Thiacalix[4] tube: Synthesis, X-ray crystal structure and preliminary binding studies

Article abstract of DOI:10.1039/b106094p

The calix[4] tube family of ionophores has been extended to encompass the slightly larger thiacalix[4] tube. ¹H NMR and crystallographic studies show the ligand to adopt a flexible C^2v flattened cone arrangement. However, in contrast

Full text of DOI:10.1039/b106094p

View Article Online / Journal Homepage / Table of Contents for this issue
Thiacalix[4]tube: synthesis, X-ray crystal structure and preliminary
binding studies¤
Susan E. Matthews,”a Vitor Felix,°b Michael G. B. Drewb and Paul D. Beer*a
a Inorganic Chemistry L aboratory, Department of Chemistry, University of Oxford, South Parks
Road, Oxford, UK OX1 3QR. E-mail: Paul.Beer=chem.ox.ac.uk; Fax: ]44 1865 272 690
b Department of Chemistry, University of Reading, W hiteknights, Reading, Berkshire,
UK RG6 6AD
L e t t e r
Received (in Montpellier, France) 10th July 2001, Accepted 6th September 2001
First published as an Advance Article on the web 17th October 2001
The calix[4]tube family of ionophores has been extended to
encompass the slightly larger thiacalix[4]tube. 1H NMR and
opportunity of further functionalisation at the narrow rim.12
Careful LiAlH reduction of 3 in ether allows the isolation, in
4
crystallographic studies show the ligand to adopt a Ñexible C
48% yield, of the tetraalcohol 4, which retains the cone con-
2v
Ñattened cone arrangement. However, in contrast to the parent
formation. This compound can be e†ectively converted (62%
yield) to the tetratosylate 5 on treatment with tosylchloride
and triethylamine. This useful precursor can be further treated
with t-butyl thiacalix[4]arene (1) in the presence of potassium
carbonate in reÑuxing xylene to yield the desired symmetrical
thiacalix[4]tube (6) in 10% yield. (Scheme 1)
calix[4]tube, preliminary binding studies show no selectivity for
potassium, in agreement with the absence of a template e†ect in
tube formation and molecular modelling studies.
Since the Ðrst reported synthesis of thiacalix[4]arene (1) in
1997,1,2 considerable interest has been focused on both its
derivatisation3,4 and metal complexation5,6 properties.
Thiacalix[4]arene ideally complements the existing calixarene
family,7 providing a macrocycle with a slightly larger annulus
than calix[4]arene combined with four sulfur donor atoms,
which have been shown to co-ordinate transition metal
cations.3,8,9
We have recently reported a new class of cryptand-type ion-
ophores, the calix[4]tube 2, based on a bis(calix[4]arene) scaf-
fold that displays exceptional selectivity for potassium over all
group 1 metal cations.10 Extending this system to incorporate
thiacalix[4]arene units would provide a cavity of slightly
larger dimensions, which may lead to a new ionophore with
contrasting metal co-ordination properties.
The 1H and 13C NMR spectra (Fig. 1) of the thia-
calix[4]tube 6 indicate, by a doubling of all resonances (a
and a@, etc.), that the two calix[4]arene units are Ðxed in a C
2v
Ñattened cone arrangement and are linked by two indepen-
dent types of ethylene linkages in a similar manner to the
uncomplexed calix[4]tube 2.10 However, all the resonances
are broad at room temperature and an ideal spectrum is only
observed at 0 ¡C. The temperature dependence of the 1H
NMR signals is shown in Fig. 2. SigniÐcant broadening occurs
over the temperature range [20 to 55 ¡C with a gradual ten-
dency towards coalescence and a C symmetrical structure,
4v
that results in the proton NMR spectrum simplifying from 6
to 3 resonances. Of particular interest is the 30È50 ¡C region
where further resonances appear in the spectrum, this can be
postulated to relate to a second favourable arrangement of the
ethylene region that can be attained at higher temperatures.
This Ñexibility is aptly demonstrated by the rate of exchange
Akdas et al.11 have developed selective conditions for the
preparation of the lower rim substituted thiacalix[4]arene
tetraacetate (3) in the cone conformation, which o†ers the
between the two possible
calix[4]arene subunits, through the intermediate C cone;
C
arrangements of the
2v
4v
k
calculated from EXSY 1H NMR data13 for the
exch
thiacalix[4]tube 6 is 9.25 s~1 at 273 K in comparison to 0.9
s~1 at 328 K for the potassium selective calix[4]tube 2.
X-Ray quality single crystals of thiacalix[4]tube were
grown by slow di†usion of diisopropyl ether into a chloro-
form solution. The structure in the solid state is shown in Fig.
3 and is consistent with the low temperature NMR studies in
solution. The conformation of the cage can be quantiÐed by
the ÈOÈCH ÈCH ÈO torsion angles and the angles made by
2
2
the phenyl rings with the plane of the four linking sulfur
atoms. Two of the torsion angles are 0¡, consistent with the
2/m symmetry of the molecule. The other two are ^165.2¡.
The angles made by the phenyl rings with the 4S plane are
88.1 and 42.7¡.
This conformation is very similar to that observed in the
t-octyl calix[4]tube14 in which the torsion angles are [0.5,
161.6, 0.5 and [154.9¡. Here the cage contains a crystallo-
graphic centre of symmetry but again the two carbon atoms
that make up the trans torsion angles are disordered over two
sites. In contrast, the t-butyl calix[4]tube 2 has torsion angles
of 161.2, [47.8, [161.2 and 47.8¡; the tube again contains a
crystallographic centre of symmetry but without disorder.
¤ Electronic supplementary information (ESI) available: full experi-
mental details, including synthesis and characterisation of 4 and 5,
and the crystallographic study of 6. See http://www.rsc.org/suppdata/
nj/b1/b106094p/
” Current address: Department of Chemistry, Trinity College Dublin,
Dublin 2, Ireland.
° On sabbatical leave from the Departamento de Quimica, Uni-
versidade de Aveiro, P-3810-193 Aveiro, Portugal.
DOI: 10.1039/b106094p
New J. Chem., 2001, 25, 1355È1358
1355
This journal is ( The Royal Society of Chemistry and the Centre National de la Recherche ScientiÐque 2001

View Article Online
Scheme 1 Synthesis of thiacalix[4]tube (6). Reagents and conditions: (i) BrCH CO Et, Na CO , acetone, reÑux, 71%; (ii) LiAlH , Et O, 48%;
2
2
2
3
4
2
(iii) tosyl chloride, Et N, 62%; (iv) 1, K CO , xylene, reÑux, 4 days, 10%.
3
2
3
Preliminary alkali metal cation binding studies were under-
taken using electrospray ionisation mass spectrometry (ESI-
MS). This method has recently attracted much interest for the
evaluation of hostÈguest complexation and alkali metal cation
selectivity with calixcrown systems.15,16 The thiacalix[4]tube
6 was pre-treated with a mixture of alkali metal iodides in
chloroformÈmethanol (4 : 1) to give a host : guest ratio of
1 : 20. Disappointingly, Fig. 4 shows that signiÐcant complex
[6 ] M`] peaks were observed for both sodium and pot-
assium, together with smaller rubidium and caesium peaks,
which suggests that the thiacalix[4]tube exhibits poor selec-
tivity for these cationic guest species. This observation is in
stark contrast to 2, which displays remarkable selectivity for
potassium cations.
Qualitative 1H NMR complexation experiments with 6 and
potassium cations in CDCl ÈCD OD (4 : 1) also conÐrmed
3
3
the thiacalix[4]tube to be a poor ionophore for this metal
cation; only very small shifts (*d O 0.05) were observed in the
spectra. These preliminary co-ordination chemistry results
may account for the poor yield (10%) of the thiacalix[4]tube
cage-forming step (Scheme 1), suggesting that formation of the
Fig. 1 NMR spectrum for thiacalix[4]tube (6) in CDCl at 0 ¡C.
3
Fig. 2 A section of the 1H NMR spectrum for thiacalix[4]tube (6) in
CDCl over a range of temperatures, showing broadening of the orig-
inal signals for the ethylene linkage (L) and transient appearance of
signals (…), indicating a second species at elevated temperature.
3
Fig. 3 The structure of 6 with ellipsoids at 15%. One ordered mol-
ecule is shown.
1356
New J. Chem., 2001, 25, 1355È1358

View Article Online
Fig. 4 Electrospray mass spectrum of thiacalix[4]tube (6) after pre-complexation with 20 equiv. of sodium, potassium, rubidium and caesium as
the iodides in chloroformÈmethanol (4 : 1). These competition selectivity experiments were complicated by the in situ oxidation of the
thiacalix[4]tube under the electrospray ionisation MS conditions.
thiacalix[4]tube is not potassium cation template driven as is
the case with the parent calix[4]tube 2.
cipitated from chloroformÈmethanol. The white solid was re-
crystallised from chloroformÈacetone to yield the
Molecular modelling studies have been carried out in an
e†ort to rationalise the above Ðndings. Calculations were
undertaken to compare the pathway of a potassium cation
through the thiacalix[4]tube 6 with that previously estab-
lished for calix[4]tube 2. In the case of calix[4]tube 2, we
showed via molecular dynamics calculations that K` entered
the cage along the central axis of the calixarene, then located
at an intermediate position close to the phenyl rings before
proceeding to the centre of the cage, triggering the necessary
conformational change of the central OÈCH ÈCH ÈO torsion
thiacalix[4]tube 6 (10%). 1H NMR (500 MHz [D]CHCl
3
18 ¡C): d 0.78 [s, 36H; (CH ) ], 1.26 [s, 36H; (CH )@ ], 4.67 (s,
3 3
3 3
8H; OCH ), 5.33 (s, 8H; OCH@ ), 6.90 (s, 8H; ArH), 7.69 (s,
2
2
8H; ArH@); 13C NMR (75 MHz, [D]CHCl , 18 ¡C): d 30.76
3
[(CH ) ], 31.39 [(CH )@ ], 33.81 [C(CH ) ], 34.29 [C(CH )@ ],
3 3
3 3
2
3 3
3 3
74.06 (OCH ] OCH@ ), 128.25 (ArH ] ArH@), 130.97
2
(ArH ] ArH@), 133.04 (ArCH ), 134.55 (ArCH@ ), 145.86
2
2
(ArCH ] ArCH@), 158.34 (ArO), 161.36 (ArO@); MS (MALDI):
m/z 1541.52 [M [ 4H]`.
2
2
angles from tgtg to gggg. This latter conformation then
accommodates the K` ion in an approximate cubic eightfold
co-ordination environment.
Crystal data. 6 É 4H O É 3CH Cl ,
C
H
Cl O S ,
2
2
2
91 118 16 12 8
M \ 1937.04, monoclinic, space group C2/m, a \ 20.72(2),
b \ 19.67(3), c \ 12.796(15) A, b \ 101.48(1)¡, U \ 5111(11)
Our molecular dynamics calculations show that the mecha-
nism for insertion of the K` ion in the thiacalix[4]tube 6 is
equivalent to that found for the parent tube. However the size
of the metal co-ordination sphere is signiÐcantly larger in that
the optimum MÈO distance (obtained by molecular mecha-
nics calculations using our published method17) is 2.96 A,
compared to a calculated distance of 2.82 A and an experi-
mental distance in the crystal structure of 2.759(6) to 2.809(6)
A for calix[4]tube 2. Indeed, a search of the Cambridge Crys-
tallographic Databases for KÈO (crown ether) distances
showed a mean KÈO distance of 2.85 A for 1689 observations
(17 outliers were ignored). This mismatch in size may be the
reason that the K` ion does not remain in the cryptand
region but readily withdraws and may also account for the
poor binding ability of the thiacalix[4]tube 6.
AŽ
3, Z \ 2.
CCDC reference number 170661. See http://www.rsc.org/
suppdata/nj/b1/b106094p for crystallographic data in CIF or
other electronic format.
Acknowledgements
This work was supported by an EPSRC Postdoctoral Fellow-
ship (S. E. M.). The University of Reading and the EPSRC are
gratefully acknowledged for funding towards the crystallo-
graphic image plate system. The EPSRC Mass Spectrometry
Service (University College, Swansea) is gratefully acknow-
ledged for the provision of MALDI mass spectra.
In summary, synthetic procedures for the preparation of
calix[4]tubes have been successfully extended to the prep-
aration of a thiacalix[4]tube. This adopts the same unusual
Ñattened cone arrangement in both solution and crystalline
states although exchange between the two extremes is fairly
rapid at room temperature. Such Ñuxionality, combined with
the absence of an apparent templation e†ect and molecular
modelling results, tend to account for the lack of selective pot-
assium binding of this new ionophore.
Notes and references
1
H. Kumagi, M. Hasegawa, S. Miyanari, Y. Sugawa, Y. Sato, T.
Hori, S. Ueda, H. Kamiyama and S. Miyano, T etrahedron L ett.,
1997, 38, 3971.
2
3
4
5
T. Sone, Y. Ohba, K. Moriya, H. Kumada and K. Ito, T etra-
hedron, 1997, 53, 10689.
N. Iki, N. Morohashi, F. Narumi, T. Fujimoto, T. Suzuki, S.
Miyano and C. Kabuto, T etrahedron L ett., 1999, 40, 7337.
H. Katagiri, N. Iki, T. Hattori, C. Kabuto and S. Miyano, J. Am.
Chem. Soc., 2001, 123, 779.
Y. Higuchi, M. Narita, T. Niimi, N. Ogawa, F. Hamada, H.
Kumagai, N. Iki, S. Miyano and C. Kabuto, T etrahedron, 2000,
56, 4659.
R. Lamartine, C. Bavoux, F. Vocanson, A. Martin, G. Senlis and
M. Perrin, T etrahedron L ett., 2001, 42, 1021.
V. Bohmer, Angew. Chem., Int. Ed. Engl., 1995, 34, 713.
G. Mislin, E. Graf, M. W. Hosseini, A. Bilyk, A. K. Hall, J. M.
HarrowÐeld, B. W. Skelton and A. H. White, Chem. Commun.,
1999, 373.
Experimental
General details and full characterisation of 4 and 5 are provid-
ed in the ESI, as well as details of the crystallographic study.
6
7
8
Thiacalix[4]tube (6)
Tetrakis[(4-methylphenyl)sulfonyloxyethoxy]-p-tert-butyl-
thiacalix[4]arene (5; 400 mg, 0.27 mmol) and p-tert-butylthia-
calix[4]arene (1; 192 mg, 0.27 mmol) were heated at reÑux in
xylene (20 ml) in the presence of K CO (183 mg, 1.33 mmol)
9
A. Bilyk, A. K. Hall, J. M. HarrowÐeld, M. W. Hosseini, G.
Mislin, B. W. Skelton, C. Taylor and A. H. White, Eur. J. Inorg.
Chem., 2000, 5, 823.
10 P. Schmitt, P. D. Beer, M. G. B. Drew and P. D. Sheen, Angew.
Chem., Int. Ed. Engl., 1997, 36, 1840.
11 H. Akdas, G. Mislin, E. Graf, M. W. Hosseini, A. D. Cian and J.
Fischer, T etrahedron L ett., 1999, 40, 2113.
2
3
for 4 days. The solvent was evaporated, the residue re-
dissolved in chloroform and then washed with water. After
drying the solvent was evaporated and the crude material pre-
New J. Chem., 2001, 25, 1355È1358
1357

View Article Online
12 Similar conditions have also been reported: N. Iki, F. Narumi, T.
Fujimoto, N. Morohashi and S. Miyano, J. Chem. Soc., Perkin
T rans. 2, 1998, 2745.
14 S. E. Matthews, P. Schmitt, V. Felix, M. G. B. Drew and P. D.
Beer, J. Am. Chem. Soc., in press.
15 F. Allain, H. Virelizer, C. Moulin, C. K. Jankowski, J. F. Dozol
and J. C. Tabet, Spectroscopy, 2000, 14, 127.
13 The slow exchange rate was determined from 1D-EXSY (p/2 [ t
1
[ p/2 [ t [ p/2-obs) experiments. Rate constants were then
m
16 M. T. Blanda, D. B. Farmer, J. S. Brodbelt and B. J. Goolsby, J.
Am. Chem. Soc., 2000, 122, 1486.
calculated using the program CIFIT with T1 relaxation rates
estimated from inversion recovery experiments (A. D. Bain and J.
A. Cramer, J. Magn. Reson., Ser. A, 1996, 118, 21.
17 J. Costa, R. Delgado, M. G. B. Drew, V. Felix, R. T. Henriques
and J. C. Waerenborgh, J. Chem. Soc., Dalton T rans., 1999, 3253.
1358
New J. Chem., 2001, 25, 1355È1358