Pd-containing organopolyoxometalates derived from Dawson polyoxometalate [P2W15V3O62]9-: Lewis acidity and dual site catalysis

Article abstract of DOI:10.1021/ol501644c

Grafting of a palladium complex to the Dawson vanadotungstate polyanion [P₂W₁₅V₃O₆₂]^9- via an organic ligand generates a large family of pincer-type hybrid polyoxometalates. The palladium-POM derivatives have dual catalytic properties. Unlike their parent inorganic polyanions, they catalyze allylations while retaining their oxidant character, which leads to single-pot dual site catalysis. This opens a new route for multicatalytic reactions.

Full text of DOI:10.1021/ol501644c

Letter
pubs.acs.org/OrgLett
Pd-Containing Organopolyoxometalates Derived from Dawson
Polyoxometalate [P2W15V3O62]9−: Lewis Acidity and Dual Site
Catalysis
Benoît Riflade,† David Lachkar,§ Julie Oble,†,‡ Joaquim Li,† Serge Thorimbert,*,†,‡
Bernold Hasenknopf,*,†,‡ and Emmanuel Lacote*,†,§,∥
̂
†Sorbonne Universites
́
, UPMC Univ Paris 06, Institut Parisien de Chimie Molec
‡CNRS, UMR 8232, IPCM, 4 place Jussieu, 75005 Paris, France
§ICSN-CNRS, Bat
iment 27, 1 avenue de la Terrasse, 91198 Gif-sur-Yvette Cedex, France
∥Universite
de Lyon, Institut de chimie de Lyon, UMR 5265 CNRS-Universite Lyon I-ESCPE Lyon, 43 Bd du 11 novembre 1918,
́
ulaire, 4 place Jussieu, 75005 Paris, France
̂
́
́
69616 Villeurbanne, France
S
* Supporting Information
ABSTRACT: Grafting of a palladium complex to the Dawson vanadotungstate polyanion
[P2W15V3O62]9− via an organic ligand generates a large family of pincer-type hybrid polyoxometalates.
The palladium-POM derivatives have dual catalytic properties. Unlike their parent inorganic
polyanions, they catalyze allylations while retaining their oxidant character, which leads to single-pot
dual site catalysis. This opens a new route for multicatalytic reactions.
ewis acidic polyoxometalates (POMs) have emerged
Lrecently as a family of POMs with new catalytic
properties.1 Generally, POMs are nanosized polyanionic
clusters of highly oxidized early transition metals (most often
VV, MoVI, and/or WVI) and oxido ligands,2 thus not Lewis acid
candidates. To overcome this problem one can insert one (or
several) Lewis acidic cation(s) into a lacunary framework1a−g or
tether it1h via a suitable ligand in an organo-polyoxometalate.3
The Dawson POM [P2W15V3O62]9− is an oxidation catalyst
(as are its organo-hybrids).4,5 Yet it is not Lewis acidic. Its
oxido ligands can support transition metal ions in structures
that are stable in organic solution,5a,6 however with no open
coordination sites for catalysis. Unfortunately, no
[P2W15V3O62]9−-derived monolacunary species exists, where
one could insert a “naked” Lewis acidic cation, so another
strategy had to be pursued to give Lewis acidity to
[P2W15V3O62]9−.
Figure 1. General strategy for installing Lewis acidity onto 1.
We recently showed that Pd-pincer POMs such as 3 could be
prepared via capping of the 3-vanadium triad of
[P2W15V3O62]9− (1) with picolinamide-diols, followed by
pincer-complex formation via palladation/C−H activation
(Figure 1).7 The CO insertion into the POM framework
ensures that the palladium is electronically conjugated (and not
simply tethered8) to the POM framework. The new POMs
were assessed as catalysts for Mizoroki−Heck couplings.
However, the basic conditions led to degradation of the
clusters.
PCP and related pincer Pd(II) complexes are electrophilic
enough to catalyze the allylation of imines.9 Our Pd-POMs
have a different geometry, but we surmised that they would also
be electrophilic. We thus decided to investigate them as
potential Lewis acids. We also hoped that the oxidative
Received: April 17, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ol501644c | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
properties of the vanadotungstic framework might be retained
in this new type of Lewis acidic POMs, that the general
solubility of POMs would allow easy catalyst recycling, and that
the ligand on the POM would allow tuning of the catalytic
activity which is not achievable with purely inorganic structures.
Our first task was to extend the scope of the palladation with
regard to the pyridine and the carbon undergoing the C−H
insertion. We previously evidenced that an aryl group in the 2-
position of the pyridine could lead to pincer complexes on this
substituent, rather than onto the substituent on the diol.7 So we
focused on pyridines with no aryl substituent at the 2-position
in order to limit the structural space to the former type of
pincers (see 3, Figure 1).
The hybrid ligands 2a−j were prepared in good yields from
(TBA)5H4[P2W15V3O62] (1) and various amido-diols, under
microwave heating at 80 °C for 35 min in the presence of 30
mol % of para-toluenesulfonic acid (PTSA) in dimethyl-
acetamide (DMAc).10
Figure 2. Structures of the POM-palladacycles obtained.
A typical POM ligand (2a) was isolated in 82% yield after
two successive precipitations in diethyl ether and ethanol and
centrifugation. No trace of protonation of the pyridine was
detected. POMs 2a−j were obtained as TBA salts with a
varying amount of dimethylammonium cations. The ratio was
determined by NMR and elemental analysis. Full analytical data
can be found in the Supporting Information (SI).
1131.5 for {TBA(CH3CN)[P2V3W15O62C11H13N2Pd]}4−) and
m / z 1 0 6 7 . 7 ( c a l c d 1 0 6 6 . 4 { T B A ( C H 3 C N ) -
[P2V3W15O62C11H13N2Pd]}4−), confirming its structure (see
SI for the full description).
Preparative experiments were also carried out with the POM
ligands 2b−j (Table 1; see Figure 2 for the structures and the
SI for full description). All POMs 3a−g were isolated and
purified by successive precipitations in diethyl ether/ethanol
and centrifugations, and fully characterized by IR, multinuclear
NMR, ESI-MS, and elemental analysis.
The palladations were carried out as described.7a In a typical
experiment (Table 1, entry 1), 1 equiv of Pd(OAc)2 was added
Table 1. Preparation of the Pd-Containing POMs: Scope
The C−H activation worked with both electron-withdrawing
and -donating substituents on the pyridine, independently of
the position of the substituents (compare Table 1, entries 2, 6
and 3−5; see Figure 2 for structures). An isoquinoline
derivative also led to the cyclopalladated POM derivative (see
3g, 84%, entry 7).
The steric hindrance at the C−H site was gradually increased
to evaluate the preference for 5- or 6-membered ring formation.
The propyl derivative 3h led to 5-membered palladacycle 3h in
90% yield (entry 8). However, an additional methyl group, as in
3i, redirected the C−H insertion to the primary methyl moiety,
likely because the tertiary carbon of the isobutyl group is not
accessible for steric reasons (entry 9, 95%). A 6-membered
palladacycle was also isolated in 98% yield from benzyl
derivative 2j via an aromatic C−H insertion (entry 10).
Thus, the grafting strategy to build Pd-containing organo-
hybrids of 1 is highly versatile. Interestingly, the mass spectra of
3a−j indicated that one molecule of solvent (MeCN) was
present. Thus, the compounds have one coordination site at the
Pd center potentially available for catalytic purposes.
entry
POM
pyridine
2-pyridyl
R
product, yield (%)
1
2
2a
2b
2c
2d
2e
2f
Et
Et
Et
Et
Et
Et
Et
3a, 87
3b, 75
3c, 82
3d, 90
3e, 93
3f, 99
3g, 84
3h, 90
3i, 95
3j, 98
(6-Cl)-2-pyridyl
(6-Me)-2-pyridyl
(6-OMe)-2-pyridyl
(4-OMe)-2-pyridyl
(5-CF3)-2-pyridyl
1-isoquinolyl
3
4
5
6
7
2g
2h
2i
8
2-pyridyl
n-Pr
9
2-pyridyl
i-Bu
10
2j
2-pyridyl
CH2Ph
The palladium center on the POM might complex Lewis
bases, such as imines or aldehydes, as a traditional Lewis acid
through exchange of the solvent. Or it could undergo
nucleophilic attack from allylmetal reagents at the electrophilic
Pd site to generate η1-allyl-Pd(II) POMs, which could allylate
imines and/or aldehydes. In both cases, the reactivity would
derive from the presence of the accessible electrophilic center.
The allylation of imines thus appeared to be an excellent
benchmark to determine whether the Pd embedding approach
can bring new properties to tungstovanadate ion 1 without
POM degradation.
to 2a in acetonitrile at rt. After 5 min all the palladium acetate
was solubilized. The two 31P signals of the Dawson structure
moved slightly from δ = −5.7, −11.9 ppm in 2a to δ = −6.0,
−12.0 ppm in 3a. The 1H NMR signal of the amide proton at δ
= 10.41 ppm in 2a disappeared. In addition, the AB system
corresponding to the two CH2OV (δ ≈ 5.5 ppm in 2a) turned
into a broad singlet (δ = 5.48 ppm in 3a). The 13C spectra
indicated that the methyl of the organic ligand in 2a (δ = 7.6
ppm) had been turned into a methylene group in 3a (δ = 25.3
ppm). Thus, a 5-membered palladacycle was formed (see
Figure 2). Major ESI-MS signals for 3a are at m/z 1131.3 (calcd
In a typical experiment, N-(phenylmethylene)-benzene-
sulfonamide 4a was heated at 40 °C in a microwave oven
B
dx.doi.org/10.1021/ol501644c | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
with 2 equiv of allyltributylstannane in the presence of 4 mol %
of 3a in DMF (Table 2, entry 1).
structure. The trends observed with the substituents are
consistent with an increased or lowered reactivity of the
electrophiles toward the allylmetal intermediates.
In order to learn more about the reaction and the possible
influence of the pyridine substitution we selected four of the
organo-Pd POMs with no (3a, 3g), donating (3e) and
withdrawing (3f) substituents for further examination. The
conversion of the allylation of diphenylimine was followed over
a period of 2 h (Figure 3). After that time, the reaction was over
Table 2. Catalytic Allylation of Sulfonylimines and
Aldehydes
a
Figure 3. Conversion as a function of time for the allylation of
diphenylimine catalyzed by (a) 3a (4 mol %, deep blue diamonds); (b)
3a (0.4 mol %, orange dots); 3e (4 mol %, light blue stars); 3f (4 mol
%, red squares); 3g (4 mol %, olive green triangles).
with unsubstituted pyridine and isoquinoline POM-Pd 3a and
3g. The selectivity of the reaction is total, corresponding to a
TON of 25 and an average TOF of 12.5 h−1. For the POM-Pd
containing polar substituents (3e, 3f), the reaction was still
selective, but did not reach completion within 2 h (the catalyst
was however still active (TON = 17.5; TOF = 8.75 h−1).
Finally, decreasing the loading resulted in a less efficient
reaction, which stopped at about 24% conversion (however,
since conversion dropped less sharply than the loading, the
TON improved to 60; TOF = 30 s−1). This confirms our
hypothesis that the activity of the catalysts might be tuned via
the substitution of the organic ligand of the organo-POM.
Finally, POM 3a was precipitated after addition of diethyl
ether at the end of the reaction. This led to its recycling by
centrifugation. It could be reused up to eight times without loss
of its catalytic activity (e.g., 89% yield at the seventh run; see
Table S1 in the SI).
We next investigated whether 3a retained the catalytic
properties of the vanadotungstates in the presence of the
palladium atom. Bifunctional substrate 5h including both a
sulfide and an aldehyde moiety was chosen as model substrate
to probe this issue. It was placed in the presence of 4 mol % of
3a and submitted to allylstannane at 120 °C. Then H2O2 was
added in the same pot, and the reaction was left overnight at rt
(Scheme 1). Sulfonyl alcohol 8 was isolated in 82% yield with
no trace of alcohol oxidation. No reaction took place when the
a
All yields are isolated yields. Unless otherwise noted, conversion into
b
product was 100% after 2 h. Starting material was recovered intact.
c
d
Conversion was 70% after 2 h. 0.4 mol % of 3a was used.
Conversion was 24% after 2 h. The reaction was heated at 80 °C.
The reaction was heated at 120 °C.
e
f
g
The sulfonylimines with either electron-withdrawing (Table
2, entry 3) or -donating groups (entries 4−5, 7−10) delivered
the expected products. The allylamines were obtained in good
to excellent yields in all cases but one. Only the allylation of
phenyl-tert-butylimine failed (entry 6). The reaction did not
proceed in that case, which may be a consequence of both the
steric hindrance and the electron-donating character of the
bulky tert-butyl substituent on the nitrogen. When the loading
of 3a was lowered to 0.4 mol %, the reaction proceeded but
stopped at about 24% conversion (see below). Isoquinolyl
POM-Pd 3g behaved exactly as 3a (entry 9), while 3e−f were
less reactive (70% conversion after 2 h, entries 7−8).
Aromatic aldehydes were examined next. Again, those were
allylated without a problem (entries 11−18). An electron-
withdrawing nitro substituent on the aromatic ring helped the
reaction (entry 14), while strong electron-donating substituents
required us to heat the reaction more (entries 15−17). Finally,
POM 2a which has the same structure as 3a except for the
absence of any Pd atom left the imine unchanged.
Scheme 1. Dual Catalytic Activity of 3a
The Pd-POMs thus catalyze allylations of imine and
aldehyde electrophiles at the Pd centers in the hybrid POM
C
dx.doi.org/10.1021/ol501644c | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(3) (a) Long, D.-L.; Tsunashima, R.; Cronin, L. Angew. Chem., Int.
Ed. 2010, 49, 1736. (b) Dolbecq, A.; Dumas, E.; Mayer, C. R.;
Mialane, P. Chem. Rev. 2010, 110, 6009. (c) Proust, A.; Matt, B.;
Villaneau, R.; Guillemot, G.; Gouzerh, P.; Izzet, G. Chem. Soc. Rev.
2012, 41, 7605.
(4) (a) Hou, Y.; Hill, C. L. J. Am. Chem. Soc. 1993, 115, 11823.
(b) Zheng, H.; Newkome, G. R.; Hill, C. L. Angew. Chem., Int. Ed.
2000, 39, 1771. (c) Morgan, J. R.; Cloninger, M. J. J. Polym. Sci. Part.
A: Polym. Chem. 2005, 43, 3059. (d) Pradeep, C. P.; Long, D.-L.;
Newton, G. N.; Song, Y. F.; Cronin, L. Angew. Chem., Int. Ed. 2008, 47,
4388. (e) Han, Y.; Xiao, Y.; Zhang, Z.; Liu, B.; Zheng, P.; He, S.;
Wang, W. Macromolecules 2009, 42, 6543. (f) Li, J.; Huth, I.;
same sequence was carried out without the POM. Similarly,
alcohol 7h was left mostly unchanged when reacted with H2O2
in the absence of any POM. Only minor amounts of the
corresponding sulfoxide and sulfone were obtained. Finally,
benzyl alcohol was not oxidized in the reaction conditions.
The combined observations prove that POM 3a has dual
catalytic properties. From what precedes the Pd center is
responsible for the allylation since 2a does not catalyze it, and it
is not poisoned by the sulfide. As for other vanadotungstates,4a,i
the vanadium triad is likely responsible for the chemoselective
oxidation, although its exact mechanism is not known at this
stage.
To conclude, a surface organometallic approach can be used
to add new catalytic properties to the Dawson vanadotungstate
polyanion [P2W15V3O62]9−. Pyridine-derived diol-amide ligands
installed on the POM lead to electrophilic pincer-like POM-
Pd(II) hybrids. The latter are recyclable catalysts for the
allylation of imines and aldehydes, as well as oxidation catalysts.
Both types of catalysis are compatible, and a dual catalytic, one-
pot process was devised. In addition, the coordination sphere of
the Pd center can be easily modified, leading to reactivity
tuning. Further work will focus on extending the concept of
dual-site POM catalysts toward other structures and reactions,
as well as to design a chiral environment for asymmetric
versions of the reactions studied.
̂
Chamoreau, L.-M.; Hasenknopf, B.; Lacote, E.; Thorimbert, S.;
Malacria, M. Angew. Chem., Int. Ed. 2009, 48, 2035. (g) Pradeep, C. P.;
Misdrahi, M. F.; Li, F.-Y.; Zhang, J.; Xu, L.; Long, D.-L.; Liu, T.;
Cronin, L. Angew. Chem., Int. Ed. 2009, 48, 8309. (h) Pradeep, C. P.;
Li, F.-Y.; Lydon, C.; Miras, H. N.; Long, D.-L.; Xu, L.; Cronin, L.
Chem.Eur. J. 2011, 17, 7472. (i) Oble, J.; Riflade, B.; Noel, A.;
̈
Malacria, M.; Thorimbert, S.; Hasenknopf, B.; Laco
̂
te, E. Org. Lett.
2011, 13, 2990.
(5) (a) Weiner, H.; Finke, R. G. J. Am. Chem. Soc. 1999, 121, 9831.
(b) Keita, B.; Contant, R.; Mialane, P.; Secheresse, F.; de Oliveira, P.;
́
Nadjo, L. Electrochem. Commun. 2006, 8, 767. For other recent
catalytic oxidations by vanadium-based polyoxometalates, see:
(c) Khenkin, A. M.; Leitus, G.; Neumann, R. J. Am. Chem. Soc.
2010, 132, 11446. (d) Mizuno, N.; Kamata, K. Coord. Chem. Rev.
2011, 255, 2358. (e) Miao, W.-K.; Yan, Y.-K.; Wang, X.-L.; Xiao, Y.;
Ren, L.-J.; Zheng, P.; Wang, C.-H.; Ren, L.-X.; Wang, W. ACS Macro
Lett. 2014, 3, 211.
(6) (a) Nomiya, K.; Hasegawa, T. Chem. Lett. 2000, 410. (b) Nomiya,
K.; Kasahara, Y.; Sado, Y.; Shinohara, A. Inorg. Chim. Acta 2007, 360,
2313. (c) Nomiya, K.; Torii, H.; Kato, C. N.; Sado, Y. Chem. Lett.
2003, 32, 664. (d) Rapko, B. M.; Pohl, M.; Finke, R. G. Inorg. Chem.
1994, 33, 3625.
ASSOCIATED CONTENT
* Supporting Information
■
S
Procedures for synthesis, full characterization of all new
compounds described; recycling of the catalysts. This material
is available free of charge via the Internet at http://pubs.acs.org.
(7) (a) Riflade, B.; Oble, J.; Chenneberg, L.; Derat, E.; Hasenknopf,
̂
B.; Lacote, E.; Thorimbert, S. Tetrahedron 2013, 69, 5572. A similar
AUTHOR INFORMATION
Corresponding Authors
C−H activation has been suggested as a key step for remote C−H
functionalizations; see: (b) He, G.; Chen, G. Angew. Chem., Int. Ed.
2011, 50, 5192. (c) Zaitsev, V. G.; Shabashov, D.; Daugulis, O. J. Am.
Chem. Soc. 2005, 127, 13154. (d) Nadres, E. T.; Daugulis, O. J. Am.
Chem. Soc. 2012, 134, 7. (e) Ting, C. P.; Maimone, T. J. Angew. Chem.,
Int. Ed. 2014, 53, 3115.
■
*E-mail: serge.thorimbert@upmc.fr.
*E-mail: bernold.hasenknopf@upmc.fr.
*E-mail: emmanuel.lacote@univ-lyon1.fr.
(8) (a) Bar-Nahum, I.; Cohen, H.; Neumann, R. Inorg. Chem. 2003,
42, 3677. (b) Berardi, S.; Carraro, M.; Iglesias, M.; Sartorel, A.;
Scorrano, G.; Albrecht, M.; Bonchio, M. Chem.Eur. J. 2010, 16,
10662. (c) Tong, J.; Wang, H.; Cai, X.; Zhang, Q.; Ma, H.; Lei, Z.
Appl. Organomet. Chem. 2014, 28, 95.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Univ. P. et M. Curie, CNRS, ICSN and the French
National Research Agency (ANR, Grant 08-PCVI-0005) for
funding. FR2769 is acknowledged for technical assistance.
(9) (a) Solin, N.; Kjellgren, J.; Szabo,
2003, 42, 3656. (b) Solin, N.; Kjellgren, J.; Szabo,
Soc. 2004, 126, 7026. (c) Piechaczyck, O.; Cantat, T.; Mez
Floch, P. J. Org. Chem. 2007, 72, 4228. (d) Johansson, R.; Wendt, O.
Dalton Trans. 2007, 4, 488. (e) Selander, N.; Szabo, K. J. Chem. Rev.
́
K. J. Angew. Chem., Int. Ed.
K. J. J. Am. Chem.
ailles, N.; Le
́
́
́
REFERENCES
■
2011, 111, 2048. (f) Nakamura, H.; Iwama, H.; Yamamoto, Y. J. Am.
Chem. Soc. 1996, 118, 6641.
(1) (a) Boglio, C.; Lemier
̀
e, G.; Hasenknopf, B.; Thorimbert, S.;
te, E.; Malacria, M. Angew. Chem., Int. Ed. 2006, 45, 3324.
y, P.; Hasenknopf, B.; Thorimbert, S.;
Laco
̂
(10) The amido-diols 2a−j were synthesized in 2 to 5 steps from the
pyridine-containing acyl chlorides, diethylaminomalonate, and the
suitable aryl or alkyl bromides.
(b) Boglio, C.; Micoine, K.; Rem
́
Laco
̂
te, E.; Malacria, M.; Alfonso, C.; Tabet, J.-C. Chem.Eur. J. 2007,
13, 5426. (c) Kikukawa, Y.; Yamaguchi, S.; Nakagawa, Y.; Uehara, K.;
Uchida, S.; Yamaguchi, K.; Mizuno, N. J. Am. Chem. Soc. 2008, 130,
15872. (d) Suzuki, K.; Sugawa, M.; Kikukawa, Y.; Kamata, K.;
Yamaguchi, K.; Mizuno, N. Inorg. Chem. 2012, 51, 6953. (e) Absillis,
G.; Parac-Vogt, T. N. Inorg. Chem. 2012, 51, 9902. (f) Vanhaecht, S.;
Absillis, G.; Parac-Vogt, T. N. Dalton Trans. 2012, 41, 10028.
(g) Dupre,
Laco
te, E.; Hasenknopf, B.; Malacria, M. Chem.Eur. J. 2010, 16,
7256. (h) Dupre, N.; Brazel, C.; Fensterbank, L.; Malacria, M.;
Thorimbert, S.; Hasenknopf, B.; Laco
te, E. Chem.Eur. J. 2012, 18,
12962.
(2) Long, D.-L.; Burkholder, E.; Cronin, L. Chem. Soc. Rev. 2007, 36,
105.
́ ́
N.; Remy, P.; Micoine, K.; Boglio, C.; Thorimbert, S.;
̂
́
̂
D
dx.doi.org/10.1021/ol501644c | Org. Lett. XXXX, XXX, XXX−XXX