Grafting of Secondary Diolamides onto [P2W15V3O62]9- Generates Hybrid Heteropoly Acids

Article abstract of DOI:10.1002/anie.201510954

The Dawson tungstovanadate [P₂W₁₅V₃O₆₂]^9- can be grafted to secondary diolamides. The electron-withdrawing character of the polyanion increases the acidity of the amide proton, leading to an organo-po

Full text of DOI:10.1002/anie.201510954

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201510954
German Edition: DOI: 10.1002/ange.201510954
Polyoxometalates
Grafting of Secondary Diolamides onto [P2W15V3O62]9À Generates
Hybrid Heteropoly Acids
David Lachkar, Debora Vilona, Elise Dumont,* Moreno Lelli,* and Emmanuel Lacôte*
Abstract: The Dawson tungstovanadate [P2W15V3O62]9À can
be grafted to secondary diolamides. The electron-withdrawing
character of the polyanion increases the acidity of the amide
proton, leading to an organo-polyoxometalate, which can be
used as a Brønsted organocatalyst. High-field NMR and DFT
modeling indicate that the amide proton stays on the nitrogen
and that the exalted acidity derives from the interaction
between the organic and inorganic parts of the organo-
polyoxometalate. The amide-inserted vanadotungstates thus
form a new family of (hybrid) heteropolyacids, offering new
perspectives for the application of POM-based catalysis in
organic synthesis.
ones.[3] One approach focused on the generation of mixed
salts from the fully inorganic heteropoly acids and more or
less complex amines.[4] This had led to very efficient systems.
However, the POM itself plays no significant role in what
essentially are organocatalytic reactions with complex coun-
terions.
To create truly hybrid heteropoly acids in which activity
derives from the orbital interplay between the inorganic and
organic parts, we considered the potential of Dawson
tungstovanadate [P2W15V3O62]9À. This polyanion can be
derivatized by diolamides to generate conjugated organo-
POMs.[5] POM/organic electronic communication is enabled
by the insertion of the carbonyl into the polyoxometallic
structure. We surmised that since POMs are good electron
reservoirs, they might stabilize the anions derived from
deprotonation of the amide on the hybrid structures by
accepting the extra negative charge of the conjugated amide.
This delocalization of the conjugated charge over the whole
POM would thus increase the acidity of the amide (Figure 1).
With the help of high-field NMR and computational model-
ing we show here that this is the case.
P
olyoxometalates (POMs) are nanometric polyanionic clus-
ters, where highly oxidized early transition metals—typically
WVI, VV, or MoVI—are bound by oxido bridges. The stability
of these polyanions makes their protonated forms—called
heteropoly acids—highly acidic. Heteropoly acids are widely
used catalysts. They can be employed in heterogeneous
catalysis, in industrial processes, and scores of other applica-
tions.[1] However, their structural diversity and tunability is
relatively narrow and they also have a limited solubility in
organic solvents. Thus, their use with sensitive organic
substrates remains limited.[2] New families of heteropoly
acids and new strategies to generate them would thus be
welcome.
In that context, POM organo-hybrids are especially
attractive. They combine the properties of the POMs and
the great structural space of organic structures, which can be
used to install new properties on POMs, or to tune existing
[*] D. Lachkar, Dr. E. Lacôte
ICSN CNRS
1 avenue de la Terrasse, 91198 Gif-sur-Yvette Cedex (France)
D. Vilona, Dr. E. Lacôte
Univ Lyon, CPE Lyon, UniversitØ Claude Bernard Lyon1, CNRS,
Institut de chimie de Lyon, C2P2 UMR 5265
69616 Villeurbanne (France)
E-mail: emmanuel.lacote@univ-lyon1.fr
D. Vilona, Dr. M. Lelli
Figure 1. Deprotonation of the POM-diolamides.
Univ Lyon, UniversitØ Claude Bernard Lyon1, ENS Lyon, CNRS,
Institut de chimie de Lyon, ISA-CRMN UMR 5280
69100 Villeurbanne (France)
E-mail: moreno.lelli@ens-lyon.fr
The phenyl (1a), p-nitro (1b) and p-methoxy (1c)
derivatives have been previously prepared in our group
(Figure 2).[5a] We added electron-poor p-trifluoromethyl 1d to
the array of POMs. Together, these four representative POMs
(with the general formula TBA5[P2V3W15O59{(OCH2)2C-
D. Vilona, Dr. E. Dumont
Uni Lyon, ENS de Lyon, CNRS, UniversitØ Lyon 1, Laboratoire de
chimie, UMR 5182
69342 Lyon (France)
E-mail: elise.dumont@ens-lyon.fr
=
(Et)NHC( O@POM)R}]) cover the whole range of elec-
Supporting information for this article can be found under:
http://dx.doi.org/10.1002/anie.201510954.
tronic effects on the conjugated aromatic ring, which might
Angew. Chem. Int. Ed. 2016, 55, 5961 –5965
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5961

Angewandte
Communications
Chemie
for 2a,b. Comparison with authentic samples showed that the
latter corresponded to remaining traces of the free diolamide
ligands of 1a, 2a,b that did not get anchored to the POM
scaffold. In both the free diolamides and the organo-POMs,
the proton resonances of the amide in the pyridine derivatives
are from 1 to 1.5 ppm downfield compared to their analogous
1
benzamide derivative. In the H-15N HSQC experiments one
can only observe the 15N resonance of atoms that are directly
bound to a proton. Thus there is no cross peak corresponding
to the pyridine nitrogen of 2a and 2b. The 15N resonances shift
from about 120 ppm in the free ligands to about 160 ppm in
the organo-POMs and they do not vary much in the different
POMs. Simultaneously, the 1H signals are also strongly shifted
downfield (2 ppm average shift) in the organo-POMs, as
mentioned above.
=
Figure 2. TBA5[P2V3W15O59{(OCH2)2C(Et)NHC( O@POM)R}] organo-
POMs considered in this work.
influence the acidity. POM 1d was prepared in 99% yield
using our established method. We also prepared pyridine
derivatives 2a,b,[5c] which have an internal basic site.
The most noticeable feature of POMs 1a–d is the
chemical shift of the proton of the grafted amide in CD3CN.
It ranges from 9.15 ppm for 1c to 9.86 ppm for 1b, with POM
1a at 9.32 ppm and 1d at 9.68 ppm. The chemical shift for 2a
and 2b is 10.40 ppm. This suggests that the iminium resonance
form of the amide might be stabilized by the POM structure.
However, it might also be possible that the amide would be
deprotonated, and the proton released would remain on the
POM as a counterion. To discriminate between these two
options, we decided to monitor the amide resonance with 2D
1H-15N HSQC experiments at natural nitrogen isotopic
abundance (see Supporting Information for details) on
POMs 1a, 2a,b.
The observation of crosspeaks demonstrates that the
target proton remains bound to the amide in the POMs, as
1
does the significant H chemical shift difference between 1a
and 2a,b. The latter certainly reflects the different magnetic
environment of the amide when substituted by a phenyl ring,
relative to an electronically different pyridine ring. This
difference would likely be less pronounced if the protons were
attached to the metal oxide surface, far from the substituents.
1
The strong downfield shift observed in H and 15N supports
our assumption that upon coordination of the amide oxygen
to the vanadate atoms the amide assumes a higher imidate
character (Figure 1, top right) with a partial positive charge
on the nitrogen that is likely responsible for the downfield
shifts.
In order to further validate our hypothesis, we performed
DFT calculations at the BP86/LANL2DZ level of theory
within the Gaussian09 suite of programs[6] for the protonated
and deprotonated hybrid-POM molecules. An implicit sol-
vent correction was added to model the acetonitrile. The
example of 1a is shown in Figure 4. In the protonated form
Figure 3 shows the overlap of the HSQC spectra acquired
on the POMs. Strong cross-peaks were observed for the amide
with a 15N chemical shift of about 160 ppm and H chemical
1
shifts ranging from 9.4 for 1a to 10.4 ppm for 2a and 2b. In
the same spectrum, we also observed weak cross-peaks at
around 120 ppm for nitrogen and 6.9 ppm for 1a, 8.6–8.7 ppm
=
(upper panel, a), the HN-C(Ph) O bond length was calcu-
=
lated at 1.36 Š, and that of the C O bond was 1.28 Š. The
=
phenyl group appears slanted and the V-O( C)-V bridge is
À
À
distorted, with the V O bond facing the C H of the phenyl
À
longer than the V O bond on the other side (2.26 Š vs.
À
2.35 Š). The N C bond connecting the amide group to the
diol part is 1.51 Š long. In the deprotonated form (lower
=
panel, b), the N-C(Ph) O bond length shortens to 1.31 Š,
=
while the C O bond is lengthened (1.37 Š). The phenyl group
=
is much less slanted, and the V-O( C)-V bridge is sym-
À
À
metrical (equal V O bond lengths at 2.13 Š). The N C bond
connecting the amide group to the diol part is unchanged
(1.50 Š). This trend is similar for all the systems calculated,
including the pyridine derivatives (Table S1 and Fig-
ures S1,S2).
1
We then computed the H NMR chemical shifts of the
amide proton of all POMs studied (in their N-protonated
forms), using the GIAO protocol, at the PBE1PBE/
LANL2DZ//BP86/ LANL2DZ level of theory. The calculated
values were in very good linear agreement with the observed
shifts (Figure S3, R2 = 0.952), with a systematic deviation of
0.33 ppm. Finally, we calculated the proton affinities for 1a.
This was carried out to explore the possibility of a protonation
on the oxo ligands of the POM structure. The proton affinity
Figure 3. Superposition of the 1H-15N HSQC spectra of the hybrid-
POM solution in CD3CN (1a: 5 mm, 2a: 2.3 mm, 2b: 7.5 mm;
T=298 K). The spectra were acquired at 14.1 T (600 and 60.8 MHz for
1H and 15N respectively) using 512”128 complex points with 64 scans,
4.0 s of recycle delay, and FID acquisition times of 17.5 ms and
53.2 ms in the indirect and direct dimension, respectively. The 2D FID
was processed with a 1024”512 complex matrix points with square-
cosine windows functions for both the direct and indirect dimensions.
5962 www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 5961 –5965

Angewandte
Communications
Chemie
electron-poor heteroaromatic ring, which pull electrons and/
or of a difference in anisotropy cones compared to arenes.
The natural population analysis conducted on DFToptimized
structures indicates no significant variation of the charges
redistribution between arenes and pyridines (see Table S3),
which tends to support the second argument.
The 1D 1H spectra of compounds 2a and 2b evidenced no
extra peak that could indicate the presence of protonated
pyridinium ion (i.e. an internal prototropy from the amide to
the pyridine ring). No extra peak attributable to protonated
pyridinium could be found in the HSQC spectra either, and
this was consistent with the chemical shift calculations.
However, to exclude the possibility that the pyridinium
proton would not be observed because of a rapid exchange
with the residual water in the solvent, we acquired 2D 1H-15N
1
HMBC spectra (Figure S4). The 2D H-15N HMBC experi-
ment correlates the nitrogen spin with the non-exchangeable
protons attached to the carbon which is bound to nitrogen via
the weak long-range 1H-15N 2J or 3J-scalar couplings. It is thus
possible to determine the 15N chemical shift of the pyridine
nitrogen independently of its protonation state and to clearly
assess if it is protonated or not. This experiment provides in
just a few hours of acquisition an unambiguous confirmation
of the pyridine protonation state but also a confirmation of
the correct assignment of the amide nitrogen. We observed
a correlation among the aromatic proton of the pyridines and
the nitrogen resonance at 303.8 ppm for 2a and 306.6 ppm for
2b. This chemical shift value is very close to the chemical shift
of the free pyridine (302 ppm in CHCl3) and it is far from the
chemical shift observed for the protonated pyridinium ions
(204 ppm in CHCl3 for the pyridinium chlorohydrate).[7] This
unambiguously shows that the pyridine is not protonated in
organo-POMs 2a and 2b.
The absence of prototropy in 2a,b suggests that the
protons of the inserted amide are only weakly acidic. To
evaluate the potential of the organohybrids as Brønsted
organocatalysts we thus targeted a reaction where the key
event does not involve protonation of the weakly basic
carbonyl group. The organo-catalyzed reduction of quinolines
appeared as a perfect benchmark.[8] The mechanism of that
reaction involves the formation of an ion pair between the
deprotonated form and the protonated quinoline substrate.
The latter is then hydrogenated by the Hantzsch ester. We felt
that the highly negative charge of the POM would ensure the
formation of a tight ion pair, potentially giving it a stronger
influence over the course of the reaction.
Figure 4. Calculated structure of POM 1a, fully optimized at the DFT
level for the a) N-protonated and b) conjugated base forms.
at the nitrogen position is more favorable by at least
25.3 kcalmolÀ1 compared to a terminal oxygen and by at
least 19.5 kcalmolÀ1 compared to a bridging oxo (Table S2).
This means that according to calculations, the POM surface is
not basic enough to deprotonate the inserted amide. Similar
conclusions can be drawn from all the other systems (see
Table S2).
All these models confirm the positioning of the proton on
the nitrogen, not on the POM structure. The calculated bond
lengths are fully consistent with our hypothesis. In the
À
protonated form, the C N bond has less double-bond
=
character, and is longer. The C O bond is also shorter
À
because of the conjugation to the C N. Steric strain appears
between the POM and the phenyl group, which results in
a distortion. The organic ligand “pulls” the oxygen to escape
À
this tension, resulting in longer V O bonds, and a loss of
À
symmetry. Upon deprotonation, the C N bond is consider-
À
ably shortened, while the C O bond is lengthened. This
reflects the increase of the imidate form, as expected since
more electron density is accommodated on the oxygen, and
In a first experiment, we heated a suspension of POM 1a
(1 mol%), 2-phenyl quinoline 3a, Hantzsch ester (2.5 equiv)
in toluene for 13 h (Table 1, entry 1). Gratifyingly, 62% of the
expected 4a were isolated. No reaction took place in THF
(entry 2), but the reaction was quantitative in acetonitrile
(entry 3). We selected this solvent for the rest of the study. At
room temperature the reaction was slower, taking 3 days to
yield 68% of 4a (entry 4). With 0.5 mol% of catalyst, the
yield was 94% (entry 5). With 0.1 mol% of catalyst, the
conversion suffered and only 30% of 4a was isolated
(entry 6).
À
the POM. The C O bond lengthening also releases the strain,
with a V-O-V crown back to normal.
The electronic nature of the arene substituent influences
the NMR chemical shifts. The amide signals in 1b and 1d are
much further downfield, as one would expect since electron-
withdrawing groups are likely to pull electron density from
À
the iminium, thus making the N H more polarized, while
a donating group would do the contrary (as in 1c). For
pyridine-substituted POMs 2a,b, the signals are even further
downfield. This can be a consequence of the presence of the
We next carried out control experiments. No reaction took
place in the absence of POM (entry 7), or in the presence of
Angew. Chem. Int. Ed. 2016, 55, 5961 –5965
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5963

Angewandte
Communications
Chemie
Table 1: Optimization of the conditions for the organo-POM-catalyzed
hydrogenation of quinolines.
Table 2: Scope of the hydrogenation.
Entry[a] Cat. (mol %)
Solvent
T
Yield
[%]
Entry
Substrate
POM
Product, yield [%]
1
2
3
4
5
6
7
8
1a (1)
1a (1)
1a (1)
1a (1)
1a (0.5)
1a (0.1)
–
toluene
THF
D
D
D
62
–
99
[b]
1
2
3
4
5
6
7
3a
3a
3a
3a
3b
3c
3d
1a
1b
1d
1c
1a
1a
1a
4a, 94
4a, 91
4a, 96
4a, 81
4b, 95
4c, 100
4d, 100
MeCN
MeCN RT 68[c]
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
D
D
D
D
D
D
D
D
94
30
[b]
–
–
–
[b]
TBABr
[b]
9
diolamide (R=Ph, 0.5)
TBA5H4[P2W15V3O62] (0.5)
TBA6.5H2.5[P2W15V3O62] (0.5)
TBA5H4[P2W15V3O62] (0.5) + diolamide
(R=Ph, 0.5)
10
11
12
62
34
59
optimized reaction (Table 2, entry 1), both POMs with
electron-withdrawing substituents 1b and 1d led to similarly
high yields of 4a (entries 2,3). Hybrid 1c which has a donating
para-methoxy substitutent was slightly less efficient (81% vs.
94 with 1a, entry 4). On the substrate side, all substrates
considered led to an efficient reaction (entries 5–7). This
rapid scope study shows that the POM-based catalyst
tolerates changes in the periphery of the inserted amide,
providing an opportunity for tuning.
To summarize, we have introduced a new family of acidic
POMs, in which the acidity of an amide proton is exalted upon
grafting to the vanadium-oxide crown of a Dawson vanado-
tungstate. Cooperative interactions between the inorganic
and organic parts of the organo-POM lead to the first
covalent hybrid heteropoly acids.[5b–c,9] Beyond POM chemis-
try, this is a new way of generating acidic organocatalysts,
a subclass of catalysts with a rather narrow functional
diversity. The organic end could be further decorated to
enhance the activity or build a chiral pocket around the
proton.
[a] Unless otherwise noted: 2-phenylquinoline (1 equiv), Hantzsch ester
(2.5 equiv), reaction left for 13 h. [b] No reaction. [c] Reaction was left for
3 days.
tetrabutylammonium bromide (entry 8). The “free” diol-
amide alone did not catalyze the reaction either (entry 9). On
the other hand, the purely inorganic [P2W15V3O62]9À sur-
rounded by five TBA+ cations and four protons delivered
62% of 4a (entry 10). After cation exchange on a TBA-
loaded resin to strip it from some of its surrounding protons,
the activity of the same POM (with 6.5/2.5 TBA+/H+
counterions) dropped significantly (34% completion after
13 h, entry 11). Finally, the physical mixture of
TBA5H4[P2W15V3O62] and the diolamide (thus with no
covalent attachment) returned a yield comparable to that of
TBA5H4[P2W15V3O62] alone (59%, entry 12).
The controls show that the POM is necessary for the
reaction to proceed. As one would expect from the literature,
the diolamide in itself is not capable of acting as a protic
catalyst, nor is the POM counterion. The experiments of
entries 10–11 show that the purely inorganic structure is
a possible catalyst because of its surrounding protons.
However, with four protons it is already less active than 1a,
and its activity drops further with the proton concentration.
As the full deprotonation of 1a would provide a species with
only one proton, we can surmise that the latter would be even
less active. Finally, there is no cooperativity of the inorganic
and organic partners if they are not grafted (entry 12). As the
NMR shows that the proton of the amide is on the nitrogen,
and not on the POM surface we deduce that the activity stems
from the activation of the amide in the grafted species. The
inorganic structure does not act as a support for a proton, like
the usual heteropolyacids, but it exalts the acidity of the
organic part.
Acknowledgements
This work was supported by CNRS, the RØgion Rhône-Alpes
(doctoral stipend to D.V.), CPE Lyon, UniversitØ Claude
Bernard Lyon 1, ENS Lyon and TGIR RMN-THC FR 3050
(CRMN-Lyon). We thank the Pôle Scientifique de ModØlisa-
tion NumØrique at ENS-Lyon for modeling time.
Keywords: acid derivatives · hydrogen transfer ·
organic-inorganic hybrid composites · organocatalysis ·
polyoxometalates
How to cite: Angew. Chem. Int. Ed. 2016, 55, 5961–5965
Angew. Chem. 2016, 128, 6065–6069
We next shortly examined the influence of changes in the
POM structure and in the substrate. Given the molecular
weight of the POM, we selected the conditions of entry 5 for
a short scope study (Table 2). Compared to the previously
[1] a) S.-S. Wang, G.-Y. Yang, Chem. Rev. 2015, 115, 4893; b) V. F.
Odyakov, E. G. Zhizhina, Y. A. Rodikova, L. L. Gogin, Eur. J.
Inorg. Chem. 2015, 3618; c) I. Kozhevnikov in Handbook of
5964 www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 5961 –5965

Angewandte
Communications
Chemie
Green Chemistry, Vol. 2 (Eds.: P. T. Anastas, R. H. Crabtree),
derivative. See: f) Y. Hou, C. L. Hill, J. Am. Chem. Soc. 1993, 115,
Wiley-VCH, Weinheim, 2009, pp. 153 – 174.
11823; g) C. P. Pradeep, D.-L. Long, G. N. Newton, Y. F. Song, L.
Cronin, Angew. Chem. Int. Ed. 2008, 47, 4388; Angew. Chem.
2008, 120, 4460; h) Y. Han, Y. Xiao, Z. Zhang, B. Liu, P. Zheng, S.
He, W. Wang, Macromolecules 2009, 42, 6543.
[2] a) D. Hueber, M. Hoffmann, B. Louis, P. Pale, A. Blanc, Chem.
Eur. J. 2014, 20, 3903; b) D. Hueber, M. Hoffmann, P. de Fremont,
P. Pale, A. Blanc, Organometallics 2015, 34, 5065; c) N. DuprØ, P.
Remy, K. Micoine, C. Boglio, S. Thorimbert, E. Lacôte, B.
Hasenknopf, M. Malacria, Chem. Eur. J. 2010, 16, 7256; d) Q.
Han, C. He, M. Zhao, B. Qi, J. Niu, C. Duan, J. Am. Chem. Soc.
2013, 135, 10186; e) Y. Kikukawa, K. Suzuki, M. Sugawa, T.
Hirano, K. Kamata, K. Yamaguchi, N. Mizuno, Angew. Chem. Int.
Ed. 2012, 51, 3686; Angew. Chem. 2012, 124, 3746; f) K. J. Stowers,
K. C. Fortner, M. S. Sanford, J. Am. Chem. Soc. 2011, 133, 6541.
[3] a) M.-P. Santoni, G. S. Hanan, B. Hasenknopf, Coord. Chem. Rev.
2014, 281, 64; b) H. Li, L. Wu, Soft Matter 2014, 10, 9038; c) A.
Dolbecq, E. Dumas, C. R. Mayer, P. Mialane, Chem. Rev. 2010,
110, 6009; d) P. Yin, D. Li, T. Liu, Chem. Soc. Rev. 2012, 41, 7368;
e) A. Proust, R. Thouvenot, P. Gouzerh, Chem. Commun. 2008,
1837; f) D.-L. Long, R. Tsunashima, L. Cronin, Angew. Chem. Int.
Ed. 2010, 49, 1736; Angew. Chem. 2010, 122, 1780.
[6] Gaussian09, RevisionD.01, M. J. Frisch, et al., Gaussian, Inc.,
Wallingford CT, 2009.
[7] R. O. Duthaler, J. D. Roberts, J. Am. Chem. Soc. 1978, 100, 4969.
[8] a) M. Rueping, A. P. Antonchick, T. Theissmann, Angew. Chem.
Int. Ed. 2006, 45, 3683; Angew. Chem. 2006, 118, 3765; b) J. W.
Yang, M. T. Hechavarria Fonseca, B. List, Angew. Chem. Int. Ed.
2004, 43, 6660; Angew. Chem. 2004, 116, 6829; c) C. Zhu, T.
Akiyama, Org. Lett. 2009, 11, 4180.
[9] For other examples of organo-POM catalysts, see: a) C. Brazel, N.
DuprØ, M. Malacria, B. Hasenknopf, E. Lacôte, S. Thorimbert,
Chem. Eur. J. 2014, 20, 16074; b) G. Modugno, A. Monney, M.
Bonchio, M. Albrecht, M. Carraro, Eur. J. Inorg. Chem. 2014,
2356; c) W.-K. Miao, Y.-K. Yan, X.-L. Wang, Y. Xiao, L.-J. Ren, P.
Zheng, C.-H. Wang, L.-X. Ren, W. Wang, ACS Macro Lett. 2014,
3, 211; d) M. Carraro, G. Fiorani, L. Mognon, F. Caneva, M.
Gardan, C. Maccato, M. Bonchio, Chem. Eur. J. 2012, 18, 13195;
e) N. DuprØ, C. Brazel, L. Fensterbank, M. Malacria, S. Thor-
imbert, B. Hasenknopf, E. Lacôte, Chem. Eur. J. 2012, 18, 12962;
f) S. Berardi, M. Carraro, M. Iglesias, A. Sartorel, G. Scorrano, M.
Albrecht, M. Bonchio, Chem. Eur. J. 2010, 16, 10662; g) I. Bar-
Nahum, R. Neumann, Chem. Commun. 2003, 2690; h) H. Zeng,
G. R. Newkome, C. L. Hill, Angew. Chem. Int. Ed. 2000, 39, 1771;
Angew. Chem. 2000, 112, 1841.
[4] a) S. Luo, J. Li, H. Xu, L. Zhang, J.-P. Cheng, Org. Lett. 2007, 9,
3675; b) J. Li, X. Li, P. Zhou, L. Zhang, S. Luo, J.-P. Cheng, Eur. J.
Org. Chem. 2009, 4486.
[5] a) J. Li, I. Huth, L.-M. Chamoreau, B. Hasenknopf, E. Lacôte, S.
Thorimbert, M. Malacria, Angew. Chem. Int. Ed. 2009, 48, 2035;
Angew. Chem. 2009, 121, 2069; b) B. Riflade, A. Nol, M.
Malacria, S. Thorimbert, B. Hasenknopf, E. Lacôte, Org. Lett.
2011, 13, 5990; c) B. Riflade, D. Lachkar, J. Oble, J. Li, S.
Thorimbert, B. Hasenknopf, E. Lacôte, Org. Lett. 2014, 16, 3860;
d) I. Azcarate, I. Ahmed, R. Fara, M. Goldmann, X. Wang, H. Xu,
B. Hasenknopf, E. Lacôte, L. Ruhlmann, Dalton Trans. 2013, 42,
12688; e) I. Azcarate, Z. Huo, R. Farha, M. Goldmann, H. Xu, B.
Hasenknopf, E. Lacôte, L. Ruhlmann, Chem. Eur. J. 2015, 21,
8271. The polyoxoanion can also be converted into its triester
Received: November 25, 2015
Revised: February 26, 2016
Published online: April 8, 2016
Angew. Chem. Int. Ed. 2016, 55, 5961 –5965
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5965