Efficient conversion of NO into N2O by selected electroreduced heteropolyanions

Article abstract of DOI:10.1039/a708728d

It is shown that a series of one- and two-electron-reduced heteropolyanions of the Keggin- and Dawson-types convert quantitatively NO into N₂O in acidic aqueous media. The corresponding oxidized forms of the heteropolyanions are fully regenerated. Unsubstituted and lacunary heteropolyanions, as well as the substituted heteropolyanions of these two groups, prove active in this context.

Full text of DOI:10.1039/a708728d

View Article Online / Journal Homepage / Table of Contents for this issue
Efficient conversion of NO into N O by selected electroreduced
2
heteropolyanions
Abderrahman Belhouari,a Bineta Keita,a Louis Nadjo*,a and Roland Contantb
a L aboratoire dÏElectrochimie et de Photoelectrochimie (CNRS ER 549), Universite de Paris XI,
Batiment 420, 91405 Orsay cedex, France
b L aboratoire de Chimie des Metaux de T ransition (CNRS URA 419), Universite de Paris V I,
4, place Jussieu, 75252 Paris cedex 05, France
L e t t e r
It is shown that a series of one- and two-electron-reduced heteropolyanions of the Keggin- and Dawson-types
convert quantitatively NO into N O in acidic aqueous media. The corresponding oxidized forms of the
2
heteropolyanions are fully regenerated. Unsubstituted and lacunary heteropolyanions, as well as the substituted
heteropolyanions of these two groups, prove active in this context.
Transformaion quantitative de NO en N O par des heteropolyanions electroreduits. De nombreux
2
heteropolyanions ayant la structure de Keggin ou celle de Dawson presentent un comportement catalytique vis
a vis de la reduction de NO, au potentiel de leur premiere vague de reduction. Les composes satures ou
lacunaires, et les substitues appartenant a ces deux groupes de structures, se revelent efficaces pour cette
reduction. Sur quelques exemples dÏheteropolyanions, dont le premier couple redox est soit monoelectronique,
soit bielectronique, il a ete montre que la transformation de NO en N O est quantitative.
2
Among the large variety of catalytic processes in which het-
eropolyanions intervene efficiently in the liquid phase, our
group has been interested in the reduction of nitrogen oxides
for several years. Patents issued from this laboratory describe
electrocatalytic procedures based on oxometalates for the
reduction of nitrite anion and nitric oxide.1,2 Based on unam-
biguous cyclic voltammetric experiments, it was demonstrated
that a large variety of Keggin-type as well as Dawson-type
anions were suitable for this purpose. These families were
extended by their lacunary transition-metal-substituted deriv-
atives.1,2 Almost concomitantly, Toth and Anson3 published
work that also demonstrated that Keggin-type iron-
substituted polyoxometalates catalyze the reduction of nitrite
and nitric oxide; this occurs primarily through the formation
of an ironÈnitrosyl complex that is further reduced at more
negative potentials by electrons accumulated in the tungsten
framework. The product obtained was ammonia. These two
pioneering works aroused the interest of other groups.4h6
Most data from these groups corroborated the cyclic voltam-
metry work of Toth and Anson on iron-substituted hetero-
polyanions with an extension to Dawson-type anions4,6 and
to some other transition metal cations.6 ConÐrmation of the
data concerning the catalytic behaviour of the lacunary and
the parent unsubstituted heteropolyanions in our patents also
began to appear.4,6 None of them go beyond the results in the
patents. Thus, a series of problems remains to be solved. They
include, in particular, the identiÐcation and quantiÐcation of
the products obtained on the catalytic waves and the mecha-
nistic pathways through which the catalysis proceeds. These
are currently under investigation in our group. Recently, we
have published an attempt at an extension to suitable experi-
mental conditions to obtain sizeable amounts of chemicals.7
The main product, detected after electrolysis at potentials
more negative than the Ðrst redox couple of the hetero-
appropriate substrate, and also to the several reduction pos-
sibilities of nitrogen oxides, it seemed desirable to tackle the
problem by steps. In the following, we are interested in the
product (or products) obtained by the interaction of NO with
the Ðrst electronation step of any heteropolyanion. Nitric
oxide has been selected as the substrate to avoid the possible
disproportionation process associated with nitrite anion at
pH O 3.
The mounting and polishing of the glassy carbon electrode
(GC, Tokai, Japon, and V25 Le Carbone Lorraine, France)
and the electrochemical set-up have been described pre-
viously.7 Analyses were performed on a Girdel 3000 gas chro-
matograph and on
a
Perkin Elmer Lambda
5
spectrophotometer. The gas chromatograph was equipped
with a Porapack Q column for the analysis of N O and a 5 Ó
2
molecular sieve column for N . NH OH and N H have been
2
2
2
4
analyzed by cyclic voltammetry with a platinum electrode.
The gases are injected directly into the chromatograph using a
gas-tight syringe. Hyponitrous acid was analysed by UV-
visible spectroscopy at 207 nm after precipitation of
P W
O
6~ by NEt `.
2
18 62
4
Na SO , H SO and HClO (Prolabo) were used as
2
4
2
4
4
received. The heteropolyanions (HPAs) were synthesized, puri-
Ðed and characterized by one of us, and the syntheses of those
requiring new methods have been published.8,9 All the com-
pounds are listed below and their abbreviations will designate
the corresponding anions or the compounds themselves when
no confusion seems possible: Na P Mo
O
(P Mo ),
6 2
18 62
2
18
K P W
6 2 18 62
a FeP W ),
17
O
(P W ),
18
a K FeP W Mo O
15
a
and a FeP W
O
(a and
2
2
1
2
2
17 61
1
15
(a FeP W Mo ),
2
2
7
2
2
1
61
2
2
2
K
P W Mo O (P W Mo ), a and a K P W MoO
10 2 15 61 15 6 2 17 62
(a and a P W Mo), K P W Mo O (P W Mo ),
2 2 17 6 2 15 62 15
K P W Mo O (P W Mo ), K P W Mo O
6 2 14 62 14 6 2 12 62
(P W Mo ), a K LiP W (a P W ), a K P W
17 61 1 2 17
(SiMo ), (NH ) SiW
12 4 4
2
2
2
2
1
3
2
3
4
2
4
6
O
O
2
12
6
1
9
2
2
10 2 17 61
polyanions in this work, was N O.7
(a P W ), H SiMo
O
O
2
2 2 17
(SiW ).
12
4
12 40
12 40
The aim of the present communication is to try and detect,
and then to quantify, the products of the catalytic process.
Owing to the high capacity of heteropolyanions to accumulate
electrons in their framework and then deliver them to an
The reduction of NO by one- or two-electron-reduced
HPAs has been carried out under an argon atmosphere in
0.2 M Na SO /H SO (the pH l medium) by mixing two
2
4
2
4
New J. Chem., 1998, Pages 83È86
83

View Article Online
solutions: one saturated with NO and the other containing
the reduced HPA, usually called heteropolyblue (HPB). The
solution of HPB was obtained under an argon atmosphere, in
a two-compartment cell, by exhaustive electrolytic reduction
and coulometry (one- or two-electron reduction) by setting the
working electrode (glassy carbon sheet of 5 cm2 surface area)
potential at 60 mV negative to the peak potential observed in
cyclic voltammetry for the Ðrst reduction process of the HPA.
The HPB solutions are stored under argon in a Ñask tightly
stoppered by a septum. Pure NO (N20 grade, Air Liquide
France) was puriÐed by bubbling through a 4 M NaOH solu-
tion (to get rid of N O and NO impurities), then washed by
2
2
bubbling through fully deionized Millipore water. The whole
set-up was deaerated for at least one hour before introducing
NO for puriÐcation. This puriÐed gas was bubbled for at least
15 min into 50 ml of the thoroughly argon-deaerated pH l
solution. The NO-saturated solution was also protected
against dioxygen by tight septum stoppers. The NO was
assayed by UV-visible spectroscopy in the presence of
FeII(EDTA) as the complexing agent. The absorbance reading
at 430 nm indicates a concentration of 1.5 ] 10~3 M NO at
room temperature. The NO-saturated solution was withdrawn
by a gas-tight syringe and injected directly into the HPB solu-
tion. Completion of the reaction was indicated by the disap-
pearance of the HPB colour. Then, 2 to 3 ml of concentrated
HClO were added to the mixture to facilitate gas evolution.
4
Warming of the mixture to roughly 70 ¡C allows all the gas to
be collected.
The efficiency of several HPAs for the catalysis of NO was
Ðrst evaluated qualitatively in cyclic voltammetry experiments.
P Mo is chosen as a representative example. Its cyclic volt-
2
18
ammogram in aqueous media begins by at least two two-
electron reversible redox processes;10,11 only the Ðrst one is of
interest in the present study. Fig. 1 was recorded with a thor-
oughly polished glassy carbon electrode in the pH 1 medium
and shows the following main observations in sequence. The
Ðrst three redox waves of P Mo are well-behaved in the
absence of NO (dotted line, Fig. 1A). Upon saturation of the
HPA solution with NO, the current intensities of these waves
are remarkably enhanced (full line, Fig. 1A) and their polaro-
gram shapes can be used to evaluate the reaction rate con-
stant. However, further interpretation of the morphology of
these catalytic waves and the study of the associated mecha-
nistic pathways and rate constants are beyond the scope of
this paper and will be published elsewhere. In this short com-
munication, only the phenomena concerning the Ðrst wave of
the HPAs retain our interest and are better sketched in Fig.
1B. In the absence of HPA, nitric oxide is reduced close to the
supporting electrolyte discharge. The improvement in oper-
Fig. 1 Cyclic voltammograms in a pH 1 medium at a scan rate of 10
mV s~1. Working electrode: 5 mm diameter glassy carbon disk; solu-
tions deaerated with pure argon. (A) Dotted line: cyclic voltam-
2
18
mogram in supporting electrolyte with 10~4M P Mo
O
6~,
2
18 62
showing the Ðrst three redox sytems of the HPA under an argon
atmosphere (the current intensity has been multiplied by 5). Full line:
cyclic voltammogram of the preceding solution saturated with NO
(the current intensity has been multiplied by 2). The parameter
c \ [NO]/[HPA] \ 15. (B) (a) Supporting electrolyte alone; (b) sup-
porting electrolyte saturated with NO; (c) supporting electrolyte with
10~4M P Mo
O
6~, thoroughly deaerated with argon; and (d)
2
18 62
solution c saturated with NO
ational potential *E for the reduction of NO brought about
i
by the presence of the reduced HPA has been evaluated. For
Table 1 Improvement in the operational potential *E obtained for
i
this purpose, a current density of 53 lA cm~2 for the
the reduction of NO in the presence of heteropolyanions in pH 1
reduction of NO alone is chosen. *E is deÐned as the di†er-
mediuma
i
ence between the potential values at which this current density
HPAb
*E /Vc
CAT (%)c
is observed in the absence and in the presence of catalyst, pro-
vided the catalyst current is subtracted. The overall rate con-
stant evaluated from the polarogram-shaped catalytic wave12
in Fig. 1B is 4.6 ] 103 M~1 s~1; roughly the same value,
5 ] 103 M~1 s~1, was obtained by double potential step
i
P Mo
1.200
1.200
1.080
1.070
0.810
475
344
900
1160
1150
2
18
P W Mo
2
12
6
a P W Mo
1 2 17
SiMo
12
P W
2
18
chronocoulometry.13 Table
1
gives *E for
a
series of
i
a The *E values are determined for a current density of 53 mA cm~2
unsubstituted, lacunary and substituted HPAs, as well as the
results for the catalytic efficiency deÐned as CAT \ 100
[I (HPA ] NO) [ I (HPA)]/I (HPA).
i
after subtraction of the HPA current. This current density, in the
absence of heteropolyanion, is obtained at [0.8 V vs. SCE for NO-
saturated solutions. Working electrode surface area: 0.19 cm2.
c \ [NO]/[HPA] \ 15. b Only the Ðve HPAs for which the yields for
the NO reduction product were studied are listed. Numerous other
p
p
p
All the HPAs in Table 1, and the others cited in this com-
munication, show the same kind of catalytic e†ects on their
respective Ðrst redox couples in the presence of NO. The
choice of HPAs has been made to highlight the very large
HPAs,
namely,
P W Mo ,
P W Mo ,
a P W Mo,
2
14
4
2
15
3
2 2 17
a FeP W Mo , P W Mo , a and a FeP W , a and a P W
,
2
2
15
2
2
15
2
1
2
2
17
1
2 2 17
and SiW , have been used under appropriate conditons and give
12
*E values that could be achieved in the present experiments.
Many details are worth noting. We observe from the current
intensities of the catalytic waves that even the very favourable
very good results, which will be published elsewhere. c Scan rate of 10
i
mV s~1.
84
New J. Chem., 1998, Pages 83È86

View Article Online
Table 2 Reduction of NO into N O by selected one- and two-electron-reduced HPAsa
2
Electrolysis
potential vs.
SCE/V
Moles of
HPA reduced
(]106)
Moles of N O
produced
(]106)
2
HPA
C¡/mM
Y (%)b
One-electron-reducible HPAs
a P W Mo
1.7
3.5
1c
2
]0.1
8.5
17.5
5
10
30
4.4
8.8
2.4
5.2
15
103 ^ 6
100 ^ 6
96 ^ 6
1 2 17
]0.1
P W
[0.06
[0.06
[0.06
2
18
104 ^ 6
100 ^ 6
6
Two-electron-reducible HPAs
P Mo
1
2
0.74
]0.32
]0.18
]0.05
5.2
8.2
3.7
5.2
8
3.52
100 ^ 6
98 ^ 6
95 ^ 6
2
18
SiMo
12
P W Mo
2
12
6
a Solution: 5 ml of 0.2 M Na SO /H SO (pH 1 solution) with C¡ of HPA. The reduction of each HPA is carried out on its Ðrst redox wave. For
2
4
2
4
other details, see text. b Yield for the production of N O per mole of HPB. c Analysis and quantiÐcation by UV-visible spectroscopy.
2
potential shifts do not seem to reduce smoothly the catalytic
A
moles of N O
B
efficiency. This strongly suggests that an intermediate adduct
forms between HPA and NO. This Ðnding is being elaborated
on using a large variety of HPAs. Another point shows again
that unsubstituted compounds, as well as the parent lacunary
species, are active in this context. This was described in our
patents1,2 for Keggin- as well as Dawson-type anions, and
have recently been rediscovered and conÐrmed.4,6 Finally, we
emphasize the case of multiply substituted HPAs, a new strat-
egy initiated by our group,9,14 in which we study the substi-
tution, in various locations of the W framework, by selected
ensembles of identical or di†erent heterometals. In this
context, special emphasis will be put elsewhere on several
examples not shown here and, in particular, on
a FeP W Mo .
2
Y \ 100n
moles of reduced HPA
with n \ 2 and 1, respectively, for one- and two-electron-
reduced HPAs. The corresponding values are gathered in
Table 2 and call for several comments. Handling gaseous
species quantitatively requires particular care. Therefore, each
result in Table 2 is the average of at least three replications of
the same experiment. All the yields are close to 100%. The
obtention of such high yields is also favoured by the relatively
positive reduction potential of the present HPAs, the reduced
forms of which are not or are hardly sensitive to residual
oxygen in the electrolysis medium or to traces of oxygen that
could enter the cell during the di†erent stages of the experi-
ments. It was checked by cyclic voltammetry and UV-visible
spectroscopy that, in each experiment, the oxidized form of
the HPA is fully regenerated after visual completion of the
reaction, as indicated by the disappearance of the blue colour
of the reduced oxometalates. The yields in Table 2 also seem
to be independent of the reduction potential of the HPAs used
in this work, thus reinforcing our previous mechanistic propo-
sal of a ““chemical catalysisÏÏ pathway. The detailed succession
of electrochemical and chemical sequences occurring during
this reaction is under investigation. For instance, the
2
2
15
2
The results, in the following, are restricted to those HPAs
for which the identiÐcation and quantiÐcation of the products
of the catalytic process have been completed. The technique of
mixing the reduced heteropolyanion with NO has been pre-
ferred here. It ensures that a homogeneous reaction proceeds
cleanly, without any interference from the electrode potential
and material on possible reaction intermediates and/or pro-
ducts. However, in the present case, its main advantage is to
allow us to work with small volumes of HPB and pure NO,
which is rather expensive. With such small-volume, tightly
closed vessels, the determination of reaction yields on gaseous
species is accurate. It has been checked, previously7 and in the
present experiments, that continuous electrolysis of any of the
selected HPAs in the presence of NO does give the same qual-
reduction of P Mo
O
6~ is a two-electron process, even in
2
18 62
the absence of protons. The way in which P Mo
O
8~
2
18 62
delivers these two electrons to a substrate, simultanueously or
in sequence, is an interesting problem.
In conclusion, Keggin- and Dawson-type heteropolyanions
show catalytic behaviour for the reduction of NO in acidic
aqueous media. This study was restricted to their Ðrst redox
couples, for which the phenomenon is already visible.
Unsubstituted and lacunary HPAs, as well as the substituted
heteropolyanions of these two groups, prove active in this
system. Five selected HPAs featuring one- and two-electron-
reducible species were selected for further study in the present
work. The only product detected in the reduction of NO is
itative results. For quantitative measurements
a larger
volume, tightly closed electrochemical set-up, with NO bub-
bling, becomes necessary. Table 2 gathers the results. We have
used combinations of several analytical techniques; in repeat-
ed experiments, N O was the only chemical detected. To com-
2
plete the Ðrst reduction process observed in cyclic
voltammetry for each HPA, either one or two electrons are
required. As the same product is observed at the end of the
reaction, we assume that one-electron-reduced HPAs proceed
through the overall stoichiometry:
N O, obtained in quantitative yield, whatever the hetero-
polyanion. This work is interesting in its own right, but also
2
constitutes a Ðrst step in a new strategy that consists in trying
to accumulate several electrons on the Ðrst reduction waves of
heteropolyanions, in order to study the electrochemical and
chemical consequences of this multiple electronation possi-
bility on the reaction products of the catalytic reduction of
substrates like NO.
2NO ÈÈÈ N O ] H O ] 2HPA
(1)
2
2
2HPB, 2H`
and that two-electron-reduced HPAs proceed through:
`
1HPB, 2H
2NO ÈÈÈ N O ] H O ] HPA
(2)
2
2
There is ample indication in the literature that the reduction
of NO by one- and two-electron reductants yields mainly
Acknowledgements
This work was supported by the CNRS (ER 549 and URA
419) and by the Universities of Paris XI and Paris VI.
N O.15h17
2
The yield for the generation of N O is then deÐned, follow-
2
ing the two stoichiometries, as
New J. Chem., 1998, Pages 83È86
85

View Article Online
10 J. P. Ciabrini, R. Contant and J. M. Fruchart, Polyhedron, 1983, 2,
1229.
References
11 J. F. Garvey and M. T. Pope, Inorg. Chem., 1978, 17, 1115.
12 R. Greef, R. Peat, L. M. Peter, D. Pletcher and J. Robinson,
Instrumental Methods in Electrochemistry, Ellis Horwood;
England, 1985, Ch. 6.
1 B. Keita, L. Nadjo, R. Contant, M. Fournier and G. Herve
(CNRS), French Patent, 89/1,728, 1989.
2 B. Keita, L. Nadjo, R. Contant, M. Fournier and G. Herve
(CNRS), Eur. Patent Appl. EP 382,644, 1990; Chem. Abst., 1991,
114, 191882u.
13 B. Keita, K. Essaadi, L. Nadjo, R. Contant and Y. Justum, J. Elec-
troanal. Chem., 1996, 404, 271.
3 J. E. Toth and F. C. Anson, J. Am. Chem. Soc., 1989, 111, 2444.
4 S. Dong and M. Liu, J. Electroanal. Chem., 1994, 372, 95.
5 S. Dong, X. Xi and M. Tian, J. Electroanal. Chem., 1995, 385, 227.
6 T. McCormac, B. Fabre and G. Bidan, J. Electroanal. Chem., 1997,
427, 155.
14 B. Keita, A. Belhouari, L. Nadjo and R. Contant, J. Electroanal.
Chem., in the press.
15 D. E. Hendriksen and R. E. Powell, Inorg. Chem., 1982, 21, 1693.
16 F. T. Bonner, L. S. Dzelzkalns and J. A. Bonucci, Inorg. Chem.,
1978, 17, 2487.
7 B. Keita, A. Belhouari, L. Nadjo and R. Contant, J. Electroanal.
Chem., 1995, 381, 243.
17 F. T. Bonner and K. A. Pearsall, Inorg. Chem., 1982, 21, 1973.
8 R. Contant, Inorg. Synth., 1990, 27, 104.
9 R. Contant, M. Abbessi, J. Canny, A. Belhouari, B. Keita and L.
Nadjo, Inorg. Chem., 1997, 36, 4961.
Received 11th August 1997; Paper 7/08728D
86
New J. Chem., 1998, Pages 83È86