Assembly of a coordination cage with four aromatic channel receptors on the outside

Article abstract of DOI:10.1039/b204260f

Silver salts and triphosphine ligands with biphenyl substituents assemble to give coordination cages with four external aromatic channel receptors in a pseudo-tetrahedral arrangement.

Full text of DOI:10.1039/b204260f

Assembly of a coordination cage with four aromatic channel receptors on
the outside
Philip W. Miller, Mark Nieuwenhuyzen, Xingling Xu and Stuart L. James*
School of Chemistry, The Queen’s University of Belfast, David Keir Building, Stranmillis Road, Belfast,
Northern Ireland, UK BT9 5AG. E-mail: s.james@qub.ac.uk
Received (in Cambridge, UK) 7th May 2002, Accepted 24th July 2002
First published as an Advance Article on the web 8th August 2002
Silver salts and triphosphine ligands with biphenyl sub-
stituents assemble to give coordination cages with four
external aromatic channel receptors in a pseudo-tetrahedral
arrangement.
triphos cages.3a,b The nitrate complex 1a was chosen for
inclusion studies since its non-bulky anions should have left the
channel space free for guests. When CHCl3–CH3CN solutions
were layered with benzene, crystals were obtained for which the
molecular structure, determined by single crystal X-ray diffrac-
tion,† is shown in Fig. 1. The cage core, similar to that of
[Ag6(triphos)4(triflate)4]2+,3a was confirmed, with all bond
lengths and angles within normal ranges. The Ag+ ions define a
distorted octahedron whose faces are alternately capped by
nitrates and back-flipped or ‘endo-methyl’3a L.
The twenty-four biphenyl groups form the expected four
aromatic channels, each of length ca. 9.5 Å, in a pseudo-
tetrahedral arrangement, and three of the channels were found to
contain single benzene guests. The structure has crystallo-
graphically imposed C3 symmetry about an axis though the
centre of the vacant channel, making the filled channels
symmetry-equivalent. Interestingly, the vacant channel has a
distinctly different ‘closed’ conformation compared to the other
three, with three of its six biphenyl groups pointing inwards and
resulting in a (shortest) cross-channel C…C distance of ca. 4.2
Å, compared to ca. 8.6 Å in the filled channels (Fig. 2). The
atomic displacement parameters indicate that there is some
disorder associated with the biphenyls and the guests, therefore
detailed comments on specific host–guest interactions are not
possible, although C…C distances appear to be in the expected
range of 3.5–4.0 Å. The nitrate anion of the unfilled channel sits
on the crystallographic C3 axis and weakly bridges between
three silver ions (O…Ag distances = 2.62 Å). The nitrates in
the filled channels are not actually tripodal but instead bond to
The last ten years have seen great advances in the synthesis of
nanoscale multimetallic ring and cage coordination complexes.1
These could now be used as scaffolds on which to build new
large structures, able, for example, to act as new types of
receptors. An interesting possibility is the preparation of single
species that contain several receptors, since they could show
cooperative or anticooperative guest inclusion. Also, whilst
some attention has been given to guest inclusion within
metallocages,1 relatively little study has been made of assem-
blies with binding sites on their exteriors.2 This sort of
functionalisation could allow such species to be used as
construction units in still larger discrete assemblies or extended
networks, by using guests as linkers.
We are investigating multidentate phosphines as ligands for
generating metal cages and polymers.3 For example, we
reported the triphosphine–silver cage [Ag6(triphos)4(OTf)4]2+
(triphos = CH3C(CH2PPh2)3, OTf = triflate, CF3SO32),3a an
interesting feature of which is its arrangement of twenty-four
phenyl groups in four short (ca. 5 Å) C3-symmetry aromatic
channels on the exterior. Most of the short channel space is
occupied by triflate anions, coordinated to the silver ions at the
core, and so no inclusion behaviour in the channels would be
expected. However, the para-hydrogen atoms of the phenyl
groups all point directly away from the cage core, so that para
substituents could extend these channels into effective re-
ceptors, particularly in combination with smaller anions. To test
this idea, the 4-biphenyl-substituted triphosphine L† was
synthesised (Scheme 1).
only two silver ions each (O12A…Ag1B
= 2.43 and
O13A…Ag2B = 2.84 Å) with the distance to the third silver
centre being too far for a significant contact (O11A…Ag1A =
3.77 Å).
Addition of 4 equivalents of L to six equivalents of AgX gave
the hexasilver cages [Ag6L4X4]2+ (X = NO3, 1a; CF3SO3 1b;
or 4-CH3C6H4SO3 1c) instantly and quantitatively. 31P NMR
Interestingly, when a similar solution of 1a was layered with
pentafluorobenzene, in anticipation of stronger binding of an
data (d/ppm and 1J
Ag/Hz: 1a 26.3, 555; 1b 26.2, 586; 1c
31P–109
28.9, 570) were similar to those of the phenyl-substituted
Fig. 1 Crystal structure of 1a·(C6H6)3: benzene-filled channels are in grey,
unfilled channel in light brown. P = purple, O = red, N = blue, Ag =
brown. Hydrogen atoms omitted for clarity, space-filling spheres set at 0.70
van der Waals radius. The biphenyl groups show evidence of disorder and
have not all been constrained to planarity.
Scheme 1
2008
CHEM. COMMUN., 2002, 2008–2009
This journal is © The Royal Society of Chemistry 2002

Crystal structure determinations: data for both structures were collected
on a Bruker SMART diffractometer using the SAINT-NT8 software with
graphite monochromated Mo-Ka radiation using phi/omega scans. A
crystal was mounted on to the diffractometer at low temperature under
nitrogen at ca. 120 K. Crystal stabilities were monitored and there were no
significant variations ( < ±2%). Lorentz and polarisation corrections were
applied.
The structures were solved using direct methods and refined with the
SHELXTL program package.9 The absolute configurations of 1a·(C6H6)3
and 1a·(CHCl3)(H2O) were assigned using the Flack parameter10 (0.03(11)
and 0.05(10) for 1a·(C6H6)3 and 1a·(CHCl3)2(H2O) respectively). The
complexes exhibit higher than expected atomic displacement parameters
this is indicative of disorder which we were unable to model because the
crystals exhibited weak diffraction. However we were able to establish the
atomic connectivity and the presence of solvent within the ‘channels’. The
Fig. 2 Comparison of the conformations of filled (grey, left) and unfilled
(light brown, right) channels in 1a·(C6H6)3. Hydrogen atoms are omitted for
clarity.
2
2
function minimised was S[w(¡Fo¡ 2 ¡Fc¡ )] with reflection weights w21
=
2
2
2
[s2¡Fo¡ + (g1P)2 + (g2P)] where P = [max. ¡Fo¡ + 2¡Fc¡ ]/3.
Crystal data for C308H252O18N6P12Ag6·3(C6H6) (1a·(C6H6)3): M
=
electron deficient aromatic,4 crystals of another trisolvate were
obtained, but single crystal X-ray diffraction† revealed that
pentafluorobenzene had not in fact been incorporated. The
disordered guests could only be modelled as a mixture of
chloroform and water. This surprising result may have been due
to the slightly greater size of pentafluorobenzene, or unavoida-
ble edge-to-face interactions with the biphenyls, which could be
unfavourable.5 It suggests, interestingly, that 1a can discrim-
inate between differently substituted aromatics, similarly to p-
tert-butylcalix[4]arenes.6
Several receptors connected together, as in 1a–c, can
potentially exhibit cooperative or anticooperative guest inclu-
sion. In 1a–c, the receptors are mechanically coupled to each
other via the cage core, since the biphenyls of a given channel
are connected to those in adjacent channels via shared P atoms.
It is therefore possible that the closing up of the fourth channel
is caused by guest inclusion in the other three. Alternatively, it
may be that the isolation of trisolvates is simply due to their
favourable crystal packing. It would be necessary to measure
stepwise association constants in the solution state to determine
which is the case. However, for aromatic guests such as
benzene, toluene and phenol we have not observed significant
changes in 1H NMR chemical shifts on titration with 1a,
possibly due to competition with the organic solvents necessary
to dissolve the cage. Use of water as a solvent, with a water-
soluble version of the cage, might give measurable solution
state association, as found for water-soluble calixarenes.7
In summary, a unique arrangement of four connected
receptors has been assembled on the exterior of a coordination
cage by rational ligand design. Future work will concentrate on
detailed studies of guest binding and the connecting together of
cages with bridging guests such as oligoaromatics, toward
higher-level assemblies.
5578.34, cubic, space group P213, a = 30.564(5) Å, U = 28553(8) Å3, Z
= 4, m = 0.534 mm21. A total of 42984 reflections were measured for the
angle range 4 < 2q < 40 and 8890 independent reflections were used in the
refinement. The final parameters were wR2 = 0.3419 and R1 = 0.1159 [I
> 2sI]. CCDC 178938.
Crystal data for C308H252O18N6P12Ag6·3{2/3(H2O) + 1/3(CHCl3)} (1a·
(CHCl3)(H2O)): M = 5607.51, cubic, space group P213, a = 30.7586(10)
Å, U = 29100.4(16) Å3, Z = 4, m = 0.553 mm21. A total of 74941
reflections were measured for the angle range 4 < 2q < 40 and 9086
independent reflections were used in the refinement. The final parameters
were wR2 = 0.3296 and R1 = 0.1141 [I > 2sI]. CCDC 178939. See http://
www.rsc.org/suppdata/cc/b2/b204260f/ for crystallographic files in .cif or
other electronic format. For both structures, the silver and phosphorus atoms
have been refined anisotropically, all other non-hydrogen atoms were
refined isotropically. The biphenyl substituents were restrained to have a
chemically feasible geometry and the thermal parameters have been
modelled using SIMU constraints.
1 D. L. Caulder and K. N. Raymond, Acc. Chem. Res., 1999, 32, 975–982;
P. J. Stang, Chem.–Eur. J., 1998, 4, 19–27; M. Fujita, Chem. Soc. Rev.,
1998, 27, 417–425; C. J. Jones, Chem. Soc. Rev., 1998, 27, 289–299;
Transition Metals in Supramolecular Chemistry R. W. Saalfrank and B.
Demleitner, in Perspectives in Supramolecular Chemistry, ed. J. P.
Sauvage, Wiley-VCH, Weinheim, 1999, vol. 5, pp. 1–51; S. Leininger,
B. Olenyuk and P. J. Stang, Chem. Rev., 2000, 100, 853–908.
2 Y. L. Cho, H. Uh, S.-Y. Chang, H.-Y. Chang, M.-G. Choi, I. Shin and
K.-S. Jeong, J. Am. Chem. Soc., 2001, 123, 1258–1259; H. Abourahma,
A. W. Coleman, B. Moulton, B. Rather, P. Shahgaldian and M. J.
Zaworotko, Chem. Commun., 2001, 2380.
3 (a) S. L. James, D. M. P. Mingos, A. J. P. White and D. J. Williams,
Chem. Commun., 1998, 2323; (b) X. Xu, E. J. MacLean, S. J. Teat, M.
Nieuwenhuyzen, M. Chambers and S. L. James, Chem. Commun., 2002,
78; (c) E. Lozano, M. Niewenhuyzen and S. L. James, Chem.–Eur. J.,
2001, 12, 1644; (d) X. Xu, M. Niewenhuyzen and S. L. James, Angew.
Chem., Int. Ed., 2002, 41, 764.
4 C. R. Patrick and G. S. Prosser, Nature, 1960, 187, 1021; T. Dahl, Acta
Chem. Scand., Ser. A, 1988, 42, 1–7; C. A. Hunter, X.-J. Lu, G. M.
Kapteijn and G. van Koten, J. Chem. Soc., Faraday Trans., 1995, 91,
2009–2015.
We are grateful to EPSRC, the McClay Trust and the
Northern Ireland Department of Education and Learning for
funding, and to the referees for valuable comments.
5 N. Hayashi, T. Mori and K. Matsumoto, Chem. Commun, 1998, 1905;
I. S. Neretin, K. A. Lyssenko, M. Y. Antipin, Y. L. Slovokhotov, O. V.
Boltalina, P. A. Troshin, A. Y. Lukonin, L. N. Sidorov and R. Taylor,
Angew. Chem., Int. Ed., 2000, 39, 3273; attraction between C6F6 and
electron donors: I. Alkorta, I. Rozas and J. Elguero, J. Org. Chem.,
1997, 62, 1997.
6 A. Pochini and R. Ungaro, in Comprehensive Supramolecular Chem-
istry, ed. F. Vögtle, Pergamon Press, Oxford, 1996, vol. 2, p. 103.
7 S. Shinkai, Tetrahedron, 1993, 49, 8933 and references therein.
8 SAINT-NT, program for data collection and data reduction, Bruker-
AXS, Madison, WI, 1998.
Notes and references
†
Characterising data for 1,1,1-tris{bis(4-biphenyl)phosphinomethyl-
}ethane L: 1H NMR (300 MHz, CDCl3): d 7.25–7.46 (m, 54 Haromatic), 2.47
(s, 6 H, CH2), 1.11 (s, 3 H, CH3): 31P{1H} NMR (121 MHz, CDCl3): d
227.12 (s); FAB MS: m/z (%): 1081 (12) [M+], 307 (100); microanalysis:
calc. C 85.53, H 5.87; found C 84.25, H 5.60%.
Synthesis of cage complexes 1a–c: typically, a solution of L (100 mg,
0.092 mmol) in CDCl3 or CHCl3 (4 ml) was added to a solution of the
appropriate silver salt (0.138 mmol) in CH3CN (1 ml) and 31P NMR spectra
were obtained. For 1a, the solution was layered with either benzene or
pentafluorobenzene to obtain 1a·(C6H6)3 or 1a·(CHCl3)2(H2O), respec-
tively.
9 G. M. Sheldrick, SHELXTL Version 5.0, A System for Structure
Solution and Refinement, Bruker-AXS, Madison WI, 1998.
10 H. D. Flack, Acta Crystallogr., Sect. A, 1983, 39, 876.
CHEM. COMMUN., 2002, 2008–2009
2009