Synthesis of "two-story" calix[6]aza-cryptands

Article abstract of DOI:10.1021/ol202380q

The first four members of a new family of C_3v-symmetrical "two-story" calix[6]aza-cryptands have been synthesized. These large funnel shaped aza-ligands are formed through introduction of three aromatic arms as spacers onto the small rim of a c

Full text of DOI:10.1021/ol202380q

ORGANIC
LETTERS
2011
Vol. 13, No. 20
5660–5663
Synthesis of “Two-Story”
Calix[6]aza-Cryptands
Xianshun Zeng,†,‡ Claudia Bornholdt,† Diana Over,† and Olivia Reinaud*,†
Laboratoire de Chimie et de Biochimie pharmacologiques et toxicologiques,
ꢀ
ꢀ
ꢁ
Universite Paris Descartes, Sorbonne Paris Cite, CNRS UMR 860, 45, rue des Saints Peres,
75006 Paris, France, and Key Laboratory of Display Materials (Ministry of Education),
Tianjin University of Technology, Tianjin 300384, China
Olivia.Reinaud@parisdescartes.fr
Received September 2, 2011
ABSTRACT
The first four members of a new family of C3v-symmetrical “two-story” calix[6]aza-cryptands have been synthesized. These large funnel shaped
aza-ligands are formed through introduction of three aromatic arms as spacers onto the small rim of a calix[6]arene and subsequently capped with
the tripodal aza caps tacn [1,3,5-triazacyclononane] or tren [tris(aminoethyl)amine]. A key feature for an efficient final 1:1 macrocyclization
appears to be an adequate geometrical fit between the extended calixarene scaffold and the aza caps.
Calix[6]arenes are useful building blocks for obtaining
molecular receptors.1 A prerequisite, however, for their use
as a host is their rigidification into a cone conformation.
Indeed these large macrocycles are highly flexible and
easily undergo flipping of the aromatic walls through the
annulus. We have developed strategies to close the small
rim and constrain the calixarene core into a cone. The first
one is based on coordination chemistry and led to the so-
called “funnel” complexes.2,3 The second one makes use of
covalent linkages to yield calix[6]aza-cryptands.4 These
compounds, as illustrated in Figure 1, provide a hydro-
phobic cavity defined by the six aromatic units that is open
at the large rim for guest hosting. Once coordinated in the
aza cap, a metal ion has not only a distinct first coordina-
tion sphere but also a second coordination sphere well-
defined by the oxygen-rich calix small rim separated by a
two-atom linkage from the aza donors. Wanting to in-
crease the size of the hydrophobic funnel and vary the
second coordination core around the metal ions, we
thought of inserting a spacer between the aza cap and the
macrocycle. From a synthetic point of view, the spacer
must be not only flexible enough to allow the aza cap to
geometrically fit onto the small rim of the cone but also
rigid enough to take advantage of the preorganization
provided by the calixarene scaffold during the capping-
macrocyclization. Indeed, one particularity of the so-called
calix[6]aza-cryptands is that they cannot be purified by
standard chromatographical methods. Therefore, a pre-
requisite for their isolation is a good yield in the final
step. Here, we report the synthesis of the first members of
such “two-story” calix[6]aza-cryptands based on the same
†
ꢀ
ꢀ
Universite Paris Descartes, Sorbonne Paris Cite.
‡ Tianjin University of Technology.
(1) Gutsche, C. D. In Calixarenes: An introduction (Monographs in
Supramolecular Chemistry), 2nd ed.; Stoddart, J. F., Ed.; The Royal Society
of Chemistry: Cambridge, 2008.
(2) (a) Blanchard, S.; Le Clainche, L.; Rager, M.-N.; B. Chansou, B.;
Tuchagues, J.-P.; Duprat, A. F.; Le Mest, Y.; Reinaud, O. Angew.
ꢀ ꢁ
Chem., Int. Ed. 1998, 37, 2732–2735. (b) Seneque, O.; Rager, M.-N.;
Giorgi, M.; Reinaud, O. J. Am. Chem. Soc. 2000, 122, 6183–6189. (c)
ꢀ ꢁ
Seneque, O.; Rondelez, Y.; Le Clainche, L.; Inisan, C.; Rager, M.-N.;
Giorgi, M.; Reinaud, O. Eur. J. Inorg. Chem. 2001, 2597–2604.
ꢁ
ꢀ ꢁ
(3) Perspective: Coquiere, D.; Le Gac, S.; Darbost, U.; Seneque, O.;
Jabin, I.; Reinaud, O. Org. Biomol. Chem. 2009, 7, 2485–2500.
(4) (a) Jabin, I.; Reinaud, O. J. Org. Chem. 2003, 68, 3416–3419. (b)
Darbost, U.; Giorgi, M.; Reinaud, O.; Jabin, I. J. Org. Chem. 2004, 69,
4879–4884. (c) Zeng, X.; Hucher, N.; Reinaud, O.; Jabin, I. J. Org.
ꢁ
Chem. 2004, 69, 6886–6889. (d) Zeng, X.; Coquiere, D.; Alenda, A.;
ꢀ
Garrier, E.; Prange, T.; Li, Y.; Reinaud, O.; Jabin, I. Chem.;Eur. J.
2006, 12, 6393–6402.
r
10.1021/ol202380q
Published on Web 09/30/2011
2011 American Chemical Society

synthetic strategy. The spacers are either meta-xylene or
2,6-dimethyl-pyridine units and the aza-caps are tacn
[1,3,5-triazacyclononane] or tren [tris-(aminoethyl)amine],
well-known ligands in coordination chemistry.
Scheme 1. Synthesis of Calix[6]xyl-tacn 5a, Calix[6]pyr-tacn 5b,
Calix[6]xyl-tren 6a, and Calix[6]pyr-tren 6ba
Figure 1. Left and center: funnel complexes obtained with calix-
[6]aza-ligands with a two-atom spacer; right: a putative complex
based on a “two-story” calix[6]aza-cryptand. The dashed lines
highlight the second coordination sphere of the embedded metal
ion.
a Overall yields from compound 1: 5a: 22%, 5b: 44%, 6a: 16%.
Covalent capping of the calix-macrocycle has been
successfully obtained either through the reaction of a
nucleophilic cap with a calixarene prefunctionalized
with electrophilic arms4b,d,5 or, conversely, through the
condensation of a nucleophilic calix-derivative with
an electrophilic tripodal cap.4a,c,d,6 Here, we chose the
first option, and the multistep syntheses are depicted
in Scheme 1.
The initial step is the introduction of three spacer arms
bearing a function that will be subsequently transformed
into a good electrophile. 1,3,5-Tris-O-methylated calix-
[6]arene 1 was thus first reacted with either methyl
3-(bromomethyl)benzoate or methyl 6-(bromomethyl)pico-
linate in the presence of sodium hydride. The correspond-
ing xyl-derivative 2a and pyr-derivative 2b4d were obtained
in good yields. Although both molecules differ only in the
phenyl vs pyridine groups, differences were observed in
their chemical behavior. Xyl-ester 2a revealed to be very
sensitive to hydrolysis giving rise to carboxylic acid deri-
vatives, whereas pyr-ester 2b was fully stable in the pre-
sence of water.
of pyridyl arms. Indeed, in the b case the methylene link
between the phenoxyl units and the pyr group is very
sensitive to nucleophilic attack. The smoother reducing
reagent NaBH4 was then advantageously used instead.
Transforming the alcohol functions into electrophilic cen-
ters was carried out with PBr3 to give the tris(bromo)
derivatives 4a and 4b.
For all these intermediates, the 1H NMR analyses
indicate that the calixarene macrocycle adopts a flattened
cone conformation as schematized in Scheme 1. Indeed,
the chemical shift of the methoxy protons is situated
around 2.3 ppm, indicating the position of theOMe groups
close to the C3v axis, while the bulkier free spacer arms are
directed toward the outside of the cavity. For all com-
pounds also, the two resonances of the tBu groups and
those corresponding to the aromatic protons of the calix-
arene skeleton display comparable chemical shift differ-
ences (ΔδtBu ≈ 0.5 ppm and ΔδHAr ≈ 0.55 ppm) indicating
the similar alternate in and out positions of these aromatic
units relative to the cavity.
The triester cores were then reduced into alcohols.
Whereas the reaction proceeded smoothly and in high
yield using LiAlH4 for the xyl-derivative 2a, reaction of
triester 2b with this metal hydride salt often led to the loss
The last step was the 1:1 macrocyclization of the bromo-
derivatives 4a,b with the nitrogen caps tacn and tren. All
reactions were performed with a 1:1 calix/tripodal amine
stoichiometry in the presence of sodium carbonate under
(5) Menand, M.; Jabin, I. Org. Lett. 2009, 11, 673–676.
(6) (a) Le Gac, S.; Zeng, X.; Girardot, C.; Jabin, I. J. Org. Chem.
2006, 71, 9233–9236. (b) Le Gac, S.; Jabin, I. Chem.;Eur. J. 2008, 14,
548–557.
reasonably diluted conditions (ca. 2 Â 10À3 mol.LÀ1
)
in either pure DMF (for compounds 5a, 6a, 6b) or a
Org. Lett., Vol. 13, No. 20, 2011
5661

Figure 2. 1H NMR spectra of calix[6]xyl-tacn 5a, calix[6]pyr-tacn 5b, calix[6]xyl-tren 6a, and calix[6]pyr-tren 6b (250 MHz, 300 K,
CDCl3 for 5a, 5b, 6a and 250 MHz, 300 K, CD2Cl2 for 6b). The red squares denote the resonances of 6b. S denotes the solvent
resonances.
DMF/THF mixture (for 5b). After trituration in ether, the
tacn ligands were isolated as pure compounds in 42% (X =
CH, 5a) and 53% (X = N, 5b) yields. The same purifica-
tion procedure afforded the tren ligand 6a (X = CH) as a
pure compound. The yield (31%) obtained for this tren
ligand is a little lower than that for the corresponding tacn
ones, whichmight be due to the higher flexibility of the tren
cap. In the case of the N7 derivative 6b, we have never been
compounds 5a,b and 6a [Δδ = 0.5 to 0.6 ppm, for both sets
of resonances], which shows that the aromatic units adopt
alternate in an out positions relative to the center of the
cavity and evidences a flattened cone conformation.2c,4d
Finally, the high-field shift of the methoxy protons
(around 2.4 ppm) indicate that they remain partially
included in the calix-cone, very similar to the case of
uncapped compounds 2À4. Hence, the overall conforma-
tion adopted by the calixarene scaffold has been minimally
affected by the capping step as schematized in Scheme 1.
This stands in contrast to calix[6]arenes that have been
capped with a small 2-atom spacer (Figure 1, center). For
example, when the small and rigid 1,3,5-triazacyclohexane
unit caps the same calix macrocycle through an ethyl
spacer, the methoxy groups exhibit a proton resonance at
3.75 ppm, indicating that they have been totally expulsed
from the cone.4b Likewise, with the more flexible tren cap
and ethyl spacers, the methoxy δ shift is 3.05 ppm,4a which
also attests to their relative projection away from the
cavity, but to a lesser extent. It therefore seems that the
aromatic spacers between the calixarene moiety and the
aza cap offer sufficient flexibility for the capped calix-
cryptand to keep a similar conformation before and after
the final macrocyclization step.
1
able to isolate one single species. Indeed, the H NMR
spectrum of the isolated compound always indicated the
presence of several species. One of these shows a C3v
symmetry which does correspond to the desired ligand 6b,
as described below.
These new cryptands were characterized by NMR spec-
troscopy. For compounds 5a, 5b, 6a, the 1H NMR signa-
tures displayed in Figure 2 unambiguously attest to a C3v
symmetry. The resonances characteristic of the calix core
have similar shifts, and hence the conformation is compar-
able in all three cases. The methylene groups bridging the
aromatic units are split into two sets of diastereotopic
protons, one signal for the axial and one for the equatorial
position. Contrarily to precursors 2À4, for compounds 5
and 6 this splitting was not affected by an increase of
temperature, in agreement with the inhibition of the coneÀ
cone inversion due to the covalent capping of their small
rims. The pairs of tBu and aromatic HAr resonances of
the calixarene moiety are well separated in the case of
In the case of 5a and 6a, one resonance belonging to the
xyl-aromatic spacer is shifted to low field (δ = 7.8 ppm),
which must be due to the deshielding effect one aromatic
5662
Org. Lett., Vol. 13, No. 20, 2011

arm has on the others and suggests the spatial proximity of
these three arms. Differences are observedfor the signals of
the tacn cap. For calix[6]xyl-tacn 5a, the tacn protons
appear as a single signal, while, in the case of calix[6]pyr-
tacn 5b, the protons of the cap are split. This suggests that
the mobility of the cap depends on the nature of the
aromatic linkers (xylyl or pyridyl).
In the case of 6b, the resonances denoted in Figure 2
attest to the presence of calix[6]pyr-tren 6b as a major
compound. This was fully confirmed by 2D NMR experi-
ments together with mass spectrometry. Although the
resonances in the 2 to 5 ppm area are very similar to its
xyl-analog 6a, the peaks corresponding to the calix-aro-
matic units and their tBu substituents show a smaller
splitting, indicating a more straight conformation, which
is confirmed by the lower-field shifted methoxy groups
at 2.60 ppm.
The fact that the b case did not work as well as the a case
for the tren derived ligand highlights the sensitivity of the
key capping step to small variations of the reactants (calix-
xyl vs calix-pyr and tren cap vs tacn cap). In addition, we
wanted to check the geometrical factors that may play a
role in the efficiency of the macrocylization step. We
therefore synthesized the analogue of compound 4a start-
ing from the para-xylenyl spacer instead of the meta
derivative following the same experimental procedure as
described above. However, all attempts to isolate a capped
compound analogous to 5a were unsuccessful, the capping
step resulting in the formation of products difficult to
analyze, possibly polymers. This shows that an adequate
geometrical fit is one of the key features for the successful
final macrocyclization reaction in spite of the high flex-
ibility of the calix[6]arene macrocycle.
In conclusion, we havedescribed the synthesis of the first
members of a novel family of funnel shaped aza-ligands.
They are based on a calix[6]arene that has been extended at
the small rim by three aromatic arms covalently assembled
to a tripodal poly aza cap. The resulting “two-story” aza-
cryptands maintain a C3v-symmetrical structure in solu-
tion. Part of the success in the synthesis stems from the
choice of the spacer used to connect the aza cap to the
calixarene. Indeed, whereas the 1:1 final condensation
proceeds with relatively good yields with both meta-xyle-
nyl or 2,6-dimethylpyridyl spacers, with the corresponding
para-xylenyl groups, no cryptand could be isolated. Hence,
the meta-substitution pattern seems to be necessary to
correctly orient the electrophilic arms to react three times
with the nucleophilic tripode. This underlines the impor-
tance of a good geometrical fit between the cap and the
three arms grafted on the calix[6]arene scaffold in spite of
its high flexibility and conformational mobility. Interest-
ingly also, these aromatic spacers open a new way for
introducing additional functionalities to the system,
through their nucleus substitution. Exploration of the
coordination properties and hostÀguest behavior of these
“two-story” funnel-like ligands is currently underway.
Acknowledgment. This research was supported by
CNRS and Agence National pour la Recherche [Projects
ANR-08-CESA-020 (Saturnix) and ANR-2010-BLAN-
7141 (Cavity-zyme-Cu)].
Supporting Information Available. Experimental de-
tails and characterization data for new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
Org. Lett., Vol. 13, No. 20, 2011
5663