A 4-tert-butylcalix[4]arene tetrahydroxamate podand based on the 1-oxypiperidine-2-one (1,2-PIPO-) chelate. Self-assembly into a supramolecular ionophore driven by coordination of tetravalent zirconium or hafnium(iv)

Article abstract of DOI:10.1039/c4ra00977k

An octadentate tetrahydroxamic calix[4]arene podand incorporating 1-hydroxypiperidine-2-one (1,2-PIPOH) binding units has been designed as a specific chelator for tetravalent metal cations like Zr⁴⁺ or Hf ⁴⁺. This receptor, which can be considered as the first ever abiotic ligand possessing only cyclic six-membered hydroxamate groups, has been synthesized and characterized in its tetraprotonated form (1H₄). Contrary to expectation, however, this new chelator did not form a 1:1 complex upon reaction with M(acac)₄ (M = Zr and Hf; acac = acetylacetonate), but rather self-assembled into a dimeric species of 2:2 stoichiometry. The latter could be characterized in solution by mass spectrometry and NMR spectroscopy as its monopotassium adduct ([M₂K(1)₂] ⁺), pointing to the ionophoric character of the M₂(1) ₂ complex. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4ra00977k

RSC Advances
View Article Online
View Journal | View Issue
PAPER
A 4-tert-butylcalix[4]arene tetrahydroxamate
podand based on the 1-oxypiperidine-2-one
(1,2-PIPOꢀ) chelate. Self-assembly into a
supramolecular ionophore driven by coordination
of tetravalent zirconium or hafnium(IV)†
Cite this: RSC Adv., 2014, 4, 22743
*
Pawel Jewula, Jean-Claude Chambron, Marie-Jose Penouilh, Yoann Rousselin
´
*
and Michel Meyer
An octadentate tetrahydroxamic calix[4]arene podand incorporating 1-hydroxypiperidine-2-one (1,2-
PIPOH) binding units has been designed as a specific chelator for tetravalent metal cations like Zr4+ or
Hf4+. This receptor, which can be considered as the first ever abiotic ligand possessing only cyclic six-
membered hydroxamate groups, has been synthesized and characterized in its tetraprotonated form
(1H4). Contrary to expectation, however, this new chelator did not form a 1 : 1 complex upon reaction
with M(acac)4 (M ¼ Zr and Hf; acac ¼ acetylacetonate), but rather self-assembled into a dimeric species
of 2 : 2 stoichiometry. The latter could be characterized in solution by mass spectrometry and NMR
spectroscopy as its monopotassium adduct ([M2K(1)2]+), pointing to the ionophoric character of the
M2(1)2 complex.
Received 4th February 2014
Accepted 30th April 2014
DOI: 10.1039/c4ra00977k
www.rsc.org/advances
incorporate desferrioxamine B (DFO) as chelator,2 because of
the high affinity of hydroxamates for zirconium(IV)3 and the
ready availability and non-toxicity of this industrially-produced
Introduction
Zirconium(IV) is a d0 hard tetravalent transition metal cation
and a great variety of its open-sphere complexes have been
developed as Lewis acid polymerization catalysts over the past
decades.1 By contrast, a eld of research where the coordination
chemistry of Zr4+ has been recently involved is molecular
imaging.2 Indeed, the 89Zr isotope has been identied as very
promising for the engineering of antibody-based radiotracers
for positron emission tomography (PET) imaging, because its
half-life of 78.4 h (ꢁ3 days) is commensurable with the bio-
distribution half-life of IgG antibodies (a few days), which is not
the case of 64Cu and 86Y, the latter being also more expensive to
produce. However as it is, non-osteophilic chelation systems are
required that encapsulate the radioactive cation and prevent it
from being uptaken into bone tissues. With few exceptions,
siderophore. In addition, the pendant amine function has been
used to advantage for bioconjugation.2f However, DFO is a
hexadentate chelator that cannot fully satisfy the typical coor-
dination number of Zr4+ (i.e. CN ¼ 8), and the use of oxygen-rich
octadentate ligands (for example catecholates, hydroxypy-
ridonates, or hydroxamates) has been suggested recently.2f In
this respect, the now commercially available 3,4,3-LI(1,2-HOPO)
ligand, based on 1,2-hydroxypyridinone (1,2-HOPO), could
represent a possible alternative to DFO. Indeed, the complex of
1,2-HOPOꢀ with Zr4+ has been synthesized and characterized,4
while [3,4,3-LI(1,2-HOPO)]3: forms extraordinarily stable
complexes with tetravalent Ce4+ (log b110 ¼ 41.5), and Th4+
(log b110 ¼ 40.1),5 and it is currently considered as one of the
most promising in vivo decorporation agent for actinides.6
We have recently undertaken a research program aimed at
exploring the coordination properties of 1-hydroxypiperidine-2-
one (1,2-PIPOH, Fig. 1), a six-membered cyclic hydroxamic acid
89Zr4+-based
radiopharmaceuticals under
development
´
´
Institut de Chimie Moleculaire de l'Universite de Bourgogne (ICMUB), UMR CNRS
6302, avenue A. Savary, BP 47870, 21078 Dijon Cedex, France. E-mail:
9
jean-claude.chambron@u-bourgogne.fr; michel.meyer@u-bourgogne.fr; Fax: +33 3
80 39 61 17; Tel: +33 3 80 39 61 16
† Electronic supplementary information (ESI) available: Tables of crystallographic
data for 1Bn4; copies of the mass and NMR spectra of 1Bn4 and 1H4 (including 2D
correlation maps); copies of the 1H NMR spectra of [Na(1H4)]Cl and [K(1H4)]Cl;
copies of the 1H NMR spectra of Zr2(1)2 and Hf2(1)2; copies of the mass and
NMR spectra of [Hf2K(1)2]Cl (including 2D correlation maps) and [Zr2K(1)2]Cl
(1H NMR spectrum). CCDC 983068. For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/c4ra00977k
Fig. 1 Structural formulae of 1,2-HOPO and 1,2-PIPOH.
RSC Adv., 2014, 4, 22743–22754 | 22743
This journal is © The Royal Society of Chemistry 2014

View Article Online
RSC Advances
Paper
that is found as third, pendent binding unit in a few number of take up an additional alkali-metal cation, thus acting as an
siderophores such as exochelin MN7 and tsukubachelin.8 ionophore. These results will be interpreted and rationalized in
Structural studies revealed as expected the formation of tetra- light of the structural features exhibited by the Zr(1,2-PIPO)4
leptic complexes with tetravalent d- and f-block elements, and Hf(1,2-PIPO)4 complexes.3g
including Zr4+, Ce4+, Hf4+, Th4+, and U4+.3g As the binding sites
of 1,2-PIPOꢀ are incorporated in a cyclic structure similarly to
Results and discussion
Ligand design and synthesis
1,2-HOPOꢀ, the former is also a ligand preorganized for
chelation, therefore its complexes should be more stable than
those originating from linear hydroxamates which prevail in
solution in their E or trans form. In exochelin MN and other
related siderophores, an amide nitrogen atom attached at the R-
stereogenic, C-3 carbon atom of the 1,2-PIPOH moiety connects
it to the main chain of the molecule.9 With the natural design in
mind, we decided to synthesize a tetrachelating podand by
graing four 1,2-PIPOH residues to the calix[4]arene platform
for the efficient sequestration of octacoordinated tetravalent
cations such as Zr4+. We chose calix[4]arene as an alternative to
the open-chain or branched polyamine-type backbones in order
to impart a C4-symmetric propeller-like arrangement of the
chelating units around the metal centre expected to adopt a
square-antiprismatic coordination geometry (Fig. 2). Moreover,
the large rim can, in principle, be easily functionalized by water-
solubilising groups.10
The direct precursors of 1H4 (Scheme 1) are the hydrobromide
salt of (R)-3-amino-1-benzyloxypiperidin-2-one (7) 13 and the
acetyl chloride derivative of p-tert-butylcalix[4]arene (8) 12b,14
(Scheme 2). The former was synthesized in four steps and 19%
overall yield from commercially available Cbz-protected L-orni-
thine 2, according to the reaction sequence shown in Scheme
1.13,15 At rst, this reagent was condensed with benzaldehyde in
basic conditions to afford imine 3,15 which was subsequently
oxidized with m-CPBA to oxaziridine 4. The latter was reacted
with triuoroacetic acid in dichloromethane, which allowed us
to isolate the nitrone 5 in 42% overall yield aer purication by
crystallization. Next, 5 was reacted with hydroxylamine to form
the desired cyclic hydroxamic acid,13 which was not isolated but
immediately protected by reaction with benzylbromide to afford
the synthon 6. The amino group was selectively deprotected by
33% HBr in acetic acid to give the hydrobromide salt (R)-7$HBr
in 90% yield. Subsequently, the benzyl-protected tetrapod 1Bn4
was prepared by coupling the acid chloride derivative 8 with
four equivalents of 7$HBr in the presence of triethylamine
(Scheme 2). The modied calix[4]arene 1Bn4 was obtained in
50% isolated yield by fractional crystallization from methanol.‡
Deprotection of 1Bn4 to 1H4 was accomplished by hydro-
genolysis in methanol at room temperature using 20% Pd(OH)2/
C as catalyst. Smooth conditions were used as 10% Pd/C catal-
yses the cleavage of benzyl carbamates by dihydrogen.16
However, even in these mild conditions, the reaction time (15
min) was a critical parameter. Indeed, as shown by MALDI-TOF
mass spectrometry analysis, longer periods resulted in the
hydrogenolysis of the hydroxamic acid N–O bonds. By applying
the optimal conditions, the tetrahydroxamic acid 1H4 was iso-
lated in 93% yield aer ltration on a short pad of silica gel.
In fact, calixarenes have been widely used as a platform for
constructing a variety of podands including hydroxamate
chelators,11 but tetrahydroxamates based on calix[4]arene are
very scarce.12 For example, Gopalan and coworkers reported in
1997 the design, synthesis, and use for thorium(IV) extraction of
4-tert-butylcalix[4]arenes carrying the simplest linear hydrox-
amate ligands connected from the carbonyl side to the oxygen
atoms of the lower rim through a n-butyl chain.12c Inspired by
this work, we designed the podand 1H4 in which the same
number of atoms (4) separates the phenolic oxygen and the
carbonyl carbon atoms of the 1,2-PIPOH chelating units.
This report describes the synthesis and solution character-
isation of the tetrapod 1H4, together with a detailed spectro-
scopic (NMR and HR-MS) study of the reaction products
obtained with zirconium(IV) and hafnium(IV). We show that
instead of forming the expected 1 : 1 complexes (Fig. 2),
receptor 1 affords binuclear 2 : 2 species which can moreover
X-ray crystal structure analysis of 1Bn4
Slow evaporation of a methanolic solution of the benzyl-pro-
tected receptor afforded X-ray quality crystals of
1Bn4$0.75CH3OH$0.25H2O composition. An ORTEP view17 of
the molecular structure of 1Bn4 is shown in Fig. 3, together with
the atom labelling scheme. The calix[4]arene platform adopts a
pinched-cone conformation in the solid state as found in the
structure of numerous other related compounds that carry
substituents at the lower rim in order to prevent ring ipping.18
The angles that the four aromatic rings (A–D) make with the
‡ We observed that attempts to purify crude 1Bn4 by ash column
chromatography (35–70 mm silica gel) led to a product featuring the same
MALDI-TOF mass spectrum as the one of the 1Bn4 samples isolated by
crystallization and characterized unambiguously by X-ray crystallography.
However, its 1H and 13C NMR spectra differed from the ones corresponding to
1Bn4, pointing to some unexplained isomerization or conformational change.
Fig. 2 The anticipated complex of podand 1H4 with tetravalent metal
cations Zr4+ and Hf4+
.
22744 | RSC Adv., 2014, 4, 22743–22754
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
Fig. 3 ORTEP17 picture of 1Bn4 viewed along the direction of rings A
and C. All C–H hydrogen atoms have been omitted for clarity. Thermal
ellipsoids are drawn at the 50% probability level. Hydrogen bonds are
represented by dashed lines.
determining the absolute conguration of 1Bn4. However,
assuming that the chirality of the amine 7 remains unchanged
in the course of amide bond formation, the chirality of the
modied calix[4]arene should be (R,R,R,R)-1Bn4. Interestingly,
intramolecular hydrogen bonds that maintain the calixarene
platform in its pinched-cone conformation, can be detected
between the amide protons and the carbonyl oxygen atoms
belonging to pairs of adjacent strands (Table 1).
Scheme 1 Synthesis of 7$HBr from Cbz-protected L-ornithine 2.13,15
Mass spectrometric analyses of 1Bn4 and 1H4
Both calix[4]arenes were examined by high resolution ESI mass
spectrometry, and both showed two manifolds. The most
intense massif was the one corresponding to the sodium adduct
[M + Na]+ with the most intense line appearing at m/z ¼
1712.868 for 1Bn4 and 1351.685 for 1H4, while the other was
assigned to the disodium adduct [M + 2 Na]2+ at m/z ¼ 867.929
and 687.337, respectively. In both cases, the experimental
monoisotopic distribution patterns perfectly match the calcu-
lated ones (Fig. S2–S5, ESI†). In contrast, MALDI-TOF mass
spectra evidenced only the signals of [M + Na]+.
Scheme
2 Functionalization of calix[4]arene with the 1,2-PIPOH
chelate. The C-atom numbering scheme used in the NMR data
assignment is also shown.
NMR characterisation of 1Bn4 and 1H4
1
The calix[4]arenes were fully characterized by H and 13C NMR
spectroscopy, and the signals were assigned using 2D 1H–1H
mean plane of the four bridging methylene groups (C1, C12,
C23, and C34) are 36.6(1) (A), 87.0(1) (B), 47.6(9) (C), and 84.8(1)ꢂ
(D), respectively. Therefore, rings A and C are strongly tilted,
being practically normal to each other [interplanar angle of
84.2(1)ꢂ], while rings B and D are approximately parallel to each
other [interplanar angle of 2.7(2)ꢂ]. The electron lone pairs of
the four phenolic oxygen atoms O1, O2, O3, and O4 are all
pointing towards the cavity interior. Each piperidine-2-one ring
was shown to adopt a half-chair conformation by ring puckering
analysis (Table S6, ESI†),19 and to exhibit the same congura-
tion at the asymmetric carbon atom (C47, C61, C75, and C89,
respectively), although X-ray crystallography did not permit
Table 1 Intramolecular hydrogen bond distances (A˚) and angles (ꢂ) for
a
1Bn4
N–H/O
d(H/O)
d(N/O)
<(NH/O)
N1–H1/O8
2.09
1.90
2.07
1.96
2.918(5)
2.747(5)
2.887(5)
2.764(5)
156.4
162.3
153.0
151.3
N3–H3A/O15
N5–H5A/O14
N7–H7/O9
a
˚
d(N–H) ¼ 0.88 A for all bonds.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 22743–22754 | 22745

View Article Online
RSC Advances
Paper
COSY and ROESY, 13C–1H HSQC and HMBC techniques
(Fig. S6–S18, ESI†). Signal attribution started from tert-butyl (H-
7) and amide (N–H) protons and the other protons were
assigned step by step, using the informations collected from the
COSY and ROESY spectra. The signals of 1Bn4 are relatively
broad in CDCl3, except those of the dangling benzyl protecting
groups and the doublet of equatorial H-3 (3e). The signals of
1H4 are even broader, pointing to the occurrence of hydrogen
bonds involving the hydroxamic acid units. Hence, the spectral
resolution was improved by exchanging CD3OD for CDCl3. In
fact, the spectrum of 1H4 in CD3OD is very similar to the one of
1Bn4 in CDCl3, except that the axial H-3 (3a) protons show up as
the expected doublet, whereas in the case of 1Bn4 the signal
looks like a very broad singlet. The presence of four equivalent
groups of diastereotopic ArCH2Ar methylenic protons for both
1Bn4 and 1H4 indicates that the calix[4]arene scaffolds are also
in the cone conformation in solution.20 The fact that the spectra
are diagnostic of C4-symmetric structures conrms that the
asymmetric C-12 (labelled C47, C61, C75, and C89 in the ORTEP
view of Fig. 3) has the same absolute conguration in each
branch (R) as in compound 7. Chemical shi values of homol-
ogous protons belonging to both compounds are compared in
Table 2. It turned out that the best resolved spectrum of 1H4 was
obtained aer heating the free ligand for 1 h at 60 ꢂC in CD3OD
in the presence of one equivalent of Na+ (Fig. S19, ESI†),
whereas the signal sharpening was less pronounced in the
presence of K+ (Fig. S20, ESI†). These results are in agreement
with the recognized ability of O-substituted calix[4]arenes
appended to –CH2–C(O)– groups to behave as alkali-cation
receptors.12b,14 The consequences of the central chirality of the
C-12 atom of each pendant cyclic hydroxamic acid is better seen
in the 13C NMR spectrum of 1H4 in CD3OD, as the diaster-
eotopic C-2 (C-20) and C-4 (C-40) aromatic carbon atoms clearly
show up as distinct singlets at 135.0 (134.4) and 126.9 (126.5)
ppm, respectively (Table 3). By contrast, in the case of 1Bn4
(CDCl3) these diastereotopic carbon atoms show degenerate
signals at 132.2 and 125.7 ppm, respectively.
Complexation studies
In earlier work we have shown that the cyclic hydroxamic acid 1-
hydroxypiperidine-2-one (1,2-PIPOH) was able to form octa-
coordinated ML4 complexes with tetravalent metal cations
such as Zr4+, Hf4+, Ce4+, Th4+, and U4+.3g Typically, Zr(acac)4 or
Hf(acac)4 was reacted with four equivalents of 1,2-PIPOH in
methanol at room temperature to afford the corresponding
neutral complexes Zr(1,2-PIPO)4 and Hf(1,2-PIPO)4 in 57 and
59% yields aer crystallization from CH2Cl2–heptane 1 : 3 v/v.
Single crystal X-ray analysis of these complexes showed that the
tetravalent metal cations have a distorted square-antiprismatic
geometry of D2 symmetry, because of the alternate arrangement
of the [O–(N–C)]O]ꢀ chelates, two neighbouring ones dis-
playing opposite orientations with respect to the N and C atoms
of the six-membered ring. The H and 13C NMR spectra of the
complexes in CDCl3 evidence sharp signals that are also fully
consistent with D2 symmetry.
1
Hence, in a rst attempt to prepare the zirconium complex of
the tetrahydroxamic acid 1H4, an equimolar amount of free
ligand was reacted directly in a NMR tube with Zr(acac)4 in
CD3OD at room temperature. The 1H NMR spectrum of the
crude reaction mixture showed broad signals in the same
regions as for the free ligand, but also broad features in the
aromatic range around 6.5 and 7.2 ppm, including the charac-
teristic singlet arising from the 4-H protons of uncomplexed
1H4 (Fig. S21b, ESI†). Heating the mixture at 55–60 ꢂC for 1.5 h
resulted in signicant changes in the NMR spectrum, in
particular, the disappearance of all the signals corresponding to
the free ligand and the concomitant growth of clearly identi-
able singlets protruding from broad features in the aromatic
region (Fig. S21c, ESI†). The MALDI-TOF mass spectrum of the
mixture showed, in addition to the expected signals corre-
sponding to the [Zr(1) + Na]+ cation, clusters of peaks corre-
sponding to [Zr2(1)2 + Na]+ and [Zr2(1)2 + K]+.
These latter observations prompted us to run the same
reaction (CD3OD, 60 ꢂC) in the presence of half an equivalent of
KCl. As expected, the intensity of the signal corresponding to
Table 2 Selected 1H NMR data for 1Bn4 and 1H4 in CD3ODa
b
Compd/H-
3a
3e
4,40
7
8,80
12
13,130
14,140
17,170
1Bn4
1H4
4.77
4.72
3.30
3.24
6.91
6.86
1.12
1.10
4.73
4.67
4.64
4.59
1.82/2.02
1.94/2.05
1.93
2.01
4.86/4.91
—
a
d (ppm) vs. TMS. b CH2 protons of the Bn protecting group.
Table 3 Selected 13C NMR data for 1Bn4 and 1H4
a
Compd/C-
1
2,20
3
4,40
5
6
7
8
9
12
b
1Bn4
1H4
155.1
154.5
132.2
135.0, 134.4
32.7
33.0
125.7
126.9, 126.5
145.0
146.5
33.9
34.8
31.5
31.9
74.8
75.4
171.2
172.1
50.2
51.3
c
a
d (ppm) vs. TMS. b CDCl3. c CD3OD.
22746 | RSC Adv., 2014, 4, 22743–22754
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
the K+ adduct at m/z ¼ 2872.88 was enhanced in the MALDI-TOF symmetry upon complexation, becoming asymmetric in the
mass spectrum aer KCl addition (Fig. 4 and S22, ESI†), while complex. As a matter of fact, each of the four legs of the tetrapod
the major signal found by ESI-MS corresponded to the doubly- 14ꢀ is inequivalent in the complex, which is accounted for
charged [Zr2(1)2 + 2K]2+ species (Fig. S24, ESI†). Most interest- hereaer by the use of four different labels (A, C, E, and G). The
1
ingly, the H NMR spectrum of the reaction mixture revealed coordination sphere of each tetravalent metal cation being
also dramatic changes as compared to the one recorded in the satised with four bidentate hydroxamate subunits, dimer
absence of K+ ions (Fig. S25, ESI†), clearly evidencing a strong formation can result from the binding of the bridging metal
interaction of the alkali metal with the calixarene zirconium cations by chelates belonging to two different tetrapods, either
complex in solution rather than a weak ion-pair formation in in the 2 : 2 or the 3 : 1 ratio. A reasonable illustration of the rst
the gas phase. Very similar features were noted in the 1H NMR situation (D2 symmetric complex) is shown in Fig. 6a, where two
spectrum of the corresponding hafnium complex, pointing to binding groups located trans to each other at a given metal
related stoichiometries and structures, as expected considering centre belong to the same calix[4]arene subunit. As a conse-
the almost identical ionic radii of octacoordinated Zr4+ (ri ¼ 0.84 quence, the chelating subunits of the dimer can be partitioned
4+
A) and Hf (ri ¼ 0.83 A).21 Due to slightly higher resolution, the into two groups of equivalent systems schematized by blue and
latter is reproduced in Fig. 5 together with the signal assign- pink colours in Fig. 6a. Obviously, this arrangement cannot
ments. In the aromatic region, several doublets are clustered account for the 1H NMR observations. On the contrary, the
around 7.20 ppm, while four more appear between 6.40 and situation in which three hydroxamate groups are provided by
6.60 ppm. In the 5.7–3.0 ppm range, several doublets of equal one calix[4]arene and the fourth one by the second molecule of
intensities and multiplets with more complex patterns can be ligand in order to complete the coordination sphere of each
recognized. In addition, four predominant singlets scattered metal cation, leads to a C2-symmetric complex in which all four
between 1.25 and 1.50 ppm on the one hand, and between 0.75 chelates are in different chemical environments (Fig. 6b).
˚
˚
and 1.00 ppm on the other hand, can also be identied in the
region of the tert-butyl protons (H-7).
As shown in Fig. 5 for the hafnium complex, nearly complete
assignment of the 1H NMR spectrum could be achieved by a
An illustrative summary of the various 1H NMR complexa- combined analysis of the 1H–1H COSY and ROESY, 13C–1H
tion experiments discussed is presented graphically in Fig. S26 HSQC and HMBC correlation charts (Fig. S27–S33, ESI†).
(ESI†), in the form of a stacked plot of the spectra obtained aer Thanks to the temperature dependence of certain proton reso-
reacting 1H4 with respectively KCl (1 : 1), Hf(acac)4 (1 : 1), and nances, comparison of the spectra recorded at 300 and 335 K
Hf(acac)4 + KCl (1 : 1 : 0.5).
was particularly useful for identifying overlapped signals
Assuming a [M2K(1)2]+ stoichiometry for the complexes, the (Fig. S34, ESI†). The COSY spectrum allowed at rst to recognize
1H NMR spectrum of [Hf2K(1)2]+ reproduced in Fig. 5 would the eight diastereotopic H-8/H-80 and H-3a/H-3e proton pairs,
correspond to a species in which the two Hf(1) subunits are the assignment of which being later conrmed by the analysis
equivalent, while the calix[4]arene ligand has lost its C4 of the HSQC spectrum. In particular, resonances H-3De and H-
3He could be differentiated in the 300 K HSQC spectrum of the
hafnium complex (Fig. S32, ESI†). Correlations found in the
ROESY spectrum (Fig. 7) then allowed identifying the aromatic
groups and the methylene bridges separating them on the one
hand, and the proton pairs H-8/H-80 belonging to the methylene
functions of the acetamide groups bridging the phenol and the
hydroxamate moieties. Observed NOE effects are marked with
arrows in Fig. 8. ROESY cross peaks were found between the
proton signals H-7(tBu)/H-4, H-7/H-40, H-4/H-3e, H-40/H-3e, and,
Fig. 4 MALDI-TOF mass spectrum of the reaction product of 1H4,
Zr(acac)4 and KCl in 1 : 1 : 0.5 ratio. * marks dithranol (used as matrix).
Fig. 5 1H NMR spectrum (CD3OD, 335 K) of the [Hf2K(1)2]+ complex. Fig. 6 Two possible structural arrangements of the [M2K(1)2]+ dimer.
The labels s and * denote solvent and acetylacetone, respectively.
Equivalent chelates are shown in the same colour.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 22743–22754 | 22747

View Article Online
RSC Advances
Paper
by complete analysis of the HMBC spectrum, which also
allowed to locate the aryl C-1, C-2/C-20, C-4/C-40, and C-5, as well
as the amide carbon atoms C-9. In turn, the protons of the
bound hydroxamate units were much more difficult to attribute.
Four multiplets arising from the H-12 protons attached to the
asymmetric carbon centres C-12 and those from the associated
pairs of diastereotopic protons H-13/H-130 could be distin-
guished. However, resonances corresponding to the proton
pairs H-14/H-140 and H-15/H-150 could not be assigned without
ambiguity, mainly because they are clustered in a very narrow
chemical shi range, contrary to the well scattered H-13/H-130
signals. Unfortunately, it was not possible to correlate H-12 with
other identied protons, excepted for H-12C which gives rise to a
through-space correlation with H-3Ba and H-8A in the ROESY
spectrum. Therefore, the other H-12 and H-13/H-130 proton
signals have been distinguished using a different labelling
system (o, +, and ꢃ).
Selected 1H NMR chemical shi values recorded for
[Na(1H4)]+ and [Hf2K(1)2]+ in CD3OD are compared in Table 4.
[Na(1H4)]+ was chosen as reference rather than 1H4, because of
0
0
Fig. 7 ROESY NMR spectrum of the [Hf2K(1)2]+ complex (CD3OD,
335 K, mixing time ¼ 200 ms). Starred spots correspond to impurities.
its increased rigidity.§ Whereas H-4A/H-4A and H-4E/H-4E
resonate in the same range as H-4/H-40 for [Na(1H4)]+ (7.27
0
0
ppm), this is not the case of H-4C/H-4C and H-4G/H-4G , which
give signals around 6.5 ppm. Interestingly, the H-7 proton
signals pertaining to [Hf2K(1)2]+ are also dispatched in two
groups, H-7A and H-7E are deshielded by ca. 0.17 ppm by
comparison with those related to [Na(1H4)]+, whereas H-7C and
H-7G are shielded by ca. 0.35 ppm. The three multiplets arising
from protons H-12o, H-12+, and H-12x are clustered around 4.75
ppm, which is very close to the chemical shi value found for H-
12 in [Na(1H4)]+ (4.79 ppm), while Zr complexation induces a
nearly 1 ppm (3.84 ppm) shielding of the H-12C resonance.
Whereas the equivalent diastereotopic pairs of protons H-8/H-80
show up for [Na(1H4)]+ as an AB spin system (JAB ¼ 14.1 Hz, Dn ¼
55.4 Hz) at 4.56 ppm, only H-8C/H-8C0 among the corresponding
four pairs of [Hf2K(1)2]+ gives such a system at ca. 4.24 ppm. The
three remaining pairs appear as AX spin systems. H-8A and H-8G
(ꢁ5.5 ppm) are deshielded by nearly 1 ppm upon Zr complex-
0
ation, H-8A is the only one to remain almost unaffected by the
0
presence of the metal cation (Dd ¼ ꢀ0.031 ppm), while H-8G
(3.662 ppm) is noticeably the most shielded signal among the
H-80 ones. Moreover, the 2J coupling constant for H-8G/H-8G
0
(16.2 Hz) is signicantly higher than those measured for the
other sets of H-8/H-80 resonances (2J ꢁ 12–13 Hz). However, the
most striking behaviour is that exhibited by proton H-8E, the
resonance of which being deshielded by 2.69 ppm! In
[Na(1H4)]+, proton H-3a resonates at 4.54 ppm, a value from
which the chemical shis of H-3Fa, H-3Ba, and H-3Ha do not
depart signicantly, whereas the signal of H-3Da is shied
downeld by ca. 0.74 ppm from that value. By contrast, all of the
H-3e protons show up between 3.1 and 3.5 ppm, a chemical-
shi range that includes H-3e of [Na(1H4)]+. As in [Na(1H4)]+,
the multiplets assigned to the methylenic hydroxamate protons
H-15/H-150 and the H-14/H-140 are clustered around ca. 3.7 and
Fig. 8 Map summarizing the NOE cross-correlation peaks found in
the ROESY spectrum of the [M2K(1)2]+ complexes. The coloured and
grey double-ended arrows represent trivial NOE correlations, whereas
the black ones point to key correlations for the structural analysis. The
drawing is focused on the calix[4]arene ligand that provides three out
of the four hydroxamate binding units for chelating the metal centre
M1. The fourth group, denoted E(2), bound to M1 but belonging to the
second calix[4]arene molecule, labelled calix(2), is also represented
together with its symmetry-related component E that binds M2. The
potassium cation is omitted.
more specically, H-3Ba/H-8A, H-3Fa/H-8E, H-3Ha/H-8G, H-4A/H-
0
0
0
0
§ [Na(1H4)]+ was chosen rather than [K(1H4)]+ because it showed better resolved
4G , H-4G/H-4E , H-4E/H-4C , and H-4C/H-4A . The assignments
drawn from these correlations were checked for self-consistency signals.
22748 | RSC Adv., 2014, 4, 22743–22754
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
Table 4 Comparison of selected 1H NMR data for [Na(1H4)]+ and
[Hf2K(1)2]+ (d in ppm) and complexation induced shifts (CIS, Dd) at 300
K (CD3OD)
correlation. The unusual downeld shis of signals H-8A, H-8E,
H-8G, and H-3Da could be due to deshielding interstrand (for H-
8 protons) amide NH/O contacts. These could explain the
simultaneous deshielding of H-8E and H-3Fa, which also show a
NOE contact. There is one additional NOE that was not
mentioned so far, namely the one between H-12+ and a H-15/H-
150 proton. Handling of CPK models suggests that H-12+ is H-
Proton/compd
[Na(1H4)]+a
[Hf2(1)2K]+a
Ddb
[Hf2(1)2K]+c
3a
4.537
3Ba
3Da
3Fa
3Ha
3e
3Be
3De
3Fe
3He
4,40
4A
4.445
5.275
4.634
4.157
ꢀ0.092
0.738
0.097
ꢀ0.380
4.461
5.288
4.657
4.198
0
12E and H-15/H-150 is H-15C/H-15C , see Fig. 8. All these corre-
lations are consistent with the hypothetical structure of a
binuclear complex in which the hydroxamate groups of A, C,
and G of a calixarene subunit bind to one M4+ cation together
with the hydroxamate group E of the other calixarene subunit,
providing a coordination environment that should be close to a
dodecahedron (Fig. 6b and Scheme 3).
3.435
7.273
3.436
3.163
3.436
3.157
0.001
ꢀ0.272
0.001
ꢀ0.278
3.394
3.156
3.424
3.156
To complement these investigations, diffusion ordered
spectroscopy (DOSY) NMR experiments were performed in
CD3OD (Fig. 9). The DOSY chart for the [Zr2K(1)2]+ complex
revealed the occurrence in solution of a single diffusing species
with D ¼ 301 mm2 sꢀ1 in addition to methanol-d4 (D ¼ 2214 mm2
sꢀ1), whereas the diffusion coefficient for the corresponding
hafnium complex is 367 mm2 sꢀ1 in the same solvent (D ¼ 2250
mm2 sꢀ1). Taking the value measured for the in situ generated
monomeric inclusion complex [K(1H4)]+ as a reference (D ¼ 461
vs. 2244 mm2 sꢀ1 for CD3OD), the diffusion coefficients found
for the dimeric zirconium and hafnium complexes are smaller
by a factor of 0.65 and 0.80, respectively. It must be noted that a
very similar variation was observed between a meso-substituted
zinc porphyrin precursor bearing four allyl-functionalized pol-
yether chains and the corresponding diporphyrinic cage
obtained by ring-closing metathesis.22
7.173
6.388
7.226
6.568
7.205
6.573
7.214
6.452
ꢀ0.100
ꢀ0.885
ꢀ0.047
ꢀ0.705
ꢀ0.068
ꢀ0.700
ꢀ0.059
ꢀ0.821
7.168
6.407
7.222
6.572
7.189
6.577
7.205
6.455
4C
4E
4G
0
4A
0
0
4C
4E
0
4G
7
1.193
4.559
7A
7C
7E
7G
8,80
8A
8C
8E
8G
1.355
0.816
1.375
0.860
0.162
ꢀ0.377
0.182
ꢀ0.333
1.353
0.823
1.374
0.858
5.555
4.328
7.253
5.407
4.528
4.142
3.981
3.662
0.996
ꢀ0.231
2.694
5.563
4.388
7.209
5.414
4.578
4.130
3.999
3.675
0.848
0
8A
ꢀ0.031
ꢀ0.417
ꢀ0.578
ꢀ0.897
0
0
Taken altogether, 1H NMR observations (spectral analysis
and DOSY) as well as mass spectrometry results fully support
the formation of a dimer having [M2K(1)2]+ formula upon
reaction of M(acac)4 with the calix[4]arene tetrapod 1H4
8C
8E
0
8G
12
4.793
12o
12+
12x
12C
—
—
4.834
4.748
—
d
d
4.674
3.841
ꢀ0.119
ꢀ0.952
3.867
a
T ¼ 300 K. b CIS expressed as Dd ¼ d([Hf2K(1)2]+) ꢀ d([Na(1H4)]+). c T ¼
335 K. d Hidden.
2.0 ppm, respectively. In contrast, those corresponding to the H-
13/H-130 pairs of protons are scattered between 2.8 and 1.1 ppm,
the most deshielded massif being assigned to H-13x and the
0
most shielded one to H-13C , while the largest Dd value (1.58
x
x
0
ppm) for diastereotopic protons is observed for the H-13 /13
pair.
The distribution of the proton signals that belong0 to oppo-
0
site aromatic systems, i.e. H-4A/H-4A and H-4E/H-4E (respec-
tively H-7A and H-7E) on the one hand, and H-4C/H-4C0 and H-4G/
0
H-4G (respectively H-7C and H-7G) on the other hand, with one
system shielded by comparison with the other, indicates that
the calixarene platforms take up a pinched-cone conformation
in [Hf2K(1)2]+, as found in the crystal structure of 1Bn4. Inter-
0
estingly, the most AB-like H-8/H-80 spin system (the H-8C/H-8C
Scheme
3 Summary of the coordination chemistry of 1H4 with
pair), is also the only one that does not show any H-8/H-3a NOE M ¼ Zr4+ or Hf4+
.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 22743–22754 | 22749

View Article Online
RSC Advances
Paper
the unexpectedly high deshielding effect observed for some
signals arising from the H-8 protons, as these might effectively
sense the presence of a nearby located K+ cation. To the best of
our knowledge, the [M2K(1)2]+ (M ¼ Zr and Hf) complexes
represent the rst example of a supramolecular assembly
involving two M4+ cations and two calix[4]arene moieties which
share their pendant chelating groups in the coordination
sphere of each metal centre, as all the other systems described
so far involve m-bridging ligands, such as the oxide anion23a or
CH3OH.23b
Our initial design of the calix[4]arene-based hydroxamate
tetrapod 14ꢀ relied on the hypothesis that the M4+ cation would
arrange the four 1,2-PIPOꢀ binding units in a four-bladed
propeller-like fashion, thus imparting a L or D helicity to the
metal centre (Fig. 2). As each of the chelated hydroxamate
incorporates an asymmetric carbon atom (C-12) of R congu-
ration, chirality induction could furthermore be anticipated.
Accordingly, the bidentate ligands were expected to span four
equivalent lateral edges of a twisted square antiprism dened
by the oxygen atoms located at the eight vertices, as shown
schematically in Fig. 10a. However, the recently solved X-ray
crystal structures of Zr(1,2-PIPO)4 and Hf(1,2-PIPO)4 complexes
revealed an alternative layout to the C4-symmetric arrangement,
in which the ligands are positioned along opposite edges of the
twisted square faces, in pairwise opposite orientations
(Fig. 10b).3g Clearly, this binding scheme of ideal D2 symmetry
cannot be achieved by a single tetrapod appended with four
identical chelating motifs for a 1 : 1 metal-to-ligand species, but
in turn can be satised by the aforementioned [1 + 3] dimer-
ization process. As a result, a simpler way of encapsulating an
octacoordinated cation would be to functionalize the calix[4]-
arene platform with only two pendant 1,2-PIPOH units, in order
to favour a mononuclear species of ML2 stoichiometry.
Fig. 9 Comparison of the 1H DOSY NMR spectra (CD3OD, 300 K) of an
equimolar mixture of 1H4 and KCl (a) and of 1H4, Hf(acac)4, and KCl in
1 : 1 : 0.5 molar ratio (b).
Conversely, octacoordinated metals ions larger than Zr4+ or
Hf4+ are expected to favour approaching dodecahedral coordi-
nation geometries of limiting ideal D2d symmetry in order to
featuring 1,2-PIPOH hydroxamic acid binding motifs. As NMR minimize the electrostatic repulsion energy between the oxygen
spectroscopy turned out to be the most valuable structural donors.24 For simple bidentate units, D2d dodecahedral stereo-
investigation technique, most small scale reactions were run at chemistries are typically observed when the normalized bite b is
ꢁ0.01 M concentration levels, which might favour aggregation less than 1.1 [b is dened as the O–O/O–M distance ratio or
processes. These are limited to dimerization probably thanks to
the preorganization effect of the calix[4]arene platform.
However, under more dilute conditions, typically ꢁ10ꢀ5 M, the
2 : 2 species still prevail in solutions as clearly evidenced by HR-
ESI-MS analyses. Thus, aggregation of two M(1) monomers at
higher concentrations, which could be driven by electrostatic or
hydrogen-bonding interactions, can be condently ruled out.
Unfortunately, all attempts to crystallize a zirconium or
hafnium complex failed, whether an alkali salt was present
or not.
For symmetry and stoichiometry reasons, we have arbitrarily
located the K+ cation in the middle of the molecule, where it
could interact with the 1,2-PIPOꢀ oxygen atoms, but the possi-
Fig. 10 Slightly distorted square-antiprismatic coordination geome-
bility that it could be bound by the eight oxygen atoms of the
tries. (a) Hypothesized arrangement of the bidentate (ON-OC) 1,2-
–OCH2C(O)– bridges and shuttle rapidly (with respect to the 1H
PIPOꢀ ligands required to form a complex of M(1) stoichiometry. (b)
NMR timescale) between these two equivalent sites cannot be
ruled out. Some support for this interpretation is provided by the X-ray crystal structure of Zr(1,2-PIPO)4.3g
Experimentally observed arrangement of the hydroxamate groups in
22750 | RSC Adv., 2014, 4, 22743–22754
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
alternatively as a function of the O–M–O bite angle b according
to b ¼ 2 sin(b/2)]. Experimentally, An(1,2-PIPO)4 complexes (An
¼ Th4+ or U4+) full that criterion.3g Hence, it cannot be
excluded that these ions might be tetrachelated by the four
dangling hydroxamates of ligand 14ꢀ and afford the sought 1 : 1
inclusion complex. Indeed, provided that the arms are exible
and long enough, the scaffold could properly orient the binding
groups around a single metal centre allowing them to span the
lateral edges of an approaching dodecahedron, a situation close
to the one depicted in Fig. 10a. Work going in that direction is
currently under progress.
X-ray crystallography
X-ray equipment and renement. Diffraction data were
collected on a Nonius Kappa Apex II diffractometer equipped
with a nitrogen jet stream low-temperature system (Oxford
Cryosystems). The X-ray source was graphite monochromated
˚
Mo-Ka radiation (l ¼ 0.71073 A) from a sealed tube. The lattice
parameters were obtained by least-squares t to the optimized
setting angles of the entire set of collected reections. No
signicant temperature dri was observed during the data
collections. Data were reduced by using HKL Denzo and Scale-
pack soware without applying absorption corrections,26 the
missing absorption corrections were partially compensated for
by the data scaling procedure in the data reduction. The
structure was solved by dual space methods using the SHELXD
program.27 Renements were carried out by full-matrix least-
squares on F2, using the SHELXL program on the complete set
of reections.27 Anisotropic thermal parameters were used for
non-hydrogen atoms, excepted for disordered molecules of
solvent. All H atoms bound on carbon, nitrogen, or oxygen
atoms were placed at calculated positions using a riding model
Conclusions
Developing abiotic metal-uptake agents inspired from natural
ligands has proven to be useful in the past, and our approach is
nothing but a similar one, using an iron(III) binding motif (1,2-
PIPOH) found in a few number of bacterial siderophores. So far
essentially ignored by the coordination chemists, cyclic
aliphatic hydroxamates, like the six-membered 1,2-PIPOꢀ,
represent an interesting alternative to the analogous but
aromatic, and thus slightly more acidic, 1,2-HOPO, the binding
motif of the single chelate siderophore cepabactin.25 High-val-
ent transition metals, including lanthanides and actinides,
have high coordination numbers (CN), generally equal to or
greater than 8. For CN ¼ 8, the most frequently observed
coordination polyhedra are the square-antiprism and the
trigonal-faced dodecahedron. Accommodation of additional
ligands is achieved by capping one or more faces of these
polyhedra. It is clear from this study that the design of tetra-
chelating C4-symmetric podands for metal ions favouring
octacoordination is not a trivial task and requires further tuning
for achieving a 1 : 1 metal-to-ligand stoichiometry of the com-
plexed species.
˚
˚
˚
with C–H ¼ 0.95 A (aromatic), 0.98 A (methyl), 0.99 A (methy-
˚
˚
lene), N–H ¼ 0.88 A, and O–H ¼ 0.84 A, assuming Uiso(H) ¼
1.2Ueq.(CH), Uiso(H) ¼ 1.2Ueq.(CH2), Uiso(H) ¼ 1.5Ueq.(CH3),
Uiso(H) ¼ 1.2Ueq.(NH), and Uiso(H) ¼ 1.5Ueq.(OH). Similar Uij
constraints were applied within the terminal benzyl groups to
maintain a reasonable model by using EADP constraints.27 The
geometric parameters in each group were restrained by using
AFIX 66 restraints.27 Heavily-disordered methanol (s.o.f. ¼ 0.75)
and water molecules (s.o.f. ¼ 0.25) were rened isotropically.
Crystal data for 1H4. C100H120N8O16$0.75CH3OH$0.25$H2O,
M ¼ 1718.06 g molꢀ1, monoclinic space group P21 (no. 4), a ¼
ꢂ
˚
˚
˚
15.6125(2) A, b ¼ 18.2100(3) A, c ¼ 18.3284(3) A, b ¼ 101.79(1) ,
3
V ¼ 5100.9(3) A , T ¼ 115 K, Z ¼ 2, m(Mo-Ka) ¼ 0.076 mmꢀ1, dcalc
˚
¼ 1.119 g cmꢀ3, 66 668 reections measured (8.154 # 2q #
54.934ꢂ), 22 535 unique (Rint ¼ 0.0437) which were used in all
calculations. The nal R1 was 0.0703 [I > 2s(I)] and wR2 was
0.2007 (all data).
Experimental
General methods and materials
Synthesis
Unless otherwise noted, all chemicals and starting materials
were obtained commercially and used without further puri-
Compound 8. 5,11,17,23-Tetra-tert-butyl-25,26,27,28-tetra-
cation. Toluene and tetrahydrofuran (THF) were dried over carboxymethoxycalix[4]arene14 (0.500 g, 0.57 mmol) and thionyl
sodium/benzophenone, and methanol was dried over magne- chloride (4.0 mL, 55.14 mmol) were heated in reuxing dry
sium turnings (10% w/v) and iodine (1% w/v). All dried solvents toluene (8 mL) for 4 h. Excess thionyl chloride and the solvent
1
were distilled immediately prior to use. H NMR spectra were were removed under reduced pressure to afford compound 8
recorded on a 600 MHz Bruker Avance II spectrometer and (0.340 g) in 62% yield, which was used without purication in
referenced internally using the residual protio solvent reso- the next step. 1H NMR (CDCl3, 600 MHz, 300 K): d 1.08 (s, 36H),
nances relative to Me4Si. Electrospray ionization mass spectra 3.27 (d, 4H, 2J ¼ 12.0 Hz), 4.61 (d, 4H, 2J ¼ 12.0 Hz), 5.12 (s, 8H),
were obtained either on a LTQ Orbitrap XL (HR-MS) or an 6.80 (s, 8H) ppm.
Amazon SL instrument. MALDI-TOF mass spectra were recor-
Compound 1Bn4. A solution of 7$HBr13 (0.422, 1.40 mmol),
ded using an Ultraex II LRF 2000 instrument and 1,8-dihy- Et3N (0.152 g, 1.50 mmol, 0.21 mL) in THF (25 mL) was stirred
droxy-9,10-dihydroanthracen-9-one (dithranol) as matrix. All of for 1 h and transferred via cannula into a stirred solution of the
ˆ
these measurements were performed at the “Pole Chimie chlorocarbonylmethoxy derivative 8 (0.340 g, 0.35 mmol) in THF
´
Moleculaire”, the technological platform for chemical analysis (20 mL). The reaction mixture was reuxed for 45 h then diluted
and molecular synthesis (http://www.wpcm.fr) of the Institut de with 150 mL of CH2Cl2, washed with water (3 ꢃ 20 mL), dried
´
´
Chimie Moleculaire de l'Universite de Bourgogne and over MgSO4, ltered and concentrated on a rotary evaporator to
Welience™.
give a yellowish crude product (0.678 g), which was fractionally
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 22743–22754 | 22751

View Article Online
RSC Advances
Paper
crystallized from hot CH3OH (2 mL) to afford pure compound reaction mixture was heated at 60 ꢂC overnight, aer which time
1
1
1Bn4 (0.249 g) in 49% yield. Mp 187.4–188.1 ꢂC. 1H NMR (CDCl3, the changes in the H NMR spectrum were recorded. H NMR
600 MHz, 300 K): d 8.473 (br s, 4H, NH), 7.389 (m, 8H, H-19), (CD3OD, 600 MHz, 300 K): d 7.02 (s, 8H, H-4, H-40), 4.64 (d, 8H, 2J
7.295 (m, 12H, H-20, H-21), 6.763 (br d, 8H, 2J ¼ 12.6 Hz, H-4, H- ¼ 12.6 Hz, H-3a, H-12), 4.61 (AB, 8H, JAB ¼ 13.6 Hz, Dn ¼ 95 Hz,
2
40), 4.990 (d, 4H, J ¼ 10.2 Hz, H-170), 4.933 (br s, 4H, H-3a), H-8, H-80), 3.72–3.59 (m, 8H, H-15, H-150), 2.12–1.96 (m, 12H, H-
4.861 (hidden, 4H, H-12), 4.852 (d, 8H, 2J ¼ 10.2 Hz, H-17), 4.803 13, H-14, H-140), 1.94–1.82 (m, 4H, H-130), 1.14 (s, 36H, H-7)
(br d, 8H, 2J ¼ 9.6 Hz, H-8), 3.385 (m, 4H, H-150), 3.273 (br d, 4H, ppm (H-3e hidden under the solvent signal).
H-15), 3.219 (d, 4H, 3J ¼ 13.2 Hz, H-3e), 1.89–2.01 (m, 8H, H-130,
Reaction of complex [K(1H4)]+ with Zr(acac)4 and 1H4. To the
H-14, H-140), 1.710 (br s, 4H, H-13), 1.077 (s, 36H, H-7) ppm. 13C previous solution were added 1 equiv. of solid 1H4 (5.8 mg, 4.36
NMR (CDCl3, 150 MHz, 300 K): d 171.2 (C-9), 167.8 (C-16), 155.1 mmol) and 2 equiv. of Zr(acac)4 (4.1 mg, 8.34 mmol). The reaction
1
ꢂ
(C-1), 145.0 (C-5), 135.7 (C-18), 132.2 (C-2), 129.6 (C-19), 128.6 mixture was heated at 60 C overnight and the H NMR spec-
1
(C-21), 128.5 (C-20), 125.7 (C-4), 75.8 (C-17), 74.8 (C-8), 50.8 (C- trum subsequently recorded. H NMR (CD3OD, 600 MHz, 295
15), 50.2 (C-12), 33.9 (C-6), 32.7 (C-3), 31.5 (C-7), 27.9 (C-14), 21.1 K): d 7.258 (d, 2H, 2J ¼ 13.8 Hz, H-8E), 7.226 (d, 2H, 4J ¼ 2.4 Hz,
(C-13) ppm. HR-MS (ESI) m/z calcd for [C100H120N8O16Na]+: H-4E), 7.215 (d, 2H, J ¼ 3.0 Hz, H-4E ), 7.210 (d, 2H, J ¼ 2.4 Hz,
4
4
0
4
4
0
1711.871 (monoisotopic mass), found: 1711.868. Elemental H-4A ), 7.174 (d, 2H, J ¼ 2.4 Hz, H-4A), 6.574 (d, 2H, J ¼ 2.4 Hz,
4
4
0
analysis % calcd for C100H120N8O16: C, 71.07; H, 7.16; N, 6.63. H–4C ), 6.564 (d, 2H, J ¼ 2.4 Hz, H-4G), 6.453 (d, 2H, J ¼ 3.0 Hz,
4
2
0
Found: C, 70.30; H, 7.20; N, 6.67.
H-4G ), 6.385 (d, 2H, J ¼ 2.4 Hz, H–4C), 5.545 (d, 2H, J ¼ 12.6
2
2
Compound 1H4. A mixture of 1Bn4 (0.100 g, 60 mmol) and Hz, H-8A), 5.412 (d, 2H, J ¼ 16.2 Hz, H-8G), 5.274 (d, 2H, J ¼
Pd(OH)2/C (20%, 50 mg) in CH3OH (8 mL) was stirred under an 12.6 Hz, H-3Da), 4.850 (hidden, 2H, 12ꢂ), 4.758 (dd, 2H, 3J ¼ 10.2
atmosphere of dihydrogen for 15 min. It was subsequently Hz, 3J ¼ 6.0 Hz, H-12+), 4.707 (d, 2H, 3J ¼ 4.2 Hz, H-12x), 4.628 (d,
2
2
0
ltered through a pad of silica (70–200 mm), which was washed 2H, J ¼ 13.2 Hz, H-3Fa), 4.520 (d, 2H, J ¼ 12.6 Hz, H-8A ), 4.443
2
2
with CH3OH and CH2Cl2. The ltrate was evaporated to dryness (d, 2H, J ¼ 13.2 Hz, H–3Ba), 4.336 (d, 2H, J ¼ 12.0 Hz, H-8C),
to afford pure 1H4 (0.072 g) in 93% yield. Mp 194.0–195.0 ꢂC. 1H 4.154 (d, 2H, 2J ¼ 12.6 Hz, H-3Ha), 4.142 (d, 2H, 2J ¼ 12.6 Hz, H-
2
3
0
0
NMR (CD3OD, 600 MHz, 300 K): d 8.847 (br s, 4H, NH), 6.855 (br 8C ), 3.978 (d, 2H, J ¼ 13.8 Hz, H-8E ), 3.858 (dd, 4H, J ¼ 9.0 Hz,
s, 8H, H-4, H-40), 4.723 (d, 4H, 2J ¼ 13.2 Hz, H-3a), 4.666 (AB, 8H, H-12C, H-15, H-150), 3.75–3.50 (m, 14H, H-15, H-150), 3.660 (d,
2
2
JAB ¼ 13.8 Hz, Dn ¼ 23.2 Hz, H-8,80), 4.588 (m, 4H, H-12), 3.672 2H, H-8G ), 3.438 (d, 2H, J ¼ 12.0 Hz, H-3Fe), 3.387 (d, 2H, J ¼
0
(m, 4H, H-150), 3.598 (m, 4H, H-15), 3.236 (d, 4H, 2J ¼ 13.2 Hz, 13.2 Hz, H-3Be), 3.163 (d, 2H, 2J ¼ 13.2 Hz, H-3De), 3.159 (d, 2H,
H-3e), 2.05 (m, 12H, H-130, H-14, H-140), 1.940 (br m, 4H, H-13), 2J ¼ 13.8 Hz, H-3He), 2.745 (m, 2H, H-13x), 2.679 (m, 2H, H-13+),
1.100 (s, 36H, H-7) ppm. 13C NMR (CD3OD, 150 MHz, 300 K): d 2.2–1.9 (several m, 16H, H-13ꢂ, H-13 , H-14, H-140), 1.765 (two
+
0
2
x
0
172.1 (C-9), 167.2 (C-16), 154.5 (C-1), 146.5 (C-5), 135.0 (C-20), m, 4H, H-13C, H-13ꢂ0), 1.655 (br d, 2H, J ¼ 14.4 Hz, H-13 ),
134.4 (C-2), 126.9 (C-40), 126.5 (C-4), 75.4 (C-8), 52.3 (C-15), 51.3 1.556 (m, 2H, H-140), 1.375 (s, 18H, H-7E), 1.354 (s, 18H, H-7A),
0
(C-12), 34.8 (C-6), 33.0 (C-3), 31.9 (C-7), 28.6 (C-13), 21.8 (C-14) 1.201 (m, 2H, H-13C ), 0.861 (s, 18H, H-7G), 0.814 (s, 18H, H-7C).
ppm. HR-MS (ESI) m/z calcd for [C72H96N8O16Na]+: 1351.684, MALDI-TOF MS m/z calcd for [C144H184KN16O32Zr2]+: 2868.099,
found: 1351.686; calcd for [C72H96N8O16Na2]2+: 687.336, found: found: 2868.910 (monoisotopic mass).
687.337. Elemental analysis % calcd for C72H96N8O16$4H2O: C,
61.70; H, 7.48; N, 7.99. Found: C, 61.49; H, 7.19; N, 7.78.
Complexation of 1H4 with Zr(acac)4. Zr(acac)4 (0.75 mg,
1.50 mmol) in CD3OD (0.5 mL) was added to a solution of 1H4
(1.90 mg, 1.43 mmol) in CD3OD (0.5 mL). The reaction mixture
was heated to 55 ꢂC overnight and analyzed by 1H NMR
(Fig. S21, ESI†).
Complexation experiments
Reaction of 1H4 with NaCl. 1H4 (4.08 mg, 3.07 mmol) was
One-pot reaction of 1H4 with Zr(acac)4 and KCl. A mixture of
dissolved in CD3OD (0.7 mL) and the 1H NMR spectrum of the 1H4 (0.062 g, 47 mmol), Zr(acac)4 (0.023 g, 47 mmol), and KCl (1.8
resulting solution recorded. Next, solid NaCl (0.18 mg, 3.08 mg, 23.5 mmol) in CH3OH (4 mL) was heated to 60 ꢂC. The
mmol) was added to the solution of the calix[4]arene 1H4 and the mixture became homogeneous aer 15 min at 60 ꢂC. The
1
ꢂ
reaction mixture heated to 60 C overnight. Changes in the H solvent was evaporated, leaving a yellowish solid (0.074 g).
NMR spectrum were recorded. 1H NMR (CD3OD, 600 MHz, 300 Analysis by H NMR showed that this residue was identical to
1
K): d 7.273 (s, 8H, H-4, H-40), 4.793 (dd, 4H, 3J ¼ 10.2 Hz, 3J ¼ 5.4 the material obtained by sequential synthesis.
Hz, H-12), 4.559 (AB, 8H, JAB ¼ 14.1 Hz, Dn ¼ 55.4 Hz, H-8, H-80),
Reaction of 1H4 with Hf(acac)4 followed by addition of KCl. A
4.537 (d, 4H, 2J ¼ 12.6 Hz, H-3a), 3.71–3.63 (m, 8H, H-15, H-150), solution of Hf(acac)4 (2.61 mg, 4.6 mmol) in CD3OD (0.2 mL) was
3.435 (d, 4H, 2J ¼ 12.6 Hz, H-3e), 2.15–2.22 (m, 4H, H-13), 2.00– added to a suspension of 1H4 (6.05 mg, 4.6 mmol) in CD3OD (0.4
1
2.14 (m, 8H, H-14, H-140), 1.839 (m, 4H, H-130), 1.193 (s, 36H, H- mL). The reaction mixture became immediately clear. The H
7) ppm. 13C NMR (CD3OD, 150 MHz, 300 K): d 171.6 (C-9), 167.3 NMR spectrum was run at 333 K to prevent precipitation. Next,
(C-16), 152.3 (C-1), 149.4 (C-5), 136.1 (C-2), 127.4 (C-4), 75.8 (C- KCl (0.76 mg, 10.2 mmol) was added into the NMR tube and the
1
ꢂ
8), 52.4 (C-15), 51.0 (C-12), 35.2 (C-6), 31.8 (C-7), 31.3 (C-3), 28.8 reaction mixture heated at 65 C overnight. H NMR (CD3OD,
(C-13), 21.8 (C-14) ppm.
600 MHz, 300 K; signals marked with an * correspond to the
Reaction of 1H4 with KCl. 1H4 (5.3 mg, 3.99 mmol) was dis- spectrum recorded at 335 K): d 7.253 (d, 2H, 2J ¼ 13.8 Hz, H-8E),
4
4
0
solved in CD3OD (0.7 mL) and solid KCl (0.30 mg, 4.0 mmol) was 7.226 (d, 2H, J ¼ 2.4 Hz, H-4E), 7.214 (d, 2H, J ¼ 2.4 Hz, H-4E ),
4
4
0
subsequently added to the solution of the calix[4]arene. The 7.205 (d, 2H, J ¼ 2.4 Hz, H-4A ), 7.173 (d, 2H, J ¼ 2.4 Hz, H-4A),
22752 | RSC Adv., 2014, 4, 22743–22754
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
4
4
0
0
¨
¨
6.573 (d, 2H, J ¼ 3.0 Hz, H-4C ), 6.568 (d, 2H, J ¼ 2.4 Hz, H-4G),
R. Schirrmacher, B. Wangler and C. Wangler, Molecules,
2013, 18, 6469; (h) E. W. Price and C. Orvig, Chem. Soc.
Rev., 2014, 43, 260.
4
4
6.452 (d, 2H, J ¼ 2.4 Hz, H-4G ), 6.388 (d, 2H, J ¼ 1.8 Hz, H-4C),
5.555 (d, 2H, 2J ¼ 12.6 Hz, H-8A), 5.407 (d, 2H, 2J ¼ 16.2 Hz, H-
8G), 5.275 (d, 2H, 2J ¼ 12.0 Hz, H-3Da), *4.834 (dd, 2H, 3J ¼ 10.8
Hz, 3J ¼ 10.8 Hz, H-12ꢂ), *4.748 (d, 2H, 3J ¼ 10.2 Hz, 3J ¼ 6.0 Hz,
H-12+), 4.674 (d, 2H, 3J ¼ 3.0 Hz, H-12x), 4.634 (d, 2H, 2J ¼ 13.2
3 (a) K. F. Fouche, H. J. Le Roux and F. Phillips, J. Inorg. Nucl.
Chem., 1970, 32, 1949; (b) K. Bhatt and Y. K. Agrawal, Synth.
Inorg. Met.-Org. Chem., 1972, 2, 175; (c) Y. K. Agrawal and
J. P. Shukla, J. Indian Chem. Soc., 1974, 51, 373; (d) M. Pal
and R. N. Kapoor, Indian J. Chem., 1980, 19A, 912; (e)
T. Moav, A. Hatzor, H. Cohen, J. Libman, I. Rubinstein and
A. Shanzer, Chem.–Eur. J., 1998, 4, 502; (f) F. Guerard,
Y.-S. Lee, R. Tripier, L. P. Szajek, J. R. Deschamps and
M. W. Brechbiel, Chem. Commun., 2013, 49, 1002; (g)
P. Jewula, J.-C. Berthet, J.-C. Chambron, P. Thuery,
Y. Rousselin and M. Meyer, manuscript in preparation.
4 P. E. Riley, K. Abu-Dari and K. N. Raymond, Inorg. Chem.,
1983, 22, 3940.
5 G. J.-P. Deblonde, M. Sturzbecher-Hoehne and R. J. Abergel,
Inorg. Chem., 2013, 52, 8805.
6 (a) J. Xu, P. W. Durbin, B. Kullgren, S. N. Ebbe, L. C. Uhlir and
K. N. Raymond, J. Med. Chem., 2002, 45, 3963; (b)
R. J. Abergel, P. W. Durbin, B. Kullgren, S. N. Ebbe, J. Xu,
P. Y. Chang, D. Y. Bunin, E. A. Blakely, K. A. Bjornstad,
C. J. Rosen, D. K. Shuh and K. N. Raymond, Health Phys.,
2010, 99, 401.
7 G. J. Sharman, D. H. Williams, D. F. Ewing and C. Ratledge,
Chem. Biol., 1995, 2, 553.
8 S. Kodani, M. Ohnishi-Kameyama, M. Yoshida and K. Ochi,
Eur. J. Org. Chem., 2011, 3191.
9 S. Dhungana, M. J. Miller, L. Dong, C. Ratledge and
A. L. Crumbliss, J. Am. Chem. Soc., 2003, 125, 7654.
2
2
0
Hz, H-3Fa), 4.528 (d, 2H, J ¼ 12.6 Hz, H-8A ), 4.445 (d, 2H, J ¼
13.2 Hz, H-3Ba), 4.328 (d, 2H, 2J ¼ 12.6 Hz, H-8C), 4.157 (d, 2H, 2J
2
0
¼ 13.2 Hz, H-3Ha), 4.142 (d, 2H, J ¼ 12.0 Hz, H-8C ), 3.981 (d,
2
0
´
2H, J ¼ 13.8 Hz, H-8E ), 3.841 (br dd, 4H, H-12C, H-15), 3.75–
2
3.50 (m, 14H, H-15, H-150), 3.662 (d, 2H, J ¼ 16.2 Hz, H-8G ),
3.436 (d, 4H, 2J ¼ 13.8 Hz, H-3Be, H-3Fe), 3.163 (d, 2H, 2J ¼ 12.6
Hz, H-3De), 3.157 (d, 2H, 2J ¼ 13.2 Hz, H-3He), 2.780 (tt, 2H, 2J ¼
13.8 Hz, H-13x), 2.676 (br q, 2H, H-13+), 2.137 (hidden, 4H, H-
0
´
+
x
13ꢂ, H-13 ), 2.048 (hidden, 2H, H-14 ), 2.010 (hidden, 2H, H-
14ꢂ), 1.993 (hidden, 2H, H-14+), 2.15–1.85 (m, 8H, H-14, H-140),
1.788 (br AB, 2H, H-13C), 1.735 (br AB, 2H, H-13ꢂ0), 1.649 (br d,
0
2
x
0
2H, J ¼ 14.4 Hz, H-13 ), 1.546 (m, 2H, H-140), 1.375 (s, 18H, H-
0
7E), 1.355 (s, 18H, H-7A), 1.200 (m, 2H, H-13C ), 0.860 (s, 18H, H-
7G), 0.816 (s, 18H, H-7C). MALDI-TOF MS m/z calcd for
[C144H184KN16O32Hf2]+: 3047.212 (average mass), found:
3047.194 (peak of maximum intensity).
Reaction of 1H4 with KCl followed by addition of Hf(acac)4. A
mixture of 1H4 (64.8 mg, 48.7 mmol) and KCl (1.82 mg, 24.4
mmol) in CH3OH (20 mL) was heated at reux for 4 h. Next
Hf(acac)4 (28 mg, 48.7 mmol) in CH3OH (5 mL) was added via
cannula to the reaction mixture, which was further heated at
reux for 18 h. The mixture became homogeneous aer 15 min
at 60 ꢂC. The solvent was evaporated, leaving a brown solid (90
mg). Analysis by 1H NMR and MALDI-TOF MS showed that this
residue was identical to the material obtained in the previous 10 (a) M. Lee, S.-J. Lee and L.-H. Jiang, J. Am. Chem. Soc., 2004,
experiment.
126, 12724; (b) F. Corbellini, A. Mulder, A. Sartori,
M. J. W. Ludden, A. Casnati, R. Ungaro, J. Huskens,
M. Crego-Calama and D. N. Reinhoudt, J. Am. Chem. Soc.,
2004, 126, 17050; (c) H. Huang, D. M. Li, W. Wang,
Y.-C. Chen, K. Khan, S. Song and Y.-S. Zheng, Org. Biomol.
Chem., 2012, 10, 729.
Acknowledgements
This work was supported by the Centre National de la Recher-
che Scientique (CNRS), the Conseil Regional de Bourgogne
´
(program PARI IME SMT8), and the Groupement National de 11 (a) R. Ludwig, H. Matsumoto, M. Takeshita, K. Ueda and
Recherche GNR PARIS. P.J. thanks the CNRS and the Conseil
S. Shinkai, Supramol. Chem., 1995, 4, 319; (b) L. Bennouna,
J. Vicens and Z. Asfari, J. Inclusion Phenom. Macrocyclic
Chem., 2001, 40, 96; (c) U. V. Trivedi, S. K. Menon and
Y. K. Agrawal, React. Funct. Polym., 2002, 50, 205; (d)
B. Boulet, C. Bouvier-Capely, C. Cossonnet and G. Cote,
Solvent Extr. Ion Exch., 2006, 24, 319; (e) L. Poriel,
B. Boulet, C. Cossonnet and C. Bouvier-Capely, Radiat.
Prot. Dosim., 2007, 127, 273; (f) B. Boulet, C. Bouvier-
Capely, G. Cote, L. Poriel and C. Cossonnet, J. Alloys
´
Regional de Bourgogne for
a
PhD fellowship (2009-
9201AAO037S03746).
Notes and references
1 A. Lisovskii and M. S. Eisen, Top. Organomet. Chem., 2005, 10, 63.
2 (a) I. Verel, G. W. M. Visser, R. Boellaard, M. Stigter-van
Walsum, G. B. Snow and G. A. M. S. van Dongen, J. Nucl.
Med., 2003, 44, 1271; (b) T. J. Wadas, E. H. Wong,
G. R. Weisman and C. J. Anderson, Chem. Rev., 2010, 110,
2858; (c) L. R. Perk, M. J. W. D. Vosjan, G. W. M. Visser,
´
Compd., 2007, 444–445, 526; (g) A. Leydier, D. Lecercle,
´
S. Pellet-Rostaing, A. Favre-Reguillon, F. Taran and
M. Lemaire, Tetrahedron, 2008, 64, 11319.
M. Budde, P. Jurek, G. E. Kiefer and G. A. M. S. van 12 (a) T. Nagasaki, S. Shinkai and T. Matsuda, J. Chem. Soc.,
Dongen, Eur. J. Nucl. Med. Mol. Imaging, 2010, 37, 250; (d)
Y. Zhang, H. Hong and W. Cai, Curr. Radiopharm., 2011, 4,
131; (e) T. K. Nayak, K. Garmestani, D. E. Milenic and
M. W. Brechbiel, J. Nucl. Med., 2012, 53, 113; (f) M. A. Deri,
B. M. Zeglis, L. C. Francesconi and J. S. Lewis, Nucl. Med.
Bio., 2013, 40, 3; (g) G. Fischer, U. Seibold,
Perkin Trans. 1, 1990, 2617; (b) T. Nagasaki and S. Shinkai,
J. Chem. Soc., Perkin Trans. 2, 1991, 1063; (c) L. Dasaradhi,
P. C. Stark, V. J. Huber, P. H. Smith, G. D. Jarvinen and
A. S. Gopalan, J. Chem. Soc., Perkin Trans. 2, 1997, 1187; (d)
T. N. Lambert, L. Dasaradhi, V. J. Huber and A. S. Gopalan,
J. Org. Chem., 1999, 64, 6097.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 22743–22754 | 22753

View Article Online
RSC Advances
Paper
13 L. Dong and M. J. Miller, J. Org. Chem., 2002, 67, 4759.
14 F. Arnaud-Neu, G. Barrett, S. Cremin, M. Deasy, G. Ferguson,
21 R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr.,
Theor. Gen. Crystallogr., 1976, 32, 751.
´
S. J. Harris, A. J. Lough, L. Guerra, M. A. McKervey, 22 J. Taesch, V. Heitz, F. Topic and K. Rissanen, Chem.
M. J. Schwing-Weill and P. Schwinte, J. Chem. Soc., Perkin
Trans. 2, 1992, 1119.
15 Y.-M. Lin and M. J. Miller, J. Org. Chem., 1999, 64, 7451.
16 M. R. Johnson, US Pat. Appl., 2006 004 0954, 2006.
17 L. J. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
Commun., 2012, 48, 5118.
23 (a) A. Caselli, L. Giannini, E. Solari, C. Floriani, N. Re,
A. Chiesi-Villa and C. Rizzoli, Organometallics, 1997, 16,
5457; (b) D.-Q. Yuan, W.-X. Zhu, M.-Q. Xu and Q.-L. Guo,
J. Coord. Chem., 2004, 57, 1243.
18 (a) M. A. McKervey, E. M. Seward, G. Ferguson, B. Ruhl and 24 D. L. Kepert, in Inorganic Stereochemistry, Springer Verlag,
S. J. Harris, J. Chem. Soc., Chem. Commun., 1985, 388; (b) Heidelberg, 1982.
G. Calestani, F. Ugozzoli, A. Arduini, E. Ghidini and 25 J. M. Meyer, D. Hohnadel and F. Halle, J. Gen. Microbiol.,
R. Ungaro, J. Chem. Soc., Chem. Commun., 1987, 344. 1989, 135, 1479.
19 D. Cremer and J. A. Pople, J. Am. Chem. Soc., 1975, 97, 1354. 26 Z. Otwinowski and W. Minor, Methods Enzymol., 1997, 276,
20 J. Scheerder, R. H. Vreekamp, J. F. J. Engbersen, 307.
W. Verboom, J. P. M. van Duynhoven and D. N. Reinhoudt, 27 G. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr.,
´
J. Org. Chem., 1996, 61, 3476.
2008, 64, 112.
22754 | RSC Adv., 2014, 4, 22743–22754
This journal is © The Royal Society of Chemistry 2014