Tris(triazolyl) calix[6]arene-based zinc and copper funnel complexes: Imidazole-like or pyridine-like? A comparative study

Article abstract of DOI:10.1021/ic201540x

Huisgen dipolar cycloaddition leads straightforwardly to new funnel complexes based on the calix[6]arene macrocycle bearing three functionalized triazoles as coordinating units at the small rim. Coordination to Zn(II) and Cu(I) cations was studied using <

Full text of DOI:10.1021/ic201540x

ARTICLE
pubs.acs.org/IC
Tris(triazolyl) Calix[6]arene-Based Zinc and Copper Funnel Complexes:
Imidazole-like or Pyridine-like? A Comparative Study
Benoit Colasson,† Nicolas Le Poul,‡ Yves Le Mest,‡ and Olivia Reinaud*,†
†Laboratoire de Chimie et Biochimie Toxicologiques et Pharmacologiques, CNRS UMR 8601, PRES Sorbonne Paris Citꢀe,
Universitꢀe Paris Descartes, 45 Rue des Saints Pꢁeres, 75006 Paris, France
‡Laboratoire de Chimie, Electrochimie Molꢀeculaires et Chimie Analytique, CNRS UMR 6521,
Universitꢀe Europꢀeenne de Bretagne ꢁa Brest, CS 93837, 6 Avenue Le Gorgeu, 29238 Brest Cedex 3, France
S Supporting Information
b
ABSTRACT: Huisgen dipolar cycloaddition leads straightforwardly to new
funnel complexes based on the calix[6]arene macrocycle bearing three
functionalized triazoles as coordinating units at the small rim. Coordination
1
to Zn(II) and Cu(I) cations was studied using H NMR and IR spectro-
scopies and cyclic voltammetry. The nature of the substituents on the triazole
ring affects the behavior of the ligands and their coordinating ability and
controls the hostÀguest properties of the metal receptors for exogenous
substrates. Depending on their substitution pattern but also on the metal ion
and the guest ligand, the triazole-based systems behave either imidazole-like
or pyridine-like. The ease of preparation and the versatility of 1,4-disubsti-
tuted-1,2,3-triazoles with tunable steric and electronic properties make them
promising candidates for further applications from biology to materials.
’ INTRODUCTION
donor ligands. Indeed, with a tripodal system, the effect of each
nitrogenous donor is enhanced compared to a monodentate
ligand, which may give more information. More importantly, it
allows the control of the geometry of the complexes without
saturating the coordination sphere. The free sites left at the metal
center will then allow exploration of the impact of the nitrogen
donor on the interaction of the metal ion with exogenous
molecules. The systems must present a maximum of similarities
in their structures for the comparison to be valuable. Recently, we
described the coordination properties of ligands that simulta-
neously present two tridentate cores, one tris(imidazole) and
one tris(triazole). They are based on a calix[6]arene functional-
ized at the small rim with three imidazoles and at the large rim by
three triazole moieties introduced through the CuAAC pro-
cedure.6 The related ditopic ligands were shown to stabilize
homo- and heterobimetallic complexes with remarkable control
over the stoichiometry and the regioselectivity of the complexation.7
Although this highlighted important differences in the coordina-
tion properties of the tridentate cores, the direct comparison
suffers from differences in the geometry of the tripodes due to the
nonsymmetrical cone shape structure of the calixarene. Willing
to address this question with more accuracy, we designed novel
calixarene-based ligands that present a tris(triazolyl) core at the
small rim very much like the previously reported tris(imidazolyl)
system (Imme3, see Scheme 1).8,9 Of high interest, these macro-
cyclic ligands provide a supramolecular environment that mimics
The use of the Huisgen dipolar cycloaddition has bloomed
since the discovery of the copper-catalyzed version ca. 10 years
ago.1 Indeed, the Copper-Catalyzed AlkyneÀAzide Cycloaddi-
tion (CuAAC) is a powerful tool for connecting two moieties in
high yield and under mild conditions. In the biological context,
more particularly, it is extensively used to prepare sophisticated
molecules, inhibitors, and drugs.2 Interestingly, the triazole group
is not always an innocent linker. In particular, it has been recog-
nized that, through its rigid structure, 1À4 substitution pattern,
and polarity, it can mimic a peptidic bond. More recently, its
coordination chemistry has started to be explored. The growing
number of studies related to the use of triazole as a ligand emphasizes
its ability to interact with various transition metal ions and to
produce metal complexes.3 Indeed, if they structurally mimic the
rigidity and polarity of the peptide bond, from a coordination
point of view, they resemble much more the imidazole com-
plexes. This raises the question of which properties of the triazole
group toward transition metal ions are biologically relevant.
Indeed, histidine residues are involved in the coordination of
transition metal ions either to structure proteins or to provide a
catalytic metal site in enzymes.4 It is also recognized that various
proteinscorrelatedtoneurodegenerative diseases bind copper through
their imidazole-rich backbone.5 In that context, it appeared to
us quite important to compare the properties of this group with
those of imidazole.
In order to evaluate as exhaustively as possible the coordina-
tion properties of the triazole moiety, we chose to compare tridentate
cores, i.e., tris(imidazole) versus tris(pyridine) and tris(triazole)
Received: July 19, 2011
Published: September 29, 2011
r
2011 American Chemical Society
10985
dx.doi.org/10.1021/ic201540x Inorg. Chem. 2011, 50, 10985–10993
|

Inorganic Chemistry
ARTICLE
Scheme 1. (Bottom) Synthesis of the Triazolyl Ligandsa,b and
(Top) Representation of the Imme3 and Pic3 Related Ligands
Figure 1. 1H NMR (CDCl3, 500 MHz, 300 K) spectra of ligands 3, 4,
and 5 (from bottom to top). (9) tBu, (0) OCH3, (1) ArCH2, (O)
triaCH2, ()) HAr, (b) Htria (/ minor conformations).
a Propargylbromide, Cs2CO3, DMF, 150°C, 93%. b RN3, CuSO4.5H2O,
Na ascorbate, CH2Cl2/H2O, RT, ca. 75%.
This reaction was performed in DMF at 150 °C with an excess of
propargylbromide and cesium carbonate. Tris(propargyl)-calix-
arene 2 was obtained in 93% yield. The versatility of the CuAAC
procedure enabled us to synthesize three different triazolyl-
calixarene ligands. The reaction was carried out in the biphasic
solvent mixture CH2Cl2/H2O at room temperature with copper
sulfate reduced in situ by sodium ascorbate as a catalyst. Com-
pounds 3, 4, and 5 were isolated in good yields (ca. 75%).
In each case, ESI-MS experiments indicated the clean reaction
of all three acetylene groups with two peaks at m/z correspond-
ing to the monoprotonated ligand and the sodium adduct (see
the Supporting Information). No peak associated with products
bearing only one or two triazoles was observed. 1H NMR spectra
of the three ligands in CDCl3 are very similar and in good
agreement with the C3v symmetry of the molecules in a cone
conformation (Figure 1). Two resonances at ca. 0.7 and 1.4 ppm
associated with peaks at 6.6 and 7.2 ppm account for the tBu-
phenoxyl units of the calixarene. The resonance for the methoxy
groups is a singlet at 2.1À2.2 ppm, which indicates that all three
methoxy groups point inside the cavity. These observations
denote a flattened cone conformation with three aromatic units
being oriented toward the inside of the cone of the calixarene, the
three other ones being projected away from it. Finally, the
resonances of the protons for the methylene bridges are split
into two doublets at ca. 3.3 ppm and 4.5 ppm, in agreement with
a slow coneÀcone inversion. The small resonances that can be
detected especially between the two tBu peaks attest to the
presence of minor conformers, as observed for other tridentate
ligands of the same family.8c,12a Indeed, variable temperature 1H
NMR experiments showed that these small peaks tend to vanish
upon heating the solution. Interestingly, the chemical shift of the
protons located on the triazole units, Htria, appears to be an excel-
lent reporter of the electron density on the triazolyl rings: the
more electron-withdrawing the substituent (pNO2Ph > pCH3OPh >
pCH3PhCH2), the higher field the resonance (8.4 > 8.2 > 7.8
ppm). Surprisingly, even some protons belonging to the calixar-
ene macrocycle are affected by the nature of the R substituents,
albeit located far away from the triazoles (HAr, ArCH2, tBu, and
OCH3). This shows that the substitution pattern has a remote
effect on the conformation of the calixarene cores.
the frame of a protein backbone. It also maintains a free coor-
dination site in a well-controlled environment accessible via the
cavity.10,11 Indeed, we have extensively studied the metal com-
plexes based on the Imme3 tripodal ligand and found that their
biomimetic properties do not only depend on the first coordina-
tion sphere. The second coordination sphere provided by the
oxygen-rich small rim of the calixarene also plays a decisive role
in the stabilization of a guest donor as well as the hydrophobic
cavity acting as a funnel for the selective access of an exogenous
ligand. Herein, we report the synthesis of three different triden-
tate tris(triazolyl) ligands and describe their coordination and hostÀ
guest properties with a hard dicationic metal ion, Zn(II), and with a
soft monocationic one, Cu(I), together with some electrochemical
features. A comparison with the structurally analogous tris(imidazolyl)-
and tris(pyridyl)-calix[6]arene derivatives12 (Pic3, see Scheme 1)
allows the highlighting of a number of interesting properties.
’ RESULTS
Synthesis of the Ligands. In order to obtain valuable data on
the coordination behavior of tris(triazolyl) ligands based on
calix[6]arenes, a key feature is the number of atoms separating
the coordinating N-donor site and the phenoxyl units to which
they are attached. The best situation is when two atoms separate
the N donors from the OAr units. Indeed, we have previously
observed that a longer spacer brings too much flexibility and
jeopardizes the coordination ability of the resulting ligand.13 The
introduction of triazoles at the small rim of calix[6]arenes via the
CuAAC procedure can be envisioned in two ways: an azido-
calixarene reacting with alkynes or an alkyne-substituted calixar-
ene reacting with organic azides. A calixarene building block for
the first approach was recently reported.14 However, preliminary
experiments done in collaboration with Jabin et al. for the
coordination of Zn(II) cation were unsuccessful,13 suggesting
that the triazolyl-calixarene derivatives presenting a three-atom
spacer instead of a two-atom spacer are not optimal for the
formation of the complex. In order to geometrically stick as much
as possible to the parent Imme3 and Pic3 ligands extensively
studied in our laboratory, we decided to use the alternative
strategy. We first introduced the alkyne groups on trimethox-
ycalixarene 1 via alkylation of the phenol rings (Scheme 1).15
We then evaluated the impact that the difference in triazole
electron density would have on the coordination properties of
the ligands as well as on the hostÀguest properties of the corre-
sponding Zn(II) and Cu(I) metal complexes.
10986
dx.doi.org/10.1021/ic201540x |Inorg. Chem. 2011, 50, 10985–10993

Inorganic Chemistry
ARTICLE
Scheme 2. Synthesis of the Zinc Complexes in THF and Schematic Representation of the HostÀGuest Complexesa
a R = pNO2Ph, pCH3OPh, or pCH3PhCH2; S = residual water or coordinating solvent; G = guest ligand = H2O or MeCN or PrNH2 (see text).
Zn(II) Complexes and HostÀGuest Studies. The complexes
were prepared by adding 1 equiv of Zn(H2O)6(ClO4)2 into a
solution of ligand 3, 4, or 5 in THF (Scheme 2) and isolated by
precipitation with pentane. In the case of ligand 3, the 1H NMR
spectrum in CDCl3 at 300 K clearly indicates the formation of a
well-defined Zn(II) complex (Figure 2). Although slightly broader
compared to the free ligand, the resonances attest to a C3v
symmetrical compound. The important downfield shift (+0.8 ppm)
observed for the triazolyl protons agrees with the coordination of
Zn(II) to all three triazole groups. Other changes in chemical
shifts related to the calixarene core are quite similar to those
observed for the tris(imidazole) parent complex.8 For instance,
the resonance for the methoxy protons belonging to the calixar-
ene core is strongly shifted downfield (+ 1.2 ppm). This indicates
Figure 2. 1H NMR (CDCl3, 500 MHz, 300 K) spectra of ligand 3
a conformational switch during the complexation event. Indeed,
(bottom) and complex [Zn 3 (H2O)n](ClO4)2 (top). (9) tBu, (0)
the aromatic units of the calixarene must flip one relative to the
other in order to allow all three triazoles to coordinate the zinc
cation, thus moving the methoxy groups away from the cavity, as
illustrated in Scheme 2. Finally, the peak of the residual water is
much broader than in the case of the spectrum of the free ligand
and shifted downfield by ca. 0.6 ppm. This shows that water
interacts with the metal center and that the coordinated water
molecule is in fast exchange with free water. In anhydrous CDCl3,
some resonances are sharper (Htria for instance), proving the role
of water in structuring the metal complex (see Figure S1 in
Supporting Information). Finally, the dicationic nature of the
Zn(II) complexes is confirmed by elemental analyses. All of this
information accounts for the formation of a hostÀguest complex,
as depicted in Scheme 2 with G = H2O. At this point, we cannot
discriminate between a four- or a five-coordinate species where
residual water (S = H2O) is also in interaction in the exo position
with the metal center as has been previously observed with
Imme3 coordinated to other metal dications such as Cu(II),9b
Ni(II), and Co(II)16 (vide infra in the Discussion).
3
3
OCH3, (1) ArCH2, (O) triaCH2, ()) ArH, (b) Htria (w: residual
water).
a number of new peaks denoting other conformations or
species appeared (which stands in contrast with complex
[Zn 3 (H2O)n](ClO4)2, see Figures S2 and S4, Supporting
3
3
Information). This shows that, unlike in the case of ligand 3,
the C3v complex of 4 is not the only species present in
solution. In the case of ligand 5, the presence of the Zn(II)
did not even change the spectrum. Obviously, the ability
of the calixarene-based ligands to complex the zinc cation
depends on the coordination properties of the nitrogen arms.
It has already been reported that the substitution of 1,4-
disubstituted-1,2,3-triazoles can finely tune their coordi-
nation ability.3a,o More specifically, it was shown that an
aromatic substituent decreases their coordination strength
because of electron delocalization and steric hindrance.
Hence, ligands 3, 4, and 5 seem to follow this trend. Ligand
3, in which the triazoles are substituted by a benzylic group, and
ligand 4, for which the triazoles are electronically enriched by the
anisyl groups, do bind Zn2+. On the contrary, in ligand 5, the
triazoles substituted with the electron-deficient nitrophenyl
groups are too poor σ donors to coordinate the Zn(II) cation
in the absence of guest ligands other than residual water.
For ligand 4, the 1H NMR spectrum of the product resulting
from the addition of Zn(II) is clearly different from the spectrum
of the ligand itself as well (Figure S3, Supporting Information).
The new resonances of Htria at δ = 9.03 ppm and of OCH3 at
δ = 3.21 ppm are in good agreement with the formation of the
complex. However, when the spectrum was recorded at 265 K,
10987
dx.doi.org/10.1021/ic201540x |Inorg. Chem. 2011, 50, 10985–10993

Inorganic Chemistry
ARTICLE
Figure 3. 1H NMR (CDCl3, 250 MHz, 300 K) spectra of Zn(II) com-
plexes based on 3 before (bottom, guest ligand G = H2O) and after (top,
G = MeCN) the addition of ca. 15 equiv of CH3CN. (9) tBu, (0) OCH3,
(1) ArCH2, (O) OCH2, ()) ArH, (b) Htria. (/, free CH3CN; w, residual
water; s, residual THF).
Figure 5. 1H NMR (CDCl3, 500 MHz, 260 K) spectra of the Zn(II)
funnel complexes based on ligands 4 (top) and 5 (bottom) in the pres-
ence of 4 equiv of propylamine and 3 equiv of heptylamine (G = PrNH2).
for PrNH2. No interaction could be detected at RT (the NMR
spectrum remained unchanged), and only partial binding of
propylamine was observed at 260 K (see Figure 5, top). Quite
surprisingly, whereas ligand 5 does not even form a Zn(II)
complex in chloroform, full encapsulation of propylamine was
observed at 260 K (see Figure 5, bottom). This shows that the
presence of propylamine in solution induces the formation of the
Zn(II) complex. Ligand 5 is too electron-deficient to stabilize a
Zn(II) dication in the environment provided by the calixarene
core when water or acetonitrile are the only extra-donors to
complete the coordination sphere (vide supra). However, the
presence of a good donor like PrNH2 induces the formation of
the hostÀguest adduct thanks to extra stabilization arising from
the encapsulation of the alkylamine: H-bonding to the methoxy
Figure 4. 1H NMR (CDCl3, 500 MHz) spectra of the Zn(II) funnel
complex based on ligand 3 (G = PrNH2) in the presence of ca. 4 equiv of
propylamine at 260 K (bottom) and 300 K (top), with the inset showing
the low field region.
substituents at the small rim and CH π interactions between
3 3 3
the alkyl chain and the cavity phenyl rings.8,10 Such an observa-
tion is interesting because it shows that the more electron
deficient ligand 5 is better than ligand 4 at promoting the
endocoordination of the guest if the latter is a strong σ donor.
We also tested the coordination of an amine bearing a longer
alkyl chain. In both cases (ligands 4 and 5), no coordination
inside the cavity of the calixarene was detected. The addition of
excess heptylamine eventually led to the decoordination of the
Zn(II) cation in the case of 4. These funnel complexes thus nicely
discriminate propylamine versus heptylamine for their coordina-
tion to the metal center via the cavity. This is attributable to the
tBu steric hindrance at the large rim that disfavors the inclusion of
a longer alkyl chain inside the cavity.12b
Coordination of Acetonitrile. Complex [Zn 3 (H2O)n](ClO4)2
3 3
was tested for its ability to coordinate an exogenous acetonitrile
molecule. The addition of ca. 15 equiv of CH3CN into a CDCl3
solution of complex based on 3 did not drastically modify the spec-
trum (Figure 3). Nevertheless, a very broad peak at À0.50 ppm
indicated partial endocoordination of an acetonitrile molecule to the
Zn(II) metal center within the cavity of the calixarene (G = MeCN in
Scheme 2). This interaction is also attested by the upfield shift of the
water resonance, showing a competition between acetonitrile and
water for the coordination to the metal cation, as previously reported
for the Imme3 system.8
Coordination of Amines. A better exogenous σ-donor ligand
was tested. After the addition of 4 equiv of propylamine into a
Cu(I) Complexes: HostÀGuest Interactions with CO and
CH3CN. We then explored the ability of ligands 3À5 to coordinate
a softer metal ion such as Cu(I). The synthesis of the complexes
consists of the addition of 1 equiv of Cu(CH3CN)4PF6 into a
degassed THF solution containing the ligand (Scheme 3). In each
case, a precipitate appeared, which was isolated by centrifugation
to yield the complex as a tan solid.
CDCl3 solution of [Zn 3 (H2O)n](ClO4)2, some modifications
3 3
of the spectrum could be observed at 300 K (Figure 4). The most
significant ones are an upfield shift of the Htria peak (Δδ = À0.29
ppm), the appearance of two broad resonances in the negative
region of the spectrum together with the broadening of the
resonances associated with free propylamine. At 260 K, all
resonances sharpened, and the presence of one equivalent of
propylamine guest deeply included in the calixarene cavity was
clearly evidenced with two sharp resonances at δ = À1.33 ppm
and δ = À2.01 ppm, integrating for two and three protons versus
calixarene, respectively. Hence, like the Imme3-based Zn(II)
complexes, a propylamino funnel complex (G = PrNH2 in
Scheme 2) is obtained with, however, a faster in/out exchange
process for the amino guest.8a The same experiment carried out
with ligand 4 shows a much lower affinity of the metal complex
The 1H NMR spectra of complexes [Cu 3]PF6, [Cu 4]PF6,
3
3
and [Cu 5]PF6 recorded in CDCl3 are shown in Figure 6. The
3
three complexes display a similar signature, i.e. a spectrum with
very broad resonances, as illustrated by the tBu peaks at ca.
1 ppm. Only the resonances corresponding to the triazolyl sub-
stituent R protons remain sharp. As in the tris(imidazole) case,
such behavior is attributable to a dynamic equilibrium involving
the coordination to the Cu(I) metal center by only two out of
the three nitrogenous arms (Scheme 4).9a As far as ligand 5 is
concerned, while it remains insensitive to the presence of Zn(II)
10988
dx.doi.org/10.1021/ic201540x |Inorg. Chem. 2011, 50, 10985–10993

Inorganic Chemistry
ARTICLE
Scheme 3. Synthesis of the Three Copper(I) Complexes
(R = pNO2Ph, pCH3OPh, or pCH3PhCH2)a
a As isolated, n = 0 with ligands 3 and 4, and n = 1 with 5.
in the absence of a strong donor guest ligand, its interaction with
the softer Cu(I) cation is clearly evidenced by the NMR profile.
In the solid state as well as in solution, all three complexes were
revealed to be insensitive to dioxygen, even after weeks, like the
related tris(imidazolyl) and tris(pyridyl)-Cu(I) complexes. As
compared to those complexes, it appears that the R functiona-
lization of the triazol modulates the interactions with CO or
CH3CN.
Interaction of the Copper Complexes with Carbon Monoxide.
After CO was bubbled into chloroform solutions of the complexes,
the NMR spectra changed dramatically. The phenomenon was
reversible since bubbling argon into the solutions gave back the
initial spectra. The new NMR signatures, presenting sharp and
well-resolved peaks (Figure 6), attested to the formation of well-
defined C3v symmetrical complexes. Hence, CO inhibits the
“triazole dance” around the metal center, and its coordination to
the Cu(I) cation induces the formation of a four-coordinate
species. Indeed, Cu(I) compensates the π-back-bonding to the
CO ligand with the coordination to all three triazoles, thus
adopting a tetrahedral geometry (Scheme 4). For the structurally
similar 4 and 5 ligands, the strongest back-donation is logically
observed for the more electron-rich triazole donors: υCO
=
=
2102 cmÀ1 for [Cu 4 (CO)]PF6 (R = p-MeOPh) and υCO
3 3
Figure 6. 1H NMR (CDCl3, 250 MHz, 300 K) spectra of complexes
[Cu 3]PF6 (top), [Cu 4]PF6 (middle), and [Cu 5]PF6 (bottom)
2108cmÀ1 for [Cu 5 (CO)]PF6 (R = p-NO2Ph). However, with
3 3
ligand 3, the intermediate CO stretch value (υCO = 2106 cmÀ1
)
3
3
3
before and after the addition of CO. (9) tBu, (0) OCH3, (1) ArCH2,
(O) triaCH2, ()) ArH, (b) Htria (w, residual water; s, residual THF).
suggests that subtle differences in the calixarene conformation
(evidenced by the chemical shifts for protons HAr, OCH3,
ArCH2, tBu) lead, as in the Imme3 case, to slightly different
environments for the CO guest.9a
Interaction with Acetonitrile. One interesting observation in
the spectra displayed in Figure 6 is the appearance, upon CO
bubbling, of a peak of free acetonitrile solely in the case of
ca. 8.6 ppm for the Htria (see Figure S7, Supporting Information).
In this case, the addition of an excess of acetonitrile modifies the
ligandÀcopper interaction, thus evidencing the greater affinity of
[Cu 5]PF6 toward nitriles.
3
complex [Cu 5]PF6. This peak integrates for ca. 1 equiv of
Electrochemical Studies. Cyclic voltammetry (CV) studies
show that all three tris(triazolyl)Cu(I) complexes behave simi-
larly in CH2Cl2/NBu4PF6 (see Table 1 and Supporting In-
formation). Under argon, an oxidation peak is observed at Epa
≈ 0.8 V versus Fc with subsequent reduction as a broad peak at
Epc ≈ À0.3 V on the reverse scan. The large difference between
cathodic and anodic peak potentials (ΔEp = Epa À Epc > 1 V)
evidences a high reorganization of the coordination sphere
around the copper associated with the change in redox state.17
Under these solvent conditions, the Cu(I) complexes are stabi-
lized as two or three triazolyl coordinated adducts. Their oxida-
tion leads to the formation of a five-coordinate Cu(II) species
3
CH3CN per calixarene that necessarily comes from the copper
precursor used for the synthesis of the complex. It shows that one
acetonitrile molecule remained coordinated to the Cu(I) cation
during the synthetic procedure only with the poorly donating
ligand 5. When ca. 10 equiv of CH3CN were added in a CDCl3
1
solution of [Cu 3]PF6, no significant modification of the H
3
NMR spectrum was observed at room temperature (Figure S5,
Supporting Information). However, at 260 K, a new peak arising
at À1.13 ppm attests to the coordination of an acetonitrile
molecule inside the cavity (Figure S6, Supporting Information).
Nevertheless, this peak does not integrate for 1 equiv, and in the
positive region of the spectrum, the presence of at least two
species is clearly observed. When the same addition of CH3CN
(peak Epa(a) in Figure 7 for [Cu(I) 3]), with coordination of
3
residual water molecules from the solvent: this is fully substan-
tiated by the characterization of the synthesized Cu(II) complex
based on ligand 3. Indeed, the corresponding perchlorato dicationic
was repeated with [Cu 5]PF6, the profile of the spectrum at 300 K
3
was modified with the appearance for instance of a resonance at
10989
dx.doi.org/10.1021/ic201540x |Inorg. Chem. 2011, 50, 10985–10993

Inorganic Chemistry
ARTICLE
Scheme 4. “Triazole Dance” around the Cu(I) Center and CO/MeCN Coordination
Table 1. Anodic (Epa /V versus Fc) and Cathodic (Epc /V versus Fc) Peak Potential Values from Cyclic Voltammetry at a Vitreous
Carbon Electrode of Cu(I) and Cu(II) Complexes Based on Ligands 3À5 and Pic3 and Imme3 Ligands in CH2Cl2/NBu4PF6 0.1 M
(C = 1 mM) under Argon, v = 0.1 V sÀ1
CH2Cl2
CH2Cl2 + CO
CH2Cl2 + CH3CN (45 equiv)
[Cu(I) 3]PF6
Epa = 0.77
Epc = À0.26
Epa = 0.78
Epc = À0.37
Epa = 0.79
Epc = À0.42
Epa = 1.02
Epc = À0.35
Epa = 0.52
Epc = À0.71
Epc = À0.35
Epa = 0.80
Epa =1.14
Epa = 0.62
Epc = À0.11
Epa = 0.56
Epc = 0.07
Epa = 0.63
Epc = 0.02
Epa = 0.76
Epc = À0.35
Epa = 0.18
Epc = À0.07
Epa = 0.72
Epc = -0.17
3
Epc = À0.25
Epa =1.14
[Cu(I) 4]PF6
3
Epc = À0.28
Epa =1.19
[Cu(I) 5]PF6
3
Epc = À0.42
Epa =1.02
[Cu(I) (Pic3)]PF6
3
Epc = À0.35
Epa =1.03
[Cu(I).(Imme3)]PF6
Epc = À0.71
[Cu(II) 3 (H2O)2](ClO4)2
3
3
complex displays UVÀvis and EPR signatures (see Figures S8
and S9, Supporting Information) thatare characteristic ofa square-
based pyramidal environment due to the coordination of all three
triazole arms and two extra donors (residual water or CH3CN
depending on the conditions, in endo and exo positions, as shown
in the inset of Figure 7 (vide infra)).9b,c Such a coordinative envi-
ronment stabilizes the Cu(II) complexes, which accounts for the
much more negative potential (peak Epc(a), Figure 7). This is
fully confirmed by the redox behavior in CH2Cl2 of the synthe-
solvent; see Supporting Information). It highlights the stabiliza-
tion of the Cu(I) complex through coordination of carbon monoxide,
in agreement with the formation of a tetrahedral adduct in which
the three triazolyl units are coordinated to the cuprous center.
After oxidation (peak Epa(b)), it quickly led back in both cases
to the five-coordinated Cu(II) species by water ligation. This
behavior confirms the imidazole-like character of the triazole
Cu(I) complexes in the presence of CO and contrasts with that
observed for the [Cu(I) (Pic3)]+ complex for which no inter-
3
sized [Cu(II) 3 (H2O)2]2+ complex with a reduction peak at
action with CO was detected in voltammetric studies (see Table 1
and Supporting Information).
3 3
Epc = À0.35 V (Figure 7 B).
For each Cu(I) complex, the addition of carbon monoxide shifted
drastically the oxidation peak (Epa(a) to Epa(b)) toward higher
values (ΔEpa ≈ + 400 mV), without significantly affecting the
potential value of the reduction peak (Epc(a) ≈Epc(b)) (seeFigures7
and S10ÀS13, Supporting Information). Such behavior is very
similar to that obtained with the Cu(I) complex based on theImme3
ligand under the same experimental conditions (i.e., CH2Cl2 as a
The effect of acetonitrile on the electrochemical behavior of
the Cu complexes was investigated as well. For all triazole-based
complexes, the progressive addition of aliquots of CH3CN into a
CH2Cl2 solution of the Cu(I) complexes led to a negative shift
of the oxidation peak (ΔEpa ≈ À 100 to À200 mV), together
with a positive shift of the reduction peak on the back scan. A
particularly important variation of the cathodic peak value was
10990
dx.doi.org/10.1021/ic201540x |Inorg. Chem. 2011, 50, 10985–10993

Inorganic Chemistry
ARTICLE
Figure 7. Cyclic voltammograms (v = 0.1 V sÀ1) at a vitreous carbon
electrode in CH2Cl2 + NBu4PF6 0.1 M of (A) [Cu(I) 3]PF6 (1 mM)
3
(a) under Ar (plain line) and (b) under CO (dashed line) and
(B) [Cu(II) 3 (H2O)2](ClO4)2 (1 mM) (a) before and (b) after the
3
3
addition of 150 equiv of CH3CN. Inset: Schematized five-coordinate
Cu(II) complexes based on tris(triazolyl) ligands.
observed for the [Cu 5]PF6 complex (ΔEpc > 400 mV) com-
3
pared to its [Cu 3]PF6 and [Cu 4]PF6 analogs (see Figures 8A,
3
3
Figure 8. Cyclic voltammograms (v = 0.1 V sÀ1) at a vitreous carbon
B and S14, Supporting Information). All of this denotes an
interaction with CH3CN in fast exchange at the metal center.
The high peak separation (ΔEp = Epa À Epc > 500 mV in the three
cases) suggests that important rearrangements (conformation,
coordination) still operate with the change in redox state, as in
pure CH2Cl2, in agreement with the existence of a two, three, or
four-coordination in the Cu(I) state (Scheme 4) and a five-coor-
dination in the Cu(II) state (Figure 7, inset).
electrode in CH2Cl2 + NBu4PF6 0.1 M of (A) [Cu(I) 5]PF6 (1 mM)
3
(a) before (dashed line) and (b) after (plain line) the addition of 150
equiv of CH3CN [intermediate curves: 15, 45, 75 equiv]; (B) [Cu(I) 3]-
3
PF6 (1 mM) (a) before (dashed line) and (b) after (plain line) the
addition of 300 equiv of CH3CN [intermediate curves: 15, 45, 120, 180,
225 equiv]; (C) [Cu(I) (Imme3)]PF6 (1 mM) (a) before and (b) after
3
the addition of 150 equiv of CH3CN [intermediate curves: 15, 30, 45,
75 equiv]; (D) [Cu(I) (Pic3)]PF6 (1 mM) (a) before and (b) after the
3
Interestingly, such behavior contrasts markedly with the tris-
addition of 60 equiv of CH3CN [intermediate curves: 15, 30, 45 equiv].
(/) Cathodic peak corresponding to the Cu(I) to Cu(0) reaction. (//)
Anodic peak corresponding to the Cu(0) to Cu(I) reaction.
(imidazolyl)-based Cu complex. Indeed, the same experiment
performed with complex [Cu(I) (Imme3)]+ (progressive addi-
3
tion of CH3CN into a CH2Cl2 solution of the complex, see
Figure 8C) led to the appearance of a new reversible system at
v = 0.1 V/s after the addition of 150 equiv of acetonitrile, in
agreement with previous electrochemical work in CH3CN.17b
This reversible system is assigned to the transient formation of a
Cu(I) 5-coordinate nitrilo species, which is the thermodynamic
species in the Cu(II) state (similar to that shown in Figure 7,
inset). With the triazolyl systems, such a reversible system was
never obtained, which can be ascribed to the fact that the Cu(I)
state does not reach the five-coordinate environment. Like for
the Pic3 system (Figure 8D), the much higher oxidation potential
(compared to Imme3) indicates the formation of a four- (or
lower) coordinate Cu(I) species as an intermediate in the
Cu(I)fCu(II) redox process. On the other hand, the presence
ofa singlereduction peak aftertheaddition of acetonitrileindicates
that the five-coordinate complex is by far the most stable species
in the Cu(II) state with the triazolyl ligands, which is confirmed
by the CV of the isolated Cu(II) species (Figure 7B). This stands
in contrast with the redox behavior of [Cu(II)Pic3]2+, for which
both four- and five-coordinate species were evidenced at this scan
rate.16 Such a difference may be attributable not only to electro-
nic factors (poor donors) but also to the more open access in the
exo position for a fifth exogenous ligand with the triazole environment.
substituents, the cone cavity is further stabilized by a second
water molecule that is strongly hydrogen-bonded to the first one.
In strong contrast, with pyridyl arms in place of the imidazoles,
no well-defined Zn(II) complex is formed under the same
experimental conditions. This was attributed to the lower donor
ability of pyridine versus imidazole. With the triazole derivatives,
we have encountered both situations, depending on the nature of
the R substituent. On the one hand, with the benzyl group (i.e., 3),
a stable Zn(II) complex is obtained, which displays spectro-
scopic features almost identical to the imidazole case. In parti-
cular, the calixarene conformation, according to the NMR study,
looks very similar, which substantiates the formation of an aqua
dicationic tetrahedral complex where all three triazole arms
coordinate the metal center with the N3 nitrogen of the triazole
cycle. On the other hand, when the triazolyl substituent R is a
strong electron-withdrawing group (pNO2Ph in 5), no interac-
tion with Zn(II) is observed in a noncoordinating solvent such
as chloroform. The ligand presenting intermediate donor ability
(4, R = pCH3OPh) has intermediate behavior leading to the
formation of a mixture of different species in fast exchange at
room temperature. Obviously, the donor property of the triazole
core is tuned by the R substituent, which has a direct impact on
the ability of the calixarene-based ligand to stabilize a dicationic
Zn complex.
’ DISCUSSION: COMPARISON OF THE TRIS(TRIAZOLE)
CORE VERSUS TRIS(IMIDAZOLE) AND TRIS(PYRIDINE)
Hosting Properties. The Imme3-based Zn(II) complex is a
remarkable receptor for a variety of neutral ligands (provided
they can fit in the cavity that allows binding of a single ligand (G)
in the endoposition due to geometrical constrains at the small
rim). On the NMR time scale, the ligand exchange is a slow
process. It is due to two factors: (i) a strong Lewis acidity of the
Zn(II) Coordination. In a noncoordinating solvent, the tris-
(imidazole)-based Zn(II) complex is a four-coordinate tetrahe-
dral dicationic species with all imidazolyl arms coordinated to the
metal center and one water guest ligand completing the coordi-
nation sphere. When the calix[6]arene large rim presents six tBu
10991
dx.doi.org/10.1021/ic201540x |Inorg. Chem. 2011, 50, 10985–10993

Inorganic Chemistry
ARTICLE
Scheme 5. General Overview of the Coordination and HostÀ
Guest Properties of Calix[6]Arene-Based Tris(Nitrogen)
Donors with Zn(II) (S = residual water in chloroform;
G = guest ligand)
Scheme 6. General Overview of the Coordination and
HostÀGuest Properties of Calix[6]Arene-Based
Tris(Nitrogen) Donors with Cu(I)
dicationic four-coordinate Zn(II) center that precludes a direct
dissociative process (that would lead to an unstable three-
coordinate intermediate); (ii) steric constraints at the level of
the tris(imidazol) core that disfavor binding in the exoposition
and transient formation of a five-coordinate intermediate allow-
ing guest exchange (as depicted in Scheme 5). Complexes based
on the triazole cores revealed lower affinities for organic guest
ligands such as amines or nitriles together with a faster exchange
process. This may well be attributable to the presence of a
nitrogen atom in the α position of the coordinating nitrogen of
triazole, instead of a CÀH in the imidazole case, which opens the
coordination sphere for a fifth ligand. As a result, formation of a
five-coordinate complex is favored, and the inÀout exchange of
the endo ligand is faster. This may also explain why, with ligand 3,
the Zn complex is not as good an endoreceptor as with Imme3
since a five-coordinate environment must decrease the Lewis
acidity of Zn(II). Interestingly also, whereas ligand 5 does not
stabilize a Zn complex with water or CH3CN as a guest ligand, a
better donor such as an amine can compensate for the lack of
electron density at the triazole site and allow the formation of the
corresponding funnel complex with additional stabilizing inter-
actions provided by the calixarene cavity. Quite remarkably, this
makes ligand 5 a better receptor for amines in the presence of
Zn(II) than 4, although the latter is more prone to stabilizing
Zn(II) Td species with other guest ligands (H2O, MeCN).
Cu(I) Coordination. The Pic3-based Cu(I) complex was
actually the first funnel complex to have been characterized by
X-ray diffraction. In the presence of a few equivalents of an
organonitrile, it forms a highly stable four-coordinate tetrahedral
complex with the nitrilo-ligand buried in the calixarene cavity. In
strong contrast, the Imme3 ligand does not give rise to the four-
coordinate hostÀguest adduct in the presence of acetonitrile.
With this more electron-donating ligand, only two imidazoles
coordinate simultaneously the electron-rich metal center. This
gives rise to a dance of Cu(I) between the nitrogenous arms and
no guest binding, unless a good π-acceptor such as CO is present
(as in Scheme 4). In that case, the third imidazolyl arm undergoes
coordination to Cu(I), as the resulting tetrahedral complex is
stabilized through π-back-bonding to CO. Of note, the Pic3-
based Cu(I) complex does not coordinate CO, which is attribu-
table to the lower donor ability of pyridine versus imidazole.
With Cu(I), as depicted in Scheme 6, ligands based on triazole
donors display a coordinating behavior that is close to Imme3 in
the absence of any extra donor since Cu(I) undergoes the dance
at the tris(triazole) core, and the complexes display poor affinity
for nitriles (see NMR results). In addition, all three tris-triazolyl
ligands readily give rise to stable four-coordinated species in
the presence of CO, very much like for Imme3, as shown by
spectroscopic (IR, NMR) and electrochemical means. This is in
strong contrast to Pic3Cu(I), which displays high affinity for
CH3CN and no coordination of CO. Such a difference can be
ascribed to the lower donor effect of the pyridine-based ligand.
Electrochemically, the comparative picture is more complicated.
In CH2Cl2, the triazole-based systems remain totally irreversible.
The oxidation and reduction potentials associated with the three
triazole systems lie in between those of Imme3 and Pic3, which
fits with intermediate donor properties (see Table 1). In the
presence of CH3CN, their redox behaviors, as observed by CV,
resemble more that of Pic3 with a high potential oxidation peak
corresponding rather to a four-coordinate species (or lower)
than to a five-coordinate. The redox process however remains
fully irreversible (by CV). This may be ascribed to the increased
stability of five-coordinate Cu(II) states due to a less sterically
hindered environment around the metal center.
’ CONCLUSION
In this work, we have shown that quite sophisticated ligands
associating a coordination sphere and a cavity can be synthesized
in two steps with good yields. The herein presented study
focused on the coordination and resulting hostÀguest properties
of a tris(triazole) core connected to the hydrophobic cavity of a
calix[6]arene. With a strong Lewis acidic cation such as Zn(II),
the more electron-rich triazole led to the more stable complexes.
10992
dx.doi.org/10.1021/ic201540x |Inorg. Chem. 2011, 50, 10985–10993

Inorganic Chemistry
ARTICLE
With the softer Cu(I) cation, triazole coordination was less
affected by its electron-donating capacities and yielded stable
complexes provided, however, the π-acceptor CO ligand was
present. Comparison with the analogous tris(pyridyl)- and tris-
(imidazolyl)-calixarene ligands shows that the tris(triazolyl)
cores have mixed coordinating behavior: with Zn(II), depending
on the substituent present on the triazole cores, it resembles
either the pyridyl (less electron-rich) systems or the imidazolyl
(more electron-rich) system. With Cu(I), whatever the substitu-
tion pattern is, all triazolyl systems resemble more the imidazolyl
complex (CO coordination) except for the redox behavior in the
presence of CH3CN. The study with Zn(II) also highlighted that
an important difference between these ligands lies in the different
steric hindrance next to the donor atom: the triazole has a
nitrogen atom in place of CH for the other two ligands, which
decreases the steric hindrance around the metal center and opens
more widely a fifth coordination site. Interestingly, this was
shown to have an important impact on both the kinetics and
the thermodynamics of the systems. Finally, for electronic as well
as for steric reasons, the substitution pattern of the triazoles
connected at the small rim of the calixarene also changes the
hostÀguest properties of the funnel complexes. This highlights
an efficient transmission of information between one edge of the
funnel (the coordination core) and the other (the cavity). Hence,
the ease of preparation and the versatility of ligands based on 1,4-
disubstituted-1,2,3-triazoles whose steric and electronic proper-
ties can be tuned at will make them very promising candidates for
further applications in coordination chemistry related to biology,
catalysis, and materials chemistry.
Org. Lett. 2011, DOI: 10.1021/ol102914t. (g) Crowley, J. D.; Gavey, E. L.
Dalton Trans. 2010, 39, 4035–4037. (h) Gower, M. L.; Gavey, E. L. Dalton
Trans. 2010, 39, 2371–2378. (i) Joly, J. P.; Beley, M.; Selmeczi, K.; Wenger,
E. Inorg. Chem. Commun. 2009, 12, 382–384. (j) Maeda, C.; Yamaguchi,
S.; Ikeda, C.; Shinokubo, H.; Osuka, A. Org. Lett. 2008, 10, 549–552. (k)
Mullen, K. M.; Gunter, M. J. J. Org. Chem. 2008, 73, 3336–3350. (l)
Ornelas, C.; Aranzaes, J. R.; Salmon, L.; Astruc, D. Chem.—Eur. J.
2008, 14, 50–64. (m) Fletcher, T.; Bumgarner, B. J.; Engels, N. D.;
Skoglund, D. A. Organometallics 2008, 27, 5430–5433. (n) Meudtner,
R. M.; Ostermeier, M.; Goddard, R.; Limberg, C.; Hecht, S. Chem.—Eur. J.
2007, 13, 9834–9840. (o) Ostermeier, M.; Berlin, M.-A.; Meudtner, R. M.;
Demeshko, S.; Meyer, F.; Limberg, C.; Hecht, S. Chem.—Eur. J. 2010,
13, 10202–10213.
(4) Holm, R. H.; Kennepohl, P.; Solomon, E. I. Chem. Rev. 1996, 96,
2239–2314.
(5) (a) Malmstr€om, B. G.; Leckner, J. Curr. Opin. Chem. Biol. 1998,
2, 286–292. (b) Boal, A. K.; Rosenzweig, A. C. Chem. Rev. 2009, 109,
4760–4779. (c) Kodama, H.; Fujisawa, C. Metallomics 2009, 1, 42–52.
(d) Tisato, F.; Marzano, C.; Porchia, M.; Pellei, M.; Santini, C. Med. Res.
Rev. 2010, 30, 708–749. (e) Donnelly, P. S.; Xiao, Z.; Wedd, A. G. Curr.
Opin. Chem. Biol. 2007, 11, 128–133.
(6) Colasson, B.; Save, M.; Milko, P.; Roithovꢀa, J.; Schr€oder, D.;
Reinaud, O. Org. Lett. 2007, 9, 4987–4990.
(7) Colasson, B.; Le Poul, N.; Le Mest, Y.; Reinaud, O. J. Am. Chem.
Soc. 2010, 132, 4393–4398.
(8) With zinc: (a) Sꢀenꢁeque, O.; Rager, M.-N.; Giorgi, M.; Reinaud,
O. J. Am. Chem. Soc. 2000, 122, 6183–6189. (b) Sꢀenꢁeque, O.; Rager,
M.-N.; Giorgi, M.; Reinaud, O. J. Am. Chem. Soc. 2001, 123, 8442–8443.
(c) Sꢀenꢁeque, O.; Giorgi, M.; Reinaud, O. Chem. Commun. 2001, 984–985.
(d) Sꢀenꢁeque, O.; Rondelez, Y.; Le Clainche, L.; Inisan, C.; Rager, M.-N.;
Giorgi, M.; Reinaud, O. Eur. J. Inorg. Chem. 2001, 2597–2604. (e) Sꢀenꢁeque,
O.; Giorgi, M.; Reinaud, O. Supramol. Chem. 2003, 15, 573–580.
(9) With copper: (a) Rondelez, Y.; Sꢀenꢁeque, O.; Rager, M.-N.;
Duprat, A.; Reinaud, O. Chem.—Eur. J. 2000, 6, 4218–4226. (b) Le
Clainche, L.; Giorgi, M.; Reinaud, O. Inorg. Chem. 2000, 39, 3436–3437.
(c) Le Clainche, L.; Rondelez, Y.; Sꢀenꢁeque, O.; Blanchard, S.; Campion,
M.; Giorgi, M.; Duprat, A. F.; Le Mest, Y.; Reinaud, O. C. R. Acad. Sci.,
Ser. IIc 2000, 3, 811–819.
’ ASSOCIATED CONTENT
S
Supporting Information. General procedures, synthesis,
b
and characterization of the ligands and complexes; spectroscopic
data; and electrochemical experiments. This material is available
free of charge via the Internet at http://pubs.acs.org.
(10) Coquiꢁere, D.; Le Gac, S.; Darbost, U.; Sꢀenꢁeque, O.; Jabin, I.;
Reinaud, O. Org. Biomol. Chem. 2009, 7, 2485–2500.
(11) Rebilly, J.-N.; Reinaud, O. In Supramolecular Chemistry: from
Molecules to Nanomaterials (SMC174); Wiley: New York, 2011, in press.
(12) (a) Blanchard, S.; Le Clainche, L.; Rager, M.-N.; Chansou, B.;
Tuchagues, J. P.; Duprat, A.; Le Mest, Y.; Reinaud, O. Angew. Chem., Int.
Ed. 1998, 37, 2732–2735. (b) Rondelez, Y.; Rager, M.-N.; Duprat, A.;
Reinaud, O. J. Am. Chem. Soc. 2002, 124, 1334–1340.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: olivia.reinaud@parisdescartes.fr.
(13) Unpublished work.
’ ACKNOWLEDGMENT
(14) Mꢀenand, M.; Jabin, I. Org. Lett. 2009, 11, 673–676.
(15) van Duynhoven, J. P. M.; Janssen, R. G.; Verboom, W.; Franken,
S. M.; Casnati, A.; Pochini, A.; Ungaro, R.; de Mendoza, J.; Nieto, P. M.;
Prados, P.; Reinhoudt, D. N. J. Am. Chem. Soc. 1994, 116, 5814–5822.
(16) Sꢀenꢁeque, O.; Campion, M.; Giorgi, M.; Le Mest, Y.; Reinaud,
O. Eur. J. Inorg. Chem. 2004, 43, 1817–1826.
This research was supported by CNRS and Agence National pour
la Recherche [Cavity-zyme(Cu) Project ANR-2010-BLAN-7141].
’ REFERENCES
(17) (a) Le Poul, N.; Campion, M.; Izzet, G.; Douziech, B.; Reinaud,
O.; Le Mest, Y. J. Am. Chem. Soc. 2005, 127, 5280–5281. (b) Le Poul, N.;
Campion, M.; Douziech, B.; Rondelez, Y.; Le Clainche, L.; Reinaud, O.;
Le Mest, Y. J. Am. Chem. Soc. 2007, 129, 8801–8810.
(1) (a) Tornøe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem.
2002, 67, 3057–3062. (b) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.;
Sharpless, K. B. Angew. Chem., Int. Ed. 2002, 41, 2596–2599.
(2) (a) Kolb, H. C.; Sharpless, K. B. Drug Discovery Today 2003,
8, 1128–1136. (b) Tron, G. C.; Pirali, T.; Billington, R. A.; Canonico,
P. L.; Sorba, G.; Genazzani, A. A. Med. Res. Rev. 2008, 28, 278–308.
(3) (a) Suijkerbuijk, B. M. J. M.; Aerts, B. N. H.; Dijkstra, H. P.; Lutz,
M.; Spek, A. L.; van Koten, G.; Klein Gebbink, R. J. M. Dalton Trans.
2007, 1273–1276. (b) Li, Y.; Huffman, J. C.; Flood, A. H. Chem.
Commun. 2007, 2692–2694. (c) Mindt, T. L.; Struthers, H.; Brans, L.;
Anguelov, T.; Scheinsberg, C.; Maes, V.; Tourwꢀe, D.; Schibli, R. J. Am.
Chem. Soc. 2006, 128, 15096–15097. (d) Happ, B.; Friebe, C.; Winter,
A.; Hager, M. D.; Hoogenboom, R.; Schubert, U. S. Chem. Asian J. 2009,
4, 154–163. (e) Fleischel, O.; Wu, N.; Petitjean, A. Chem. Commun.
2010, 46, 8454–8456. (f) Ni, X.-L;. Wang, S.;Zeng, X.;Tao, Z.;Yamato, T.
10993
dx.doi.org/10.1021/ic201540x |Inorg. Chem. 2011, 50, 10985–10993