Selective formation of a polar incomplete coordination cage induced by remote ligand substituents

Article abstract of DOI:10.1039/c2cc17832j

Instead of highly symmetrical T-symmetry cages common in self-assembly, the p-NMe₂-substituted triphosphine CH₃C{CH ₂P(4-C₆H₄NMe₂)₃ gives open, polar C₃ symmetry cages [Ag₆(triphos) ₄X₃]³⁺ which lack one of the expected face-capping anions; despite its subtlety this difference occurs selectively in solution and two examples have been crystallographically characterised.

Full text of DOI:10.1039/c2cc17832j

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 4061–4063
www.rsc.org/chemcomm
COMMUNICATION
Selective formation of a polar incomplete coordination cage induced by
remote ligand substituentsw
Nicola Giri,a William Clegg,b Ross W. Harrington,b Peter N. Horton,c
Michael B. Hursthousea and Stuart L. James*a
Received 14th December 2011, Accepted 28th February 2012
DOI: 10.1039/c2cc17832j
Instead of highly symmetrical T-symmetry cages common
in self-assembly, the p-NMe2-substituted triphosphine
CH3C{CH2P(4-C6H4NMe2)3 gives open, polar C3 symmetry
cages [Ag6(triphos)4X3]3+ which lack one of the expected face-
capping anions; despite its subtlety this difference occurs selectively
in solution and two examples have been crystallographically
characterised.
Self-assembly tends to give rise to non-polar structures.1a For
example, discrete cages formed through labile metal coordination,
which have been studied in solution, adopt connectivities which
are often based around Platonic or Archimedean solids,1b–e or
prisms,1f–i which can be seen as approximations to spheres or
cylinders.2 Instances where polar structures are adopted selec-
tively are therefore of fundamental interest. It has also been
shown that polar coordination assemblies can themselves be
exploited as building blocks to form still larger discrete
‘clusters of clusters’, i.e. a type of 2nd level assembly.1i Here
we describe a system which, as a result of a simple ligand
modification, gives polar assemblies selectively. Specifically,
Fig. 1 High (T) symmetry adamantoid cages [Ag6(L)4X4]2+ and low
(C3) symmetry [Ag6(L-NMe2)4X3]3+ cages missing one face-capping
whereas the unfunctionalised triphosphine CH3C(CH2PPh2)3
(L)3a,b and two of its para-functionalised derivatives (with phenyl3c
or ꢀCH2N(CH2CH2)2NCH33d substituents) have previously been
found to react with silver salts to give the expected type of non-
polar, T-symmetry adamantoid cages [Ag6(triphosphine)4X4]2+
(Fig. 1a), we report here that the para-NMe2-substituted derivative
CH3C{CH2P(4-C6H4NMe2)2}3 (L-NMe2) gives ‘incomplete’
C3 symmetry cages [Ag6(L-NMe2)4X3]3+ which lack one of the
expected face-capping anions (Fig. 1b). Although this structural
difference is subtle it occurs with remarkable selectivity, as shown
by solution-state 31P NMR for a range of anions. Two example
structures have also been crystallographically characterised.
anion (X = tosylate).
The new para-NMe2-substituted ligand L-NMe2 (Fig. 1) was
synthesised from the secondary phosphine HP(4-C6H4NMe2)
and CH3C(CH2Cl)3 as detailed in the ESI. It reacted in a 4 : 6
ligand/metal molar ratio with silver para-toluenesulfonate
(AgOTs) in CDCl3-CH3CN (4 : 1) to give a species with a
surprisingly complicated 31P NMR spectrum. 32 signals could
be resolved at or below 25 1C (Fig. 2). At higher temperatures
these signals coalesced so that at 60 1C the spectrum showed a
single P environment coupled to 107Ag and 109Ag nuclei. This
behaviour contrasts markedly with that of the unfunctionalised
ligand L, which under the same conditions gives a spectrum
with a single P environment coupled to 107Ag and 109Ag nuclei
(as expected for a symmetrical adamantoid cage) and does not
vary with temperature (Fig. 1a).
a Centre for the Theory and Application of Catalysis, School of
Chemistry and Chemical Engineering, Queen’s University Belfast,
David Keir Building, Stranmillis Road, Belfast BT9 5AG.
E-mail: s.james@qub.ac.uk
b School of Chemistry, Newcastle University, Newcastle upon Tyne,
UK NE1 7RU
Block-shaped single crystals were grown from a CDCl3-CH3CN
(4 : 1) solution of the tosylate complex by vapour diffusion of
benzene and were determined by X-ray diffraction to contain a
pseudo-adamantoid cage of the expected type but missing one
face-capping anion, i.e. [Ag6(L-NMe2)4(OTs)3]3+ (Fig. 3)z rather
c Department of Chemistry, University of Southampton, Highfield,
Southampton, UK SO17 1BJ
w Electronic supplementary information (ESI) available: Synthesis
of L, analytical details, details of modelling. CCDC 818383–818385.
For ESI and crystallographic data in CIF or other electronic format
see DOI: 10.1039/c2cc17832j
than the expected ‘complete’ cage [Ag6(L-NMe2)4(OTs)4]2+
.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 4061–4063 4061

View Online
Fig. 2 Variable-temperature 31P{1H} NMR spectra for [Ag6(L-
NMe2)4(OTs)3]3+ in CDCl3-CH3CN (4 : 1). The coloured symbols
corrrespond to the assignments shown in Fig. 4.
Fig. 4 Illustrative assignment of the 32-line 31P{1H} NMR spectrum
of [Ag6(L-NMe2)4(OTs)3]3+ which results from two pairs of mutually
coupled P environments (PA coupled to PB and PC to PD) with further
coupling to 107Ag and 109Ag. J values are in Hz.
There are two types of P centre (shown in green and black)
around the uncapped face because of the twisted C3 symmetry.
The remaining P centres attached to these ‘rim’ triphosphines
(shown in blue) and the basal triphosphine (red) provide two
further chemically distinct P centres. The four P centres cannot
be assigned to specific signals in the NMR spectrum and so a
largely arbitrary assignment has been made in Fig. 4b, for
illustrative purposes, such that red = PA, blue = PB, black = PC,
green = PD. Inequivalent P centres should couple via the silver
centres (2JPP) but through-backbone coupling (4JPP) is often not
observed with such ligands since it is close to the NMR resolution
limit. PA and PB are mutually coupled (2JPP = 146 Hz) as are
PC and PD (2JPP = 133 Hz). This P-P coupling was further
confirmed by a 31P-COSY spectrum (Fig. S5, ESI). There are
Fig. 3 X-ray crystal structure of [Ag6(L-NMe2)4(OTs)3]3+. Anions
are shown as spheres and the cage core as capped sticks. Ag = light
blue, P = pink, S = yellow, O = red. H-atoms omitted for clarity.
This incomplete cage consists of a distorted octahedral Ag6
core whose faces are alternately capped by the four L-NMe2 ligands
and three tripodal anions, with one face remaining uncapped. The
non-bonded Agꢁ ꢁ ꢁAg distances of the uncapped face are
slightly greater than those in the capped faces (see ESI) due to
the lack of a capping ligand. An analogous crystal structure was
2
limited literature data for JPP coupling via Ag (in a related
cage based on triarylphosphines 80 Hz was observed1a). The
determined for the nitrate complex, i.e. [Ag6(L-NMe2)4(NO3)3]3+
,
107
1
2
whereas crystals of the triflate complex were found instead
to be of the conventional complete adamantoid type, i.e.
[Ag6(L-NMe2)4(OTf)4]2+ (see ESI).
The 32-line solution-state 31P NMR spectrum of the AgOTs
complex is fully consistent with the incomplete cage structure
observed in the solid state. In particular, this idealised C3 symmetry
structure has four different P environments as indicated in Fig. 4a.
resulting eight signals are all further coupled to Ag (I = )
and 109Ag (I = 12) nuclei to produce the 32-line spectrum. The
magnitudes of 1J(31P-109Ag) (PA, 569 Hz; PB, 550 Hz; PC, 572 Hz;
PD, 584 Hz) are all characteristic of AgP2 centres.3 Once the P-P
coupling has been identified, the differences in intensities
between the signals can be recognised as roofing effects
(Dn/J = 5.6 for PA-PB and 1.7 for PC-PD). The averages of
c
4062 Chem. Commun., 2012, 48, 4061–4063
This journal is The Royal Society of Chemistry 2012

View Online
the four P chemical shifts and their silver-phosphorus coupling
constants (ꢀ10.97 ppm and 568 Hz respectively at 25 1C) are
in reasonable agreement with the corresponding values for the
simpler spectrum seen at 60 1C (ꢀ11.50 ppm and 562 Hz),
suggesting that the latter results from coalescence. This is most
likely to be due to scrambling of the anions around the structure.
For the analogous system with AgNO3, the spectrum at
room temperature resembled closely the high-temperature
spectrum of [Ag6(L-NMe2)4(OTs)3]3+, i.e. a slightly broad
single P environment (ꢀ8.95 ppm) with coupling to Ag
(1J31P-109Ag = 590 Hz, Fig. S4). Lower-temperature spectra
showed a number of broad signals indicating several P environ-
ments coupled to Ag. This is consistent with similar solution state
behaviour to the tosylate complex but with faster scrambling of
the anions. However, precipitation prevented a full investigation
at low temperatures. For AgOTf, the behaviour was again
consistent with the AgOTs complex, i.e. it gave a single P
environment at 40 1C (ꢀ9.35 ppm, 590 Hz) and more complex
spectra at lower temperatures. The expected 32 lines could not be
fully resolved on cooling, however, due to signal overlap (up to
27 lines could be distinguished, Fig. S2 and S3). Interestingly, at
ꢀ25 1C just two of the four P signals had coalesced, before full
coalescence of all four occurred by 40 1C (Fig. S3). This can be
explained by flipping between C3 enantiomeric forms becoming
rapid on the NMR timescale at ꢀ25 1C (which would render
equivalent all six P centres around the open rim, Fig. S3).1a
Polar coordination cages identified in solution are rare, but
include a saturated mixed-ligand Pd4LL’ structure4 (where L
and L’ are different tetrapodal N-donor ligands) and an
unsaturated (homo-ligand) Ag4L3 structure (where L is a
tripodal triphosphine) together with its capped derivative
Ag4L3L’ (where L’ is a tripodal trinitrile).1a In these cases
the adoption of polar structures can be related most likely to
steric effects which either induce a preference for hetero-ligand
complexation (i.e. self-selection of one ligand is disfavoured by
its steric bulk) or which prevent a complete cage from forming. In
this work, a specific reason for the subtle differences between
L-NMe2 and L (and previous para-functionalised analogues
where the substituents are Ph or CH2N(CH2CH2)2NCH3) is less
clear-cut. We note that the behaviour is consistent with both the
In summary we note that a remote structural change to a
ligand building block switches its preferred mode of self-
assembly away from the normal, saturated, non-polar T-symmetry
coordination cage to an unusual, incomplete, polar C3-symmetry
structure by disfavouring coordination of one of the four anions.
Despite the subtlety of this structural preference, it occurs with
remarkable selectivity as shown by 31P NMR spectroscopy. As
noted above, because of their directionality, polar assemblies
present interesting possibilities for 2nd level self-assembly into
yet larger discrete aggregates.1i Studies towards this goal are
ongoing with these systems.
We acknowledge EPSRC for funding the National Crystallo-
graphy Service (Southampton and Newcastle), and STFC for
access to synchrotron facilities at Daresbury Laboratory.
We are grateful to the referees of a previous incarnation of
this paper.
Notes and references
z Crystal data for [Ag6(L-NMe2)4(4-CH3C6H4SO3)3][4-CH3C6H4SO3]3:
C254H318Ag6N24O18P12S6 (excluding unmodelled disordered solvent),
%
Mr = 5206.68, cubic, space group Pa3, a = b = c = 39.4527(8) A,
V = 61409(2) A3, Z = 8, l = 0.71073 A, m = 0.53 mmꢀ1, T = 120 K;
149637 reflections measured, 22890 unique, Rint = 0.1093; R = 0.111
(F, 6751 data with F2 > 2s), Rw = 0.361 (F2, all data), GOF = 0.892,
1421 refined parameters, final difference map extremes +1.28 and
ꢀ0.62 e Aꢀ3
.
1 (a) J. Y. Zhang, P. W. Miller, M. Nieuwenhuyzen and S. L. James,
Chem.–Eur. J., 2006, 12, 2448 and references therein;
(b) J. S. Mugridge, R. G. Bergman and K. N. Raymond, Angew.
Chem. Int. Ed., 2010, 49, 3635; (c) D. Fujita, A. Takahashi, S. Sato
and M. Fujita, J. Am. Chem. Soc., 2011, 133, 13317; (d) W. Meng,
B. Breiner, K. Rissanen, J. D. Thoburn, J. K. Clegg and J. R.
Nitschke, Angew. Chem., Int. Ed., 2011, 50, 3479; (e) A. Westcott,
J. Fisher, L. P. Harding, P. Rizkallah and M. J. Hardie, J. Am.
Chem. Soc., 2008, 130, 2950; (f) A. Stephenson, S. P. Argent,
T. Riis-Johannessen, I. S. Tidmarsh and M. D. Ward, J. Am. Chem.
Soc., 2011, 133, 858; (g) Y. Yamauchi, Y. Hanaoka, M. Yoshizawa,
M. Akita, T. Ichikawa, M. Yoshio, T. Kato and M. Fujita, J. Am.
Chem. Soc., 2010, 132, 9555; (h) W. B. Wang, J. Huang, S. L. Xiang,
Y. Liu, J. Y. Zhang, A. Eichofer, D. Fenske, S. Bai and C. Y. Su,
Chem. Commun., 2011, 47, 3849; (i) J. Zhang, M. Nieuwenhuyzen,
J. P. H. Charmant and S. L. James, Chem. Commun., 2004, 2808.
2 We are grateful to a referee who noted that polar coordination cages
are relatively common in the area of single molecule magnetism. We
suggest that this may relate to the fact that such paramagnetic
assemblies are predominantly characterised in the solid state, and
that selective crystallisation relates to solubility effects as well as
speciation in solution.
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) P. W. Miller, M. Nieuwenhuyzen, X. Xu and S. L. James,
Chem. Commun., 2002, 2008N; (d) N. Giri and S. L. James, Chem.
Commun., 2011, 47, 1458.
4 K. Kobayashi, Y. Yamada, M. Yamanaka, Y. Sei and K. Yamaguchi,
J. Am. Chem. Soc., 2004, 126, 13896.
5 pKa values for PPh3 and P(C6H4-4-NMe2)3 are 2.73 and 8.65
respectively: T. Allman and R. G. Goel, Can. J. Chem., 1982,
60, 716.
5
more electron-donating character of L-NMe2 as well as its
greater steric bulk. Specifically, anion coordination may be
rendered weaker in complexes of L-NMe2 because of the more
basic character of the triphosphine, and the consequently
lower Lewis acidity of the Ag centres. The greater steric
crowding (exacerbated by the coplanarity of the C6H4 and
NMe2 groups) could also help to stabilise the unsaturated
structure. Although selective in solution, and to a degree in the
solid state, the effect does seem to rely on subtle energetics,
since one of the cages investigated actually crystallised as the
complete structure [Ag6(L-NMe2)4X4]2+ (X = OTf), suggesting
that crystal packing forces are of similar magnitude.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 4061–4063 4063