiW11 39Co(OH
as follows: K
dissolved in a solution of CoCl
7 mL) at 70 °C and allowed to return to room temperature. The
6
[α-SiW11
)]·15H
[α-SiW11
O
39Co(OH
O was adapted from a previous re-
39]·xH O (10 g, 3.10 mmol) was
·6H O (0.81 g, 3.4 mmol) in water
2 2
)]·15H O: The synthesis of
K
port
6
O
2
2
[
44]
6
O
2
2
2
(
product crystallized at room temperature overnight as large, red,
shiny crystals. The crystals were then washed with cold water and
dried. Yield: 9.618 g (94.8%).
2 4
tector, using N as a carrier gas. Other gases (CO, CH ) were mea-
sured with a Shimadzu GC-2010 Plus gas chromatograph, fitted
with a Restek Shin Carbon column, helium carrier gas, a methan-
izer and a flame-ionization detector. Gas chromatography cali-
brations were made by sampling known volumes of the various
gases. The typical volume of gas injected was 50 μL. Formate and
oxalate concentrations were determined by ionic chromatography
with a Metrohm 883 Basic IC plus ion-exchange chromatography
instrument, a Metrosep A Supp 5 column and a conductivity detec-
tor. A typical measurement required the sampling of 1 mL of solu-
Synthesis of (TOA)
α-SiW11 39Co(_)] was adapted from a previous report
lows: K [α-SiW11 39Co(OH )]·15H O (327 mg, 0.1 mmol) was dis-
6 6
[α-SiW11O39Co(_)]: The synthesis of (TOA) -
[32]
[
O
as fol-
6
O
2
2
solved in water (10 mL). To this red solution was added, whilst
stirring, 10 mL of a 60 mm solution of TOABr (0.6 mmol, 328 mg)
in toluene. The aqueous phase rapidly became colorless, and the
organic phase became red. The latter was separated from the aque-
ous solution, and toluene was removed under vacuum to give an
oily green product.
–
1
tion, followed by a 100-fold dilution in deionized 18 MΩcm
water, evaporation of dichloromethane, and injection of 20 μL into
the instrument. Methanol concentration was determined with a
Shimadzu GC-2010 Plus gas chromatograph fitted with a ZB-WAX
Plus column, helium as a carrier gas, and a flame-ionization detec-
Electrochemical Experiments: Analytical experiments were con-
ducted with a Princeton Applied Research Potentiostat (Model
263A). The studies were performed with a classical three-electrode
tor. Formaldehyde concentration was determined by using the
cell. Hydrodynamic voltammetries were performed with a rotating
disk electrode (RDE) (Hach Lange, France) at a rotation speed of
[45]
Nash colorimetric test
Once blue [α-SiW11
with a V-670 Jasco spectrophotometer.
(6+n)–
O
39Co]
was reoxidized to red [α-Si-
1000 rpm and at a scan rate of 10 mV/s. Cyclic voltammetry was
6–
W O39Co(OH )] , no variation of the formaldehyde concentra-
11 2
performed at a scan rate of 100 mV/s, unless otherwise noted. In
both cases, a mercury drop amalgamated on a gold electrode was
used as working electrode. The gold electrode (1 mm diameter) was
roughly polished, and the mercury drop was deposited by simple
immersion of the electrode in a mercury reservoir. The amalgam
was renewed between each series of measurements. The counter
tion over time was observed; nevertheless, samples were stored in
the refrigerator to avoid evaporation of dichloromethane. Formal-
dehyde was extracted with distilled water prior to analysis (see the
Supporting Information). Approx. 10% error must be considered.
electrode was a platinum wire, and a Calomel (1 m LiCl) electrode Acknowledgments
separated by a Vycor tip was used as reference (Radiometer Analyt-
ical, France). Dichloromethane containing TBABF
used as electrolytic solution. All studied solutions were purged by
strongly bubbling inert gas or CO for 20 min prior measurements,
4
(0.1 m) was
This work was supported by the French National Research Agency
ANR) (Carbiored ANR-12-BS07-0024-03).
(
2
the gas being previously bubbled in dichloromethane so as to limit
solvent evaporation. Finally, a prescan was performed before mea-
surement to clean the surface of the mercury drop and diminish
desorption-current phenomena.
[
[
1] C. D. Windle, R. N. Perutz, Coord. Chem. Rev. 2012, 256,
562–2570.
2] K. Sekizawa, K. Maeda, K. Domen, K. Koike, O. Ishitani, J.
Am. Chem. Soc. 2013, 135, 4596–4599.
2
Bulk Electrolysis: Bulk electrolysis experiments were carried out
with a custom-made two-compartment cell. A 1.5 cm diameter [4] M. R. Dubois, D. L. Dubois, Acc. Chem. Res. 2009, 42, 1974–
[3] J. M. Saveant, Chem. Rev. 2008, 108, 2348–2378.
1982.
pool of mercury was used as the working electrode, and the calomel
m LiCl reference electrode (Radiometer Analytical, France) was
[
[
[
5] E. E. Benson, C. P. Kubiak, A. J. Sathrum, J. M. Smieja, Chem.
1
Soc. Rev. 2009, 38, 89–99.
separated by a Vycor tip. The counter electrode used was a plati-
num wire separated from the working electrode by a frit (poros-
ity 4). The volume of the solution used in the cathodic compart-
ment was 10 mL (6 mL in the anodic compartment) and the typical
headspace volume was 21 mL. Saint-Gobain Performance Plastics
France natural rubber folding skirt stoppers were used to close the
cell. A Potentiostat VersaSTAT 3 was used to carry out most of the
bulk electrolysis experiments, except when gas production follow up
was needed, in which case a Bio-Logic SP 300 potentiostat con-
nected to a booster card was used. Bulk electrolysis solutions were
6] C. Finn, S. Schnittger, L. J. Yellowlees, J. B. Love, Chem. Com-
mun. 2012, 48, 1392–1399.
7] J. Hawecker, J. M. Lehn, R. Ziessel, J. Chem. Soc., Chem. Com-
mun. 1984, 328–330.
[8] Z. F. Chen, C. C. Chen, D. R. Weinberg, P. Kang, J. J. Concep-
cion, D. P. Harrison, M. S. Brookhart, T. J. Meyer, Chem. Com-
mun. 2011, 47, 12607–12609.
[
9] J. M. Smieja, E. E. Benson, B. Kumar, K. A. Grice, C. S. Seu,
A. J. M. Miller, J. M. Mayer, C. P. Kubiak, Prog. Natl. Acad.
Sci. USA 2012, 109, 15646–15650.
[
10] P. Kang, T. J. Meyer, M. Brookhart, Chem. Sci. 2013, 4, 3497–
purged by strongly bubbling inert gas or CO
prior to measurements, the gas was previously bubbled in dichloro-
methane so as to limit solvent evaporation. Traces of O were un-
avoidable and contributed to residual current. These traces of O
2
for at last 20 min
3502.
[11] N. Elgrishi, M. B. Chambers, V. Artero, M. Fontecave, Phys.
Chem. Chem. Phys. 2014, 16, 13635–13644.
2
2
[12] M. Beley, J. P. Collin, R. Ruppert, J. P. Sauvage, J. Chem. Soc.,
Chem. Commun. 1984, 1315–1316.
were checked through cyclic voltammetry directly performed with
Eur. J. Inorg. Chem. 2015, 3642–3648
3647
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
[
[
13] S. Chardon-Noblat, A. Deronzier, R. Ziessel, D. Zsoldos, J.
Electroanal. Chem. 1998, 444, 253–260.
[31] M. T. Colvin, M. Kozik, S. H. Szczepankiewicz, J. Phys. Chem.
B 2006, 110, 10576–10580.
14] M. Bourrez, F. Molton, S. Chardon-Noblat, A. Deronzier, An-
gew. Chem. Int. Ed. 2011, 50, 9903–9906; Angew. Chem. 2011,
[32] S. H. Szczepankiewicz, C. M. Ippolito, B. P. Santora, T. J.
Van de Ven, G. A. Ippolito, L. Fronckowiak, F. Wiatrowski, T.
Power, M. Kozik, Inorg. Chem. 1998, 37, 4344–4352.
[33] J. Paul, P. Page, P. Sauers, K. Ertel, C. Pasternak, W. Lin, M.
Kozik, Polyoxometalate Chemistry for Nano-Composite Design
(Eds.: T. Yamase, M. Pope), Springer, USA, 2002, pp. 205–215.
[34] S. Bontemps, S. Sabo-Etienne, Angew. Chem. Int. Ed. 2013, 52,
10253–10255; Angew. Chem. 2013, 125, 10443–10445.
[35] J. A. R. Sende, C. R. Arana, L. Hernandez, K. T. Potts, M. Ke-
shevarzk, H. D. Abruna, Inorg. Chem. 1995, 34, 3339–3348.
[36] D. Sattari, C. L. Hill, J. Am. Chem. Soc. 1993, 115, 4649–4657.
[37] D. K. Lyon, W. K. Miller, T. Novet, P. J. Domaille, E. Evitt,
D. C. Johnson, R. G. Finke, J. Am. Chem. Soc. 1991, 113,
7209–7221.
123, 10077–10080.
[
[
15] N. Elgrishi, S. Griveau, M. B. Chmabers, F. Bedioui, M. Fonte-
cave, Chem. Commun. 2015, 51, 2995–2998.
16] T. M. Huan, E. S. Andreiadis, J. Heidkamp, P. Simon, E. De-
rat, S. Cobo, G. Royal, A. Bergman, P. Starsser, H. Dau, V.
Artero, M. Fontecave, J. Mater. Chem. A 2015, 3, 3901–3907.
17] C. Costentin, S. Drouet, M. Robert, J. M. Saveant, Science
[
[
[
[
[
[
[
[
2012, 338, 90–94.
18] C. Costentin, S. Drouet, G. Passard, M. Robert, J. M. Saveant,
J. Am. Chem. Soc. 2013, 135, 9023–9031.
19] O. Oms, A. Dolbecq, P. Mialane, Chem. Soc. Rev. 2012, 41,
7497–7536.
20] M. T. Pope, E. Papaconstantinou, Inorg. Chem. 1967, 6, 1147–
[38] H. Weiner, J. D. Aiken, R. G. Finke, Inorg. Chem. 1996, 35,
1152.
7905–7913.
21] K. Maeda, S. Himeno, T. Osakai, A. Saito, T. Hori, J. Elec-
troanal. Chem. 1994, 364, 149–154.
22] D. M. Way, A. M. Bond, A. G. Wedd, Inorg. Chem. 1997, 36,
[39] O. A. Kholdeeva, M. P. Vanina, M. N. Timofeeva, R. I. Maksi-
movskaya, T. A. Trubitsina, M. S. Melgunov, E. B. Burgina, J.
Mrowiec-Bialon, A. B. Jarzebski, C. L. Hill, J. Catal. 2004,
226, 363–371.
[40] S. Blanchard, H. Vezin, unpublished results.
[41] B. P. Sullivan, C. M. Bolinger, D. Conrad, W. J. Vining, T. J.
Meyer, J. Chem. Soc., Chem. Commun. 1985, 1414–1415.
[42] J. W. Raebiger, J. W. Turner, B. C. Noll, C. J. Curtis, A. Mied-
aner, B. Cox, D. L. DuBois, Organometallics 2006, 25, 3345–
3351.
2826–2833.
23] J. I. Nambu, T. Ueda, S. X. Guo, J. F. Boas, A. M. Bond, Dal-
ton Trans. 2010, 39, 7364–7373.
24] C. Rinfray, S. Renaudineau, G. Izzet, A. Proust, Chem. Com-
mun. 2014, 50, 8575–8577.
[25] M. Sadakane, E. Steckhan, Chem. Rev. 1998, 98, 219–237.
[26] J. Zhang, J. K. Goh, W. T. Tan, A. M. Bond, Inorg. Chem.
2
006, 45, 3732–3740.
[43] A. Tézé, G. Hervé, R. G. Finke, D. K. Lyon, Inorganic Synthe-
ses (Ed.: A. P. Ginsberg), John Wiley & Sons Inc. 1990, 85–96.
[44] J. L. Samonte, M. T. Pope, Can. J. Chem. 2001, 79, 802–808.
[45] S. B. Jones, C. M. Terry, T. E. Lister, D. C. Johnson, Anal.
Chem. 1999, 71, 4030–4033.
[
[
[
27] B. Keita, L. Nadjo, J. Mol. Catal. A 2007, 262, 190–215.
28] T. Yamase, M. Sugeta, Inorg. Chim. Acta 1990, 172, 131–134.
29] A. M. Khenkin, I. Efremenko, L. Weiner, J. M. L. Martin, R.
Neumann, Chem. Eur. J. 2010, 16, 1356–1364.
[
30] J. Ettedgui, Y. Diskin-Posner, L. Weiner, R. Neumann, J. Am.
Received: April 10, 2015
Chem. Soc. 2011, 133, 188–190.
Published Online: July 8, 2015
Eur. J. Inorg. Chem. 2015, 3642–3648
3648
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim